1 Friday, 21 October 2022 2 (10.00 am) 3 PAUL RICH 4 Questioned by MR BLAKE (continued) 5 SIR WYN WILLIAMS: I can hear what goes on when the usher 6 makes her provisional announcements but I didn't hear 7 the fire alarm. There we are. 8 MR BLAKE: Mr Rich is ready. 9 Yesterday we spoke about the evaluation board and 10 ComPEC and MaPEC. I'm going start today in early 1996 11 and can we look at POL00031278, please. That's an email 12 to you of 10 January 1996. Can we look at the top half 13 of that page, please. 14 There is a meeting with Cardlink. Basil Shall, 15 and Wendy Powney met with Cardlink. Who are they? 16 A. They are the -- I think they're IT people and I think 17 they're POCL IT people. I know Wendy well because 18 eventually she worked for me but not at this time. 19 Q. That says: 20 "... it became quite clear that our requirements 21 process is lacking in 3 major areas: 22 "Reconciliation/Role of TIP 23 "Office Balance/Cash accounting process. 24 "Interfaces with the TMS ... 25 "If we don't give them direction (which they are 1 1 desperately seeking) they are very likely to choose 2 an option which is not in our best interests." 3 Do you know who that direction was to: "if we 4 don't give them the direction"? Who would you be 5 directing? 6 A. I don't think it's for me to direct them. I think this 7 is me -- them coming to me because I'm part of the 8 Evaluation Board in my role as partnership development 9 person, then working to Mr Peaple, and they're saying -- 10 I think I explained yesterday we allowed the group IT 11 people and our own IT people access to the suppliers as 12 the bids were evolving so that -- although they weren't 13 part of the formal procurement process and, therefore, 14 didn't form part of the value assessment factor group 15 because they weren't at sufficient arm's length to do 16 that, whereas I think John Meagher and, I think that 17 must be Jeremy Foulkes, are. I think they are part of 18 that team. 19 I think what he was doing was asking me to 20 co-ordinate within the business the relevant experts in 21 that field to try to clarify some of the reconciliation 22 cash accounting interface, the TMS. I can't remember 23 what TMS stands. Role of TIP is transaction information 24 processing. 25 I think -- I can't remember the date. This is 2 1 January '96. I think you'll find in the bundles notes 2 from or to me from relevant financial experts within the 3 business around those reconciliation and accounting 4 issues. So I was sort of promulgating the actions so 5 that his concerns would be addressed. 6 Q. Absolutely. So when we look at the bottom of this page 7 it seems as though the affected area seems to be 8 commercial. So you're looking at it from a commercial 9 perspective at that stage, in relation to those issues, 10 rather than technical; is that right or have 11 I misunderstood? 12 A. Yes, I don't think that's quite right. This is a memo 13 to me, isn't it, if you just go back to the top of it. 14 Q. It is. 15 A. So it's not from me. So: 16 "An early commercial 'must have' supported by 17 Charterhouse ... did not control there end-to-end client 18 product. 19 "This was to ensure ... could not cut POCL out of 20 the transaction." 21 Yes, I understand. Charterhouse were reviewing 22 the overall financial security and probity of the whole 23 thing for us. There are other documents that relate to 24 them, I think, in ComPEC and MaPEC papers, with their 25 input, and I think this is really more about what the 3 1 service specification requirement really was, the SSRs 2 as they became known, which were fed in both from 3 ourselves and from DSS and Benefits Agency to the -- 4 I don't know if it was five suppliers or three suppliers 5 at that stage. 6 Q. If we look at the bottom right-hand corner and those two 7 bullet points, the top one is: 8 "POCL will have to rely on the [Benefits Agency] 9 or the service provider for accurate information for 10 settlement." 11 That was a concern. So it seems as though what 12 POCL would like is to control the end-to-end process; is 13 that a fair -- 14 A. They would certainly need to be able to assure it and 15 accurately because, of course, those processes and the 16 information flows from the system would have not only 17 told the relevant client, like the Benefits Agency 18 whether or not their service had been provided and 19 settled but also would have told our own accounts team 20 that it had been in parallel. 21 I think the -- I think I remember that both -- you 22 see it says "Copy to Ian Gair, Tim Brown, Kevin 23 Corrigan/Byron for comment", Tim Brown was one of the 24 assistant finance directors at the time and he would 25 have assured it and I think I remember that the chief 4 1 auditor, Peter Dent, also had input in the one of the 2 documents I've read. 3 Q. If we move on to February of that year, 1996, can we 4 look at POL00028288, please. This is a meeting note 5 from the Evaluation Board. We can see there that the 6 chair is Bob Peaple and I think you're listed as one of 7 the attendees? 8 A. Yes, as I said yesterday, yes. 9 Q. There are representatives, as you said yesterday, 10 I think, from the Benefits Agency, as well, who attend 11 that meeting? 12 A. Yes, and the Social Security Agency for Northern Ireland 13 and DITA is the Department -- I think that's the 14 Department of Social Security's IT people. 15 Q. Thank you. Yesterday we heard from John Roberts who 16 received updates about the Evaluation Board but these 17 are the actual Evaluation Board's own minutes. Can we 18 look at page 3 of these minutes, please, and, focusing 19 in on the bottom half of the page, that's 2.7 and 2.8. 20 I'm going to read those. 2.7: 21 "With respect to the certificate for Pathway, the 22 Evaluation Board expressed severe reservations about the 23 number of significant risks against the Pathway 24 proposal. It questioned whether the cumulative effect 25 would not lead to potentially late delivery and/or 5 1 operational problems, and as such was cause for Pathway 2 to be excluded from the [Invitation to Tender]. 3 Particular concerns were expressed about the card 4 technology with shortcomings in the associated 5 management of fraud and about the dependence on Escher 6 as a small company subcontracted to Pathway." 7 Who was Tony Johnson? 8 A. He was in the -- I can't remember. I was thinking about 9 this the other day. I can't remember which organisation 10 he came from but he was part of the demonstrator team 11 within the assessment team, the one run by Andrew Stott. 12 Q. So: 13 "[Mr Johnson] explained that these issues had 14 indeed been given much consideration by the Demonstrator 15 team. It had concluded that the Pathway solution was 16 not fundamentally flawed and that from a strictly 17 technical view point the hurdle had been cleared." 18 A. Yes. 19 Q. "In its overall decision, the Evaluation Board might 20 wish to take into account the cumulative effect of any 21 contractual issues with those from the requirements 22 area." 23 So not fundamentally flawed, that's a phrase that 24 we heard yesterday as well. That's not a ringing 25 endorsement of the system, is it? 6 1 A. No, it's not a ringing endorsement, I agree with that, 2 but it was a statement we relied upon at the Evaluation 3 Board, noting the risks, among others, that are in 4 paragraph 2.7. It was -- and as I also explained 5 yesterday, I think, that when this eventually went to 6 MaPEC Major Projects Expenditure, for the main Post 7 Office Board, the paper would have had to have group IT 8 technical concurrence -- 9 Q. Absolutely. 10 A. -- and they said it is acceptable even if it's not 11 ideal. 12 Q. In terms of the three potential options, Pathway was the 13 least technically strong of the three? 14 A. Yes, I think I know what you're referring to because it 15 said the evaluation came third in -- I can't remember 16 how many, but each one cleared the hurdle and each one 17 had their own issues, is the truth. So you had to apply 18 an overall judgement about the evaluation not only on 19 its technical merits, which had to be acceptable and 20 know the risks you were going into in trying to manage 21 those, but also the other aspects from the other issues 22 around operational management, contractual and PFI 23 compliance. 24 Q. Also highlighted in the section I've just read are 25 concerns about the dependence on Escher as a small 7 1 company -- 2 A. Yes. 3 Q. -- and also concerns about the card technology at that 4 stage. 5 A. Yes. The card technology was I read -- forgive me if 6 I've got his name wrong -- Mr Cipione's technical -- 7 I was given access to his witness statement a couple of 8 days ago. You have to remember what the technology was 9 like at that point. You know, it was a very different 10 world then. So ICL was relying on magnetic stripe 11 technology for this. 12 We looked at and considered something that came up 13 from them, which was a card that included an integrated 14 circuit, known these days more as a smartcard and, at 15 that time in this country, that was very new technology, 16 more expensive, quite risky of itself, gave Benefits 17 Agency some concerns about the security aspects 18 associated with having those cards for benefit 19 encashment. So we decided that would be more risky. 20 As for the dependence on Escher, yes, I understand 21 that as well. We knew that they had a proprietary bit 22 of software called Riposte and it hadn't been proven at 23 scale because the demonstrator was, I think, an office 24 in Ireland, basically, if I remember rightly. 25 But then -- 8 1 Q. I think it may have been in the United States. 2 A. Okay, I thought they went to see something that they 3 trialled in An Post but I may misremember, forgive me. 4 Q. No, no, I think you may be right. 5 A. I think they are an American company, I think. 6 Q. Absolutely. 7 A. We knew that risk and we weren't shy about telling ICL 8 Pathway about that risk and they would have to address 9 it if they were going to go forward and, hence, why we 10 put the mitigating stuff around the risk later on. 11 Q. Can we look at 2.10, so that's over the page. Derek 12 Selwood: who is Derek Selwood, sorry? Are you able to 13 assist? 14 A. He's, again, part of the assessment team. 15 Q. "Derek Selwood confirmed that risks and issues against 16 service providers would be taken into account in the 17 evaluation and selection process. The risks against 18 Pathway would incur a substantial cost penalty given 19 their number and severity. The issues identified during 20 the Demonstrator would affect the Value Factor 21 assessment, and it would be for the Evaluation Board to 22 give due weight to that in reaching its decision." 23 Now, sometimes a proposal might just make the cut 24 but everybody in the room knows that, ultimately, the 25 penalties will mean that it's ruled out. How did you 9 1 feel at that time? Did you have any feelings about 2 Pathway? 3 A. I knew there were some risks with it, as there were with 4 others. It was very difficult, really, at that time, to 5 go out and say, "Oh, look, here's this system somewhere 6 else that's been done before", because there was nothing 7 like it and there was certainly nothing like it done 8 under a private finance initiative. So we were all, 9 I think it's fair to say, learning because you couldn't 10 say -- IBM couldn't show us -- or Cardlink couldn't show 11 us "Here's something we've done extensively like this in 12 the States and also transferred the risk of a large part 13 of fraud at the same time". 14 So you had to do that but -- I don't think it was 15 with reluctance that we chose them. We were looking to 16 try to secure an automation provider for both us and for 17 the Government. 18 Q. While we're on this document, can we just look at 19 paragraph 2.12 below. 20 A. Yes. 21 Q. I don't need to read out the paragraph there but it 22 suggests there was some confusion over the requirements 23 for EPOS at that stage. Do you remember that? 24 A. I'm trying to remember, sorry. I'm just reading the 25 material. (Pause) 10 1 I think this was about when and where EPOS would 2 be down the track, in terms of its releases. As I said, 3 we put in the requirements that it had to be -- the 4 solution had to be able to replicate existing processes 5 but we knew that it would need to migrate to better 6 automated processes later on. We were trying to give 7 priority to the DSS service, the benefit encashment 8 service. 9 Q. Absolutely. That's the very point that I make, that at 10 that stage the non-Benefits Agency part was very much in 11 its early thinking when it comes to something like -- 12 A. I think we'd fleshed it out but we also realised that it 13 might change over time, depending on what other clients 14 might want to be taken on, for example. There were 15 competitive pressures from competitors, which the 16 Federation of SubPostmasters weren't slow to tell us 17 about, from people who were -- later on, you've seen it 18 in other documents, were putting out -- a competitor was 19 putting out an automated payment terminal for bill 20 payments, which the subpostmasters were waiting for 21 Horizon to do. 22 Q. But EPOS itself was pretty fundamental to the 23 non-Benefits Agency part of Horizon? 24 A. Yes, and the join-up with the back office systems. 25 Q. It seems that in early 1996, at least, it was only very 11 1 much in its early stages and there was, according to 2 this minute, some confusion over what was required. Do 3 you agree with that? 4 A. I don't remember it, is the truth, but that's what it 5 says here so I take it as read. 6 Q. Can we move on to a board paper. It's POL00031237. In 7 fact, this document starts as a board paper but can we 8 go to page 9. 9 A. Can you just tell me what date that is, please? 10 Q. The document on page 9 will assist, actually. 11 A. Thank you. Oh, yes I remember this now. Thank you. 12 Q. I'm just looking for the date. The date of the -- 13 A. No, I understand. This (unclear: simultaneous speakers) 14 the Evaluation Board more or less immediately while -- 15 once the recommendation had been made and mandates were 16 being sought. 17 Q. So this is a minute from Bob Peaple to the Project 18 Steering Committee? 19 A. Yes. Yes, we knew it as the Joint Steering Committee, 20 so I'm not quite sure about that but fine, okay. 21 Q. Paragraph 1 sets out the "Purpose". So: 22 "The purpose of this minute is to inform you of 23 the substance of the meeting of the Evaluation Board 24 which I chaired yesterday to consider the report of the 25 evaluation team on the re-tenders submitted by Tom, Dick 12 1 and Harry." 2 We went over this yesterday with Mr Roberts. 3 I think you heard Mr Roberts' evidence? 4 A. Some of it, not all of it. 5 Q. So Harry was Cardlink, Tom was IBM and Dick was Pathway. 6 Can we look at paragraph 10, please. 7 A. Peter Mathison is the chief executive of the Benefits 8 Agency, by the way, he's not a Post Office person. 9 Q. Thank you. I don't need to read paragraph 10 but 10 essentially there were some wrinkles and they needed to 11 issue an invitation to re-tender on 16 April; do you 12 remember that? 13 A. I can remember there was some iteration. I don't 14 remember the detail I'm afraid. 15 Q. That's fine. Can we look at paragraph 13, which is over 16 the page and could we focus in on that paragraph, 17 please. 18 That addresses the technical aspects and, as you 19 said earlier, on a technical level, all three satisfied 20 the minimum levels; is that right? 21 A. Yes. 22 Q. Can we look at paragraph 16, that's over the page. It 23 was ICL that was closest to the risk transfer sought; do 24 you remember that? 25 A. ICL, as far as I can remember, were the only ones that 13 1 met that hurdle. I think Cardlink was eliminated on 2 other factors and then IBM were only prepared to take 3 risk at a pound for pound, and I think -- I may have 4 these numbers wrong, so forgive me, but it will be right 5 in the order of magnitude. I think Pathway were 6 prepared to put up about 200 million of risk transfer. 7 Q. You're absolutely right. If we go to paragraph 30 to 32 8 on page 15, please. Would it be possible to highlight 9 this? Thank you very much. 10 Essentially, Pathway (Dick) was closest to what 11 sponsors had sought to obtain, particularly in terms of 12 the risk transfer; is that right? 13 A. Sorry, that was a question? Yes, yes, it was. Sorry. 14 Q. Paragraph 33: 15 "The Board recognised that an award to Dick would 16 imply a need for a proactive management stance by 17 sponsors." 18 What did you understand by that? 19 A. Well, any set of risks needs managing and it means that 20 both sponsors had the need to put forward a suitable 21 method of managing those risks. The way we decided to 22 do that was, as I said yesterday, to create this 23 programme delivery authority, jointly staffed by BA and 24 POCL people in the same building, working together -- 25 which was an interesting cultural dynamic -- headed up 14 1 by a DSS person to give the client reassurance and 2 because they had skills, supplementing that with 3 external contractors who had technical skills where 4 necessary and then create -- I don't know what -- BA and 5 DSS had their own project team, I think, also because 6 I used to meet my equivalent, as it were, on what became 7 the PDA board, but we also set up teams within Post 8 Office in parallel to check what the PDA was doing and 9 if there were any commercial or contractual issues. 10 I say that about the commercial contractual issues 11 because, clearly, the contract hadn't been let then and 12 there was still this to come and then we had to have -- 13 we had to finalise our agreements with Benefits Agency 14 on the back of that because that was a back-to-back 15 arrangement. But we also had to cope with the nature of 16 the private finance initiative in that, in what 17 proactive management meant. 18 Q. If you are looking for who is responsible for taking 19 forward that proactive management stance, who was that 20 then? Was that the programme delivery authority or was 21 it a particular individual? 22 A. Well, the programme delivery authority on behalf of the 23 two sponsors, and that reported to a Project Steering 24 Committee, confusingly here, that's why I differentiated 25 before, which consisted of the chief executive of ICL, 15 1 the managing director of POCL, and the chief executive, 2 I think, of the Benefits Agency being supplied with 3 information. So if there were issues to resolve, that 4 programme steering committee was meant to resolve them. 5 Q. So there are lots of committees. 6 A. There are. 7 Q. We've heard about lots of committees. Who within the 8 Post Office would have been responsible for taking 9 forward the proactive management stance? 10 A. Well, the managing director essentially would have been 11 mandated by the board. He delegated quite a lot of that 12 to me to do things. He would have also taken 13 a proactive stance, as I did, with the person running 14 the programme delivery authority. 15 Q. So that's Stuart Sweetman as managing director -- 16 A. He was by then, yes, I think. 17 Q. -- yourself, and then somebody within -- 18 A. Peter Crahan was the programme director, having taken 19 over from Andrew Stott, I can't remember the exact date. 20 He was another DSS person. George McCorkell in the 21 Benefits Agency, and the PDA board was chaired by -- 22 ably, I may say -- Alec Wylie, who was the chief 23 executive of Social Security Agency for Northern 24 Ireland. 25 Q. Can we go back to the document that we were just on but 16 1 look at page 1, which is the note for the board. 2 Paragraph 11 on the second page, I won't read it 3 but it sets out the financial evaluation -- 11 and 4 below -- and then can we go to the next page, and 5 paragraph 14 is the "Non-financial evaluation". One of 6 the headings there is "non-financial characteristics" 7 and it says: 8 "this reviewed suppliers' performance against 9 a number of characteristics, including customer 10 acceptability, reliability and support, managing 11 capability, etc. All three suppliers exceeded the 12 acceptable level with the differences between them not 13 significant for the purpose of discrimination." 14 I think that's consistent with what you told us 15 earlier about the three? 16 A. Yes. 17 Q. There isn't mention there, at least, of concerns about, 18 for example, the dependence on Escher or the card 19 technology issue? 20 A. No, hadn't they been mentioned earlier or in the note? 21 Q. In fact, it mentions it slightly further down, so 22 paragraph 15 that we can see there, "The Programme 23 Evaluation Board recommended Pathway", and that was 24 endorsed by the Joint Steering Committee. 25 Then it's paragraph 18, I think, that is the part 17 1 of this particular note for the board that addresses the 2 other technical risks in a little bit more detail. 3 Could we have a look at paragraph 18? 4 A. Yes, I see, okay. 5 Q. "Some technical risks were identified with all 6 suppliers, and in some areas, Dick was considered to 7 have higher technical risks that Tom and Harry. However 8 these risks are manageable through; 9 "a strong technical assurance function, with 10 support from the Post Office IT Directorate, 11 "rigorous testing at development, trial and 12 roll-out stages, 13 "ensuring supplier contingency plans, 14 "a proactive technical management plan." 15 Is that the proactive management stance that we 16 talked about earlier that Bob Peaple was referring to? 17 A. Essentially. I think I just fleshed it out a little bit 18 more in what we actually did, rather than what we 19 thought was needed to do, as it were, and it also, 20 depended on the strength of the people we both put in to 21 that programme delivery authority when we staffed it. 22 But I think for the last, certainly, points 2 and 23 3, I think that's evidenced in all the bundles I've 24 read, so far. 25 Q. So the Post Office IT Directorate were going to take 18 1 responsibility for some of the technical assurance 2 function? 3 A. We would have placed people from within the Post Office 4 IT function either from within Post Office Counters or 5 from group IT into the PDA and, at the same time, they 6 would have been assuring the board themselves through 7 the group IT director or to whom he reported at board 8 level, executive board level that -- 9 MR BLAKE: Can I just pause you there? 10 I am being told that the transcript has stopped. 11 We may need to pause for a minute or two just while that 12 is rectified. 13 Perhaps we can take a short break. I know it's 14 very early already but, seeing as this might take a few 15 minutes, that could be our mid-morning break because we 16 have a lot to do today. 17 SIR WYN WILLIAMS: Whatever suits best, Mr Blake. 18 MR BLAKE: Can we have a ten-minute break now? 19 SIR WYN WILLIAMS: Yes. 20 MR BLAKE: Thank you very much. 21 (10.32 am) 22 (A short break) 23 (10.41 am) 24 MR BLAKE: Thank you, Chair, we can see you now. 25 I won't go back to that document but I think the 19 1 bottom line, from what we just heard this morning is 2 that Pathway got through the process but it was very 3 clear that they needed close management; is that a fair 4 summary? 5 A. As far as the -- as within the remit of the PFI and 6 I think what became clear later on and there are 7 documents in the bundle that their idea of what PFI was 8 slightly different to what the sponsors was. I think 9 there's a letter from Keith Todd later on in March '98 10 to Peter Mathison, chief executive of Benefits Agency, 11 that spells out what they believed PFI meant and the 12 level of unnecessary interference, therefore. 13 Q. Can you summarise that very briefly then. Was it your 14 position that you could have more involvement and their 15 position that you shouldn't have the level of 16 involvement that you were seeking? 17 A. Yes. Yes, if you want to summarise that briefly, "yes" 18 is the right answer. I think it was really around the 19 extent to which they could, under the PFI -- private 20 sector are innovative and can bring their skills to the 21 party -- could take the business processes and we would 22 have to rely on the outputs rather than the details of 23 the design architecture underneath it. 24 Q. Can I just take you to your witness statement. It's 25 WITN04030100 and it's page 16 that I'd like to look at? 20 1 A. This is my witness statement? 2 Q. Yes. 3 A. What paragraph? 4 Q. It's paragraph 47. You say there: 5 "Some post offices were not easily suitable for IT 6 infrastructure, as they did not have the necessary space 7 or equipment. Moreover, Pathway underestimated the 8 amount of new software development needed from its 9 subcontractors, eg Escher. A further issue was that the 10 BA/DSS [CAPS] which was to feed data to Pathway's card 11 systems, was not ready ..." 12 That sounds very much like the problems that were 13 raised at the Evaluation Board pre-Invitation to Tender 14 that we saw first thing this morning, don't they? 15 A. Some of them. One of the three but not the other two. 16 Q. There's the issue with Escher and also the card 17 system -- concerns about the card system. 18 A. No, I don't think that says -- I'm sorry, Mr Blake, 19 I don't think that says concerns about the -- 20 Q. It's the next page, sorry. 21 A. Ah yes, I see. Thank you. Sorry. 22 Q. That paragraph begins by talking about the 23 infrastructure. Keith Todd may, in due course, say that 24 Post Office couldn't reasonably have believed that their 25 premises were fit for automation and that ICL weren't 21 1 given a proper opportunity to inspect. Would you agree 2 with that? 3 A. I said it was a learning process and, certainly, one of 4 the things we knew, but we learnt more systematically, 5 if I can use that word, was that of our 19,600 post 6 offices they're not a heterogeneous set. You know, you 7 had Crown offices with 20 counter positions and a front 8 room in the Orkneys. I think what Pathway 9 underestimated, reasonably, was the amount of the 10 network that wasn't covered by ISDN. For those of 11 a certain age, that's a sort of predecessor to 12 broadband, I suppose, on which their system relied for 13 messaging and, therefore -- and typically in the 14 smallest offices, of course, or in the most rural ones 15 where BT, at that time, or Energis hadn't done. 16 So they didn't know that until they surveyed and 17 we wouldn't have been able to tell them that in this 18 statement of service requirement and, of course, the 19 ergonomics as well about counter space and, hence, why 20 we always insisted that both customer acceptability and 21 how long transactions would take -- in case it affected 22 our quality of service, queueing time, for example, in 23 bigger offices, and user acceptability, in terms of 24 agents or staff being able to use this, both physically 25 and in terms of software -- were always included. 22 1 I think that's probably why, when we first 2 realised we wouldn't be able to achieve post contract 3 what we set out to, there was something called 4 a no-fault re-plan because the other thing that was 5 mentioned in there -- I don't know if you're going to 6 ask me about that -- was the DSS issue. 7 Q. Yes, I'll come to the no-fault re-plan shortly. You 8 also highlight in your statement that Pathway had 9 underestimated the training time required. 10 A. That's true. 11 Q. You said, in terms of infrastructure, their 12 underestimation may have been reasonable. Was it 13 reasonable in respect of training time? 14 A. We thought they should have known that it would take -- 15 as Mr Cipione said, we were acquiring a system not 16 a piece of software and that system is about people and 17 the human interface with that. That was always at the 18 heart of our business, unlikely as it may sound now. 19 But the training that they set out, which they 20 did -- which they were contracted for, for part of that 21 system, which they subcontracted to a company called 22 Peritas, from memory, probably wasn't as extensive 23 enough and, certainly, we pushed back hard at that once 24 we had that pointed out to us, engaging with the 25 Federation and with subpostmasters and -- et cetera. 23 1 Q. Is that something that thought was given to during the 2 earlier stages: the evaluation stage, the Invitation to 3 Tender stage? 4 A. I think -- did we give it enough thought? That's a good 5 question. Certainly, they had to demonstrate 6 capability. You know, it wasn't like going to 7 a Microsoft and saying "I'll have that computer". It 8 was buying a whole system with all that went with it, 9 including putting it in because we were an unautomated 10 business of 19,600 offices with 70,000 people involved 11 and it doesn't take a genius to work out that that 12 culture change is going to happen overnight without 13 professional training. 14 I've read other documents later on, after my time, 15 when roll-out began, where I've only seen those since 16 you've sent me -- the Inquiry sent me the bundles as 17 I wouldn't have been there -- I've seen that, even in 18 late '99, one of our senior managers was writing to 19 Pathway saying "The training is still inadequate". 20 Q. Moving to February '97 to spring '98, as a member of the 21 PDA board, you received testing reports and something 22 called technical concurrencies. Can you briefly 23 describe what technical concurrencies were? 24 A. Can you show me had -- they mean different things at 25 different contexts, that's all. 24 1 Q. I only need a very brief explanation. 2 A. Technical concurrencies meant is this stuff fit for 3 purpose and is it working as we think it should or it is 4 likely to work as we think it should and is it capable 5 of working as we think it should? Certainly, as I've 6 described before, the Post Office Board was relying on 7 technical concurrence from our own people, from the 8 group IT people downwards, in order to assure it's 9 approval of the business case. 10 Q. You, explained in your statement at paragraph 54 and 55 11 that there were delays in spring '97 and into 1998. One 12 of those reasons was an underestimation by Pathway of 13 the effort and time needed to develop the services and 14 the other was an issue with Escher. 15 A. Yes. 16 Q. The Escher point again, something that we've heard a lot 17 about this morning, mentioned that pre-ITT stage. 18 A. Yes. It becomes a running sore is the truth. I know 19 there were -- it wasn't like ICL were ignoring it. You 20 know, they were trying to fix it and they were putting 21 more and more people, technical people, onto it. 22 I think the National Audit Office report says this 23 as well, as well as PA, that ICL thought there'd be more 24 systems integration than systems development involved. 25 You know, they thought they could take pieces of the 25 1 jigsaw and slot them in but when they tried to slot them 2 in and then tried to scale them up, it needed systems 3 development as well as integration skills, particularly 4 around complex business rules like the DSS. 5 Q. In terms of Pathway's underestimation of the effort and 6 time that was needed, was this proactive management 7 stance implemented at that stage? Do you think that ICL 8 were being effectively managed? 9 A. I think we were scrutinising them and I think if we 10 hadn't have done we would have been tempted to say, 11 "Okay, that's all right, go ahead then" and we never did 12 that. Personally -- and you have got evidence in the 13 bundles, I would never compromise quality for speed, and 14 I think I can -- we can demonstrate that by the number 15 of replans, both in terms of roll-out to offices and to 16 what was in the various releases that there were 17 following the Benefits Agency one, where we kept 18 staggering them and simplifying them to give it its best 19 chance. 20 But we wanted to try to fix before rolling out and 21 that was why the approach was taken about an initial Go 22 Live of one office and then of ten offices and then 23 a 200-type roll-out before any acceptance of the system 24 or roll-out, in order to be able to test because some of 25 the things you can't test unless it operates. 26 1 You can test some things, model office testing and 2 end testing, as it's called, or user acceptability 3 testing, all three of those, but until you actually put 4 it in action in a safe environment, if I can put it that 5 way, so you can tell with real customers and real agents 6 using it, you need to understand what the result of that 7 was. 8 Q. You talk in your statement -- it's paragraph 57 -- about 9 the causes of delay and roll-out being the same as the 10 causes of the no-fault re-plan. 11 A. Yes. 12 Q. Very briefly, what were those causes, so far as you saw 13 them? 14 A. Inability to demonstrate to us completely -- to the Post 15 Office, that it was able to work well in offices before 16 rolling it out, an inability, from the DSS point of 17 view, to securely do the benefit encashment service as 18 far as they were concerned. That was their end testing. 19 Two other points, just in support. Complexities 20 around the installation of the network and the 21 ergonomics that I mentioned earlier but ICL were more 22 aware of that by then, so by the time after the '96 23 re-fault plan, when the new master plan was -- there's 24 a document that describes it in March '97, which was 25 done for all parties as a re-plan, and then there was 27 1 disappointment after that because ICL failed to deliver 2 again to time. 3 But there was also the DSS issue of whether their 4 own systems were fit for purpose and had been 5 progressing as quickly as ICL believed to feed the 6 system. I think Keith Todd, in a letter, called -- said 7 to Peter Mathison said that he believed DSS had 8 "misrepresented" the state of the CAP systems. 9 Q. You talk about disappointment. I'm going to move to the 10 summer of 1997. Can we look at POL00039669. Now, this 11 is a new document, in the sense that it has only very 12 recently been shown to you, so if you need more time to 13 look at this document, please do let me know. 14 A. Can I have a look? It hasn't come up on the screen. 15 Q. You are going to need more time than that! 16 It should be familiar now and it's only one 17 sentence that I'd like to take you to. 18 A. You gave me this at 2.00 yesterday. 19 Q. Indeed. 20 A. Sir, I haven't fully reflected on it, I have to tell 21 you, Chair, I'm sorry about that, but I read it 22 overnight. 23 Q. It's the part just below "On Pathway" that I wanted to 24 draw your attention to. This is a letter to you from 25 Stuart Sweetman. 28 1 A. Yes, it's an internal memo because I was going on 2 holiday, yes. 3 Q. It says: 4 "The technical release slippage is a cause of 5 major concern for the [Benefits Agency], not so much for 6 its direct impact but because it is a 'very poor' signal 7 of ICL's capability, so soon after a re-plan." 8 Am I right to say it sounds as though you are 9 quite concerned by that stage? 10 A. I'm concerned and certainly DSS are. I think this is 11 the time at which we asked -- I think you will see in 12 Peter Copping, later on today -- we asked PA to come in 13 and do an independent review of where we were to see -- 14 in mid-'97, to check because, you know, one slippage on 15 a no-fault re-plan, soon after contract letting, as we 16 are all learning is one thing but then for ICL not to be 17 able to do what they said they would do at that point 18 three months later is another. 19 Q. Why would you raise it with Stuart Sweetman? 20 A. Because Stuart needed to know because he was part of the 21 programme steering committee. You know, I said earlier 22 that major issues would have had to be resolved and 23 I would have always kept the managing director, to whom 24 I was a direct report, appraised of the situation 25 because he may well have been dealing with other 29 1 stakeholders like politicians, ministers and people like 2 John Roberts, actually, as well. 3 Q. I'm going to take you to one more document of the same 4 period. This is 6 May, so a few days later. It's 5 POL00039668. I apologise, this is, again, another one 6 of those documents that you have only recently seen. 7 A. It's not on my screen, sorry, Mr Blake. Ah yes. 8 Q. It's from you to Min Burdett. Who's Min Burdett? 9 A. Min Burdett is someone within the technical assurance 10 team. She's a technician working to Mena Rego -- you 11 see Mena there and Mena was direct report to me, not as 12 part of the PDA but within Post Office Counters. Mena 13 had obviously asked her to have a look at the latest PDA 14 documents that Pathway had put forward as part of their 15 proposal to re-plan. 16 Q. Can we look at the penultimate paragraph on that page, 17 please. It says there: 18 "The main concern that POCL has is that it is not 19 clear whether the right tests are in place (either as 20 part of DIT or in the subsequent Model Office)to ensure 21 the integrity of the new Benefit accounting, settlement 22 and reconciliation process." 23 A. Yes. 24 Q. Settlement and reconciliation process, obviously, we 25 know is quite important, quite fundamental, in terms of 30 1 the work of the subpostmaster. 2 A. Yes, and I'm pleased that we were doing the work to try 3 to assure that and absolutely check it out. 4 Q. Were you concerned about it at that stage? 5 A. I was concerned -- I would have been concerned about all 6 aspects of the programme that hadn't been running to 7 time. 8 Q. Can we go to page 3 -- 9 A. I think in paragraph 2 above it shows you that we were 10 prepared to move some -- we weren't prepared, sorry, to 11 accept some of the planning -- squeezing in of new 12 functionality without it being properly tested first. 13 Q. Paragraph 3, the first bullet point, it says: 14 "Assuring that the new Accounting, Settlement and 15 Reconciliation end-to-end process works. This is a POCL 16 'must have' and yet is not clear how POCL is going to 17 know before the go-ahead is given that the new process 18 will work." 19 Again, that must have been quite a concern at that 20 stage. 21 A. It was a concern and if you note the timing of this, 22 which is -- 23 Q. It's 6 May 1997. 24 A. If you note the timing of this, there was a PDA report 25 which I thought was -- sorry a PA report, which 31 1 I thought was fair game, but one of the documents you 2 gave me yesterday showed my reaction to a draft which 3 said it had some gaps. But it was fair game for much of 4 it. Then it also was about the time -- I know you 5 questioned -- I saw that you questioned John Roberts 6 yesterday about the lessons learnt exercise he had 7 asked, which was July, I think, so two months after 8 this, and, you know, I gave him that, as he called it, 9 I think, a warts and all. 10 Q. Absolutely. Shall we turn to that? That's POL00028953. 11 A. We haven't scripted this, have we? So that's good. 12 Q. This is your covering letter. 13 A. This covering letter to my colleagues on the Counter 14 Executive Committee because it was important that they 15 knew about it because it involved actions across the 16 team. 17 Q. Can we look at page 7, please. Can we look at the 18 left-hand side of page 7. 19 Sorry, it's the page before that. Thank you. 20 Perfect. 21 This is a passage that I took John Roberts to 22 yesterday, "there is a need to review enabling 23 organisations", and it says further down there concerns 24 about the end-to-end. 25 No, sorry, over the page, sorry. Sorry, it's the 32 1 page before then. It's internal page 5. It's certainly 2 page 7 of my document. 3 A. Is this the -- what's the "Key Improvement Lessons" it's 4 under -- 5 Q. "We need a better idea of connections to our other new 6 systems", I think this is something you spoke about 7 earlier. 8 A. Yes. 9 Q. Is the concern that's being raised there that there 10 weren't proper systems in place for POCL to understand 11 why Horizon might not be working properly? 12 A. I think that what became clear in this candid report 13 from me was that I said it's a learning exercise for us 14 in becoming an automated organisation. I think I say 15 that somewhere in here, and one of those things was that 16 we had a number of projects in train that weren't to do 17 with Horizon, either existing systems or back office 18 systems, that were going through their own change and 19 what we didn't have in one place was something to join 20 those all up and plug Horizon into it, if I can put it 21 simply like that. 22 So the -- sadly, I can see that most of the 23 actions fall to me. 24 Q. Absolutely. Can we look at the top document that's 25 shown on screen and it's the bottom two bullet points on 33 1 the right-hand side. 2 A. Yes. 3 Q. "end-to-end testing procedures will need to be 4 transferred from the PDA, and supplemented as release 5 planning migrates back to POCL after the system is 6 accepted 7 "a process for live trial acceptance is being 8 devised to ensure collective ownership across POCL 9 functions." 10 So that's your name on the right-hand side. 11 A. Yes. I mean, it is also "after the system is accepted" 12 on the penultimate point which, of course, it hadn't 13 been and wasn't until late '99. 14 Q. Is that penultimate point saying that, once the PDA has 15 finished its work, that end-to-end testing will need to 16 be assured at the Post Office end? 17 A. Yes, for the Post Office systems, not for the DSS 18 systems -- 19 Q. Yes. 20 A. -- which is the other part of the end-to-end for some 21 services, obviously. 22 Q. The final bullet point, live trial acceptance, the 23 processes were going to be Post Office processes. 24 A. Yes, and, as it says there, "to ensure collective 25 ownership across POCL functions". 34 1 Q. Now, your name is on the right-hand side. 2 A. Yes. 3 Q. Does that mean that you were responsible for taking 4 those forward? 5 A. I was responsible at -- you see it says "Paul (ATSG)"? 6 Q. Yes. 7 A. That's the automation transformation steering group, 8 another fantastic name. I think it became known as 9 harnessing technology in post offices later on but it 10 was -- basically I was asked by Stuart to set that up 11 and bring together the project managers of these various 12 technical projects, to be able to talk to each other and 13 create a milestoned plan that ensured they joined up. 14 I mean, at this time we were in the early stages of 15 year 2000, for example. 16 Q. So as things went forward, who would have been taking 17 that forward? 18 A. After I'd left in early '99, that would have fallen 19 within the remit, I imagine -- I can't be certain about 20 this, you'd need to ask -- of Dave Miller, I would have 21 thought. 22 Q. Can we -- 23 A. Dave Smith was the programme manager, the guy who 24 apparently got up that interesting slide yesterday. 25 Q. Can we go on one page after the bottom page here. So 35 1 it's internal 7 but it's my page 9 and we have there, on 2 the left-hand corner -- this is again something I raised 3 with Mr Roberts -- robustness is being raised as 4 an issue there. 5 On the right-hand side: 6 "Collective nerve needed to ensure no compromised 7 on quality for sake of speed and to retain programme 8 focus." 9 That's something that you have told us about 10 already today. 11 A. Yes. 12 Q. Was there some pressure then? Why would this need to be 13 said? 14 A. Because I think -- remember I'm talking both to -- this 15 is a report, essentially, for the John Roberts Counter 16 Automation Steering Group and I've copied it to my 17 Counter Executive Committee colleagues and, therefore, 18 all the main players within the Post Office or within 19 Post Office Counters are recipients of this and I just 20 wanted us to be clear between us that we remained 21 committed to not being expedient for the sake of speed, 22 given the delays. 23 Q. Were there some who just wanted it to happen quickly? 24 A. No, I don't think so. Not within the Post Office, no. 25 But in managing the disappointment of the slippages 36 1 within Post Office and Post Office Counters, there were 2 other stakeholders to manage as well, including 3 Government ministers who'd have been shown timetables 4 before. You referred me yesterday to the Benefits 5 Agency POCL memorandum of understanding, which pre-dated 6 the awarding of the contract and, within that, there was 7 an ambition to have completed the roll-out, as far as 8 DSS were concerned, by the end of '99 and we were 9 nowhere near that. 10 Q. So were the external pressures to speed up? 11 A. Or to take a different approach. I mean, I think I said 12 elsewhere, and it may be at the bottom, I can't 13 remember, at the bottom of this document. 14 Q. Perhaps we can look at internal page 8. It's page 10 15 here. 16 A. "DSS -- political issues", and as my note to Stuart -- 17 as I was flying off to Germany that day -- said, after 18 I'd met George McCorkell for dinner the night, my 19 Benefit Agency equivalent, I think the slippage after 20 the re-plan had dented confidence within DSS, and you 21 have to -- you asked me about our relationship with the 22 Benefits Agency yesterday. Probably more DSS than BA, 23 not quite the same thing now. Different culture than us 24 and much more command and control, much less open. 25 You can see from the style of this document this 37 1 is a very open document. It underpins our -- we had 2 something called Business Excellence, Total Quality 3 culture where no denial was a characteristic and putting 4 the customer first was a characteristic at that time. 5 DSS wouldn't have thought like that. 6 They always, I think, had the ACT option in the 7 background and their systems, which not me but the NAO 8 said weren't ready for purpose at the beginning of 9 the -- were getting more fit for purpose. 10 Q. The threat of termination that's mentioned on that 11 document, though, that must have had some impact on 12 getting the job done quickly? 13 A. Well, clearly we agreed. There's documents that we 14 agreed with the Benefits Agency through the programme 15 delivery authority and its lawyers to put a breach 16 notice in November '97, I think, from memory, so after 17 this, but also committed to keep working on the 18 programme while that was put forward. 19 You heard John yesterday say we had a difference 20 of opinion about whether to follow that up later on with 21 a cure notice, a 13-week cure notice. There will be 22 lawyers here that know better than me what a cure notice 23 is, I'm afraid, but I think it's essentially putting 24 them on notice that, unless they didn't get -- fund this 25 in 13 weeks, there would be termination. 38 1 Q. During that summer there was still some significant 2 technical problems. I'm going to take you to a document 3 POL00028311. That's a programme delivery authority 4 board meeting, 21 August 1997. Can we go to page 4. 5 That's paragraph 2.1.4.6. It says there in the third 6 sentence: 7 "POCL also had problems with testing especially 8 Electronic Point of Sale System ... Pathway reported 9 that their testing strategy was under review and agreed 10 to pay particular attention to EPOSS." 11 A. Good. I notice it is Mr Coombs direction and Mr Coombs 12 was someone that Pathway brought in to strengthen their 13 technical team. He was the ICL technical director at 14 the time and I think he was parachuted in to Pathway. 15 Q. Can we go over the page to paragraph 2.3. There you 16 highlight that more emphasis was being placed on live 17 trial and quality rather than speed. So that's again 18 the very point that you made in the earlier document. 19 Again, there seems to be a looming pressure coming from 20 somewhere for speed. Would you agree with that? 21 A. Not from us. I mean, you know, we had what we thought 22 were a set of committed timetables after a re-plan in 23 March '97 and they were slipping. What do you do about 24 that, is the question, you know, and so we had another 25 re-plan and we questioned severely from the documents, 39 1 some of which you've just seen, about whether Pathway's 2 proposals to catch up in terms of release software were 3 really credible. We wouldn't have agreed to something 4 that was intrinsically risky technically, as far as we 5 were concerned, before it was bottomed out. 6 DSS, as I said, had promised their business case, 7 as I understood it, belatedly, from documents I've seen 8 and from the NAO report to remind myself -- had 9 predicated to the Treasury something like 15 million 10 a month on fraud savings, as the basis of their business 11 case, in order to -- once it had been rolled out and 12 once it was working. So, clearly, they had -- I mean, 13 the delays were hurting every party financially. 14 There was no winner here. ICL weren't getting 15 paid because the PFI structures. We were having our 16 competitive position damaged. We were disappointing 17 subpostmasters and our staff whom we said "This is 18 coming". 19 But so there wasn't, you know, severe pressure to 20 get this done but we had to, as I say, hold our 21 collective nerve to ensure that what we put out was good 22 enough. 23 Q. But there's document after document saying "We need to 24 emphasise quality rather than speed", and, I just 25 wondered, somebody must have been asking for speed 40 1 rather than quality? 2 A. As I said, I'm trying, to be honest, really honest and 3 candid and to the best of my recollection. There was 4 undoubtedly pressure on Benefits Agency from DSS and 5 I would imagine the Treasury. I don't know I'm 6 speculating on that. They were pretty close, DSS and 7 Treasury. They were a big-spending department so they 8 would be. 9 You know, they would have been saying, "What's 10 going on here? Why can't we do ACT instead?" 11 Q. You mention the PA Consulting report and that was in 12 October 1997. I'm going to spend a little bit of time 13 on that. 14 A. I thought it was a bit earlier than that, I'm sorry. 15 Q. The report itself was October 1997 and that's actually 16 one of the questions that I'm going to begin with which 17 is did you input into the report before it was 18 published? 19 A. I did, because Peter came round and saw most parties 20 from within the -- because he was commissioned by the 21 PDA from Alec Wylie, so all parties involved. I think 22 you reminded me, actually, from the document I saw last 23 night -- thank you -- about my initial reaction to it 24 was that one of my queries was I don't think he talked 25 to enough people within Post Office Counters to get 41 1 a fully rounded view on people like who were involved in 2 the automation steering or the finance people, although 3 we asked him to do it. But you'd ask Peter about that 4 yourself. 5 Q. You said at paragraph 63 of your witness statement that 6 PA Consulting report of October 1997 found no 7 fundamental technical issues with the system. That 8 sounds a little bit like the comment we heard yesterday, 9 "not fundamentally flawed". Did you have lesser 10 concerns? 11 A. Well, it's a load of lessons that were emerging and 12 I tried to dispose in that July report for all parties. 13 You know, I could only take -- I could observe what was 14 happening with other people and give my opinion on it. 15 I could only action what I thought was necessary within 16 Post Office or recommend action. 17 Q. Can we look at POL00090015. 18 A. But you are right that we did rely on quite -- that it 19 was -- it could work. It wasn't technically 20 intrinsically flawed. 21 Q. Absolutely. This is precisely a document that I'm going 22 to take you to. Again this is one of those new 23 documents -- 24 A. This is the one I was referring to. 25 Q. Yes. 42 1 A. There you go. 2 Q. So if we look at that document, it's the last page of 3 that document and it's a letter from you to Peter 4 Copping. 5 A. Yes. 6 Q. This is -- it seems like an important point. So you're 7 saying there -- 8 A. What date is this please, Mr Blake? 9 Q. It's 8 September 1997 so shortly before publication. 10 Publication was in October. 11 A. I see. Thank you. 12 Q. "Two points do occur: first, on the basic technical 13 question of whether there's a basic ICL Pathway design 14 flaw or not ... combined with the sheer scale of what 15 we're trying to do, makes the programme inherently 16 unstable. The report is silent on this explicitly at 17 the moment, though it implies the design is feasible. 18 I wonder if people who worked on the initial technical 19 evaluation (including, if I recall correctly, a PA 20 consultant) could help [me] here?" 21 So this seems to be you drilling down on precisely 22 the issue that we were addressing yesterday, that 23 because you were focusing on "fundamentally flawed" -- 24 A. We wanted to know because, as I said, we had 25 a culture -- I don't know what it's like today in Post 43 1 Office Limited -- but we had a culture of "no denial". 2 We wanted to know. If you didn't know, what could you 3 do about it? 4 Q. "Instability" is perhaps -- would you agree it would be 5 a better test than "fundamentally flawed", if you are 6 agreeing a system? 7 A. I mean, I think my language was a bit, you know, 8 imprecise there. "Unstable" could mean a number of 9 things, couldn't it, and I can't remember exactly what 10 I meant there. I didn't know if it meant "undoable" or 11 "it might fall over"; it could mean either of those 12 things. 13 Q. I think "fundamentally flawed" you would understand to 14 be doesn't really work, whereas -- 15 A. Yes, "fundamentally flawed" -- it's the "inherently" 16 bit. You know, "inherently" implies that -- or to me at 17 any rate -- that it will be difficult to fix and it 18 might be something structural and systematic. If there 19 are technical issues and flaws that can be addressed and 20 tested, that's a different matter. 21 Q. Would it be fair to say that you're asking there, is it 22 unreliable even if it works? 23 A. I don't think I had that in mind when I wrote it, is the 24 truth. 25 Q. Were those questions that were being asked at the time? 44 1 A. I'm trying to remember honestly. We certainly will 2 have -- "unreliable" will have been really manifested 3 during a live trial if it was unreliable or not and we 4 hadn't gone into live trial other than a few offices at 5 that stage. So it would have been difficult to know. 6 CF my earlier answer about; some testing you can't do 7 until you actually -- you can do all the testing -- It's 8 like playing a football match; you can do all the 9 training but you have to actually be in a match to see 10 if it works. 11 Q. So it's the acceptance and the testing live trial -- 12 A. Live trial had to come before acceptance. That was a -- 13 very fundamental stuff and the issues that came out of 14 live trial -- or around live trial in the other forms of 15 testing -- would have been logged and then addressed to 16 Pathway and the seriousness of them, either something 17 that make it unreliable or the instance of them, you 18 know, the frequency of them, would have been -- would 19 have determined their seriousness and their 20 prioritisation to get fixed before moving on, to my 21 mind. 22 Q. Do you think you got an answer to that point on whether 23 it was inherently unstable? 24 A. I think the October one, report that came out, said it 25 wasn't fundamentally flawed and I took that as including 45 1 in that. 2 Q. So your understanding of "fundamentally flawed" would 3 include whether something -- 4 A. Whether it was reliable to work, you know, because if it 5 was fundamentally flawed it wouldn't. 6 Q. Let's look at the report itself. That is at 7 POL00028092. Can we look at page 7, please. So 8 although, as you said, the finding is that the system 9 isn't fundamentally flawed there were at least some 10 concerns both in relation to POCL and in relation to 11 Pathway at that stage that were raised by PA Consulting; 12 would you agree with that? 13 A. Yes, we did. All three organisations got some lessons 14 to learn. 15 Q. If we look at those three bullet points at the top -- 16 A. Yes. 17 Q. -- and actually can we look at the top half of the page 18 including the paragraph that begins, "our key concern". 19 So there are some concerns raised in those bullet points 20 and then it goes on to say: 21 "Our key concern is that the skills required for 22 many of the new senior posts are, in our opinion, not 23 those we would have expected to find as part of POCL 24 core competencies ..." 25 So there are concerns there, it seems, about 46 1 a lack of skills within the Post Office; is that right? 2 A. Lack of those skills particularly true as it says to 3 implementation, management and contract and service. 4 I wouldn't have agreed with him about contract 5 management. I think contract management, he's talking 6 about a technical contract and how you manage that. 7 Service management, I would have agreed because, I said 8 earlier on, we were learning to be an automated 9 organisation and an automated organisation needs 10 a central service management function typically. We 11 knew that: we were trying to address it. It was 12 complicated by the fact that the PFI contract was let 13 out, so some of that service management, as Mr Cipione 14 points out, was contracted to Pathway to provide, in 15 terms of helpdesk, support desk, system desk, incident 16 log-in, all that stuff, but you still needed someone to 17 be able to manage that provider. 18 Q. The third of those bullet points, they concern EPOSS: 19 "In the longer term there's the issue of 20 developing the POCL requirements for EPOSS and the 21 supporting computer applications. We understand there 22 is much still to do here, which will require additional 23 resource." 24 EPOSS -- we heard about earlier this morning 25 didn't we -- at an early stage there hadn't been enough 47 1 thinking about EPOSS and again in October 1997 there 2 were concerns about there being much still to do in 3 relation to EPOSS? 4 A. I agree with that. There was still much to do and it's 5 not a defence or an excuse, it's not what I'm here for 6 but the prioritisation on the benefits service, which 7 could have put a whole system in jeopardy, or the whole 8 project in jeopardy, we seemed to think we should get 9 that right before moving on. We understood it needed to 10 be done and we did put extra resource on. 11 Q. Can we look at page 8 and it's M3.4 on that page. The 12 report then goes on to talk about concerns with Pathway 13 and it's over the page that I would like to look at and 14 it's the first paragraph: 15 "Much of our review at Pathway has focused on the 16 robustness of the technical solution since this has 17 become a significant concern for the sponsors and 18 Pathway. Whilst we have been able to obtain 19 satisfactory answers to all our questions, in particular 20 regarding the way security requirements will be 21 incorporated and on scalability and performance, there 22 must continue to be reservations in all these areas 23 until the final design is baselined and then realised. 24 It is also important to note the dependency for 25 technical success on Escher which is a small, 48 1 Boston-based software house and the source of the 2 Riposte messaging software which is at the heart of the 3 system." 4 Now, again that's something we spoke about earlier 5 that was -- 6 A. And he's repeating what we knew that Escher -- they were 7 dependent on Escher and what were they doing about it to 8 put it right. I've been trying to remember that 9 actually. I honestly can't recall what extra resources 10 or what actions Fujitsu took or Pathway took to -- 11 I can't remember if they brought Escher in or bought 12 them out or got them in but they certainly put extra 13 technical resource to manage them. 14 Q. Can we look at your statement which describes replans 15 that took place around this time. It's WITN04030100. 16 A. Paragraph, please? 17 Q. Paragraph 68, page 24. 18 A. Thank you. 19 Q. I think this is talking about this particular time: 20 "This was my understanding of the technical 21 difficulties with Horizon at the time but I would also 22 add that those Post Offices to whom the service had been 23 rolled out largely reported that it was going ok. They 24 did not raise many technical difficulties and indeed 25 many post offices who did not yet have access to the 49 1 system were saying that they wanted it." 2 Now, a reading of that paragraph might be that it 3 contains quite a few caveats in it. At that stage it 4 would own have been a small number of post offices who 5 had Horizon installed I think less than 200. 6 A. Yes, about 205-215. 7 Q. There was also limited functionality at that stage. 8 A. Limited functionality, yes. Both of those are true. 9 I tried to express that. Forgive me if I didn't. 10 Q. If some subpostmasters were having some technical 11 difficulties at that stage, when it was still a small 12 project, would that have been a concern? 13 A. Yes, but my memory of that is that the nature of the 14 technical difficulties weren't the ones that were being 15 experienced, apparently, during roll-out, around the 16 lack of balancing and that sort of stuff. 17 Q. So what was your understanding of -- 18 A. I think it was things like screens freezing 19 occasionally. I might be wrong, there might be one or 20 two. I can't be certain but I do know that I can 21 remember Dave -- you will have, to ask Dave Miller -- 22 Dave Miller having a meeting with many of the 23 subpostmasters who were in the trials or going round 24 there and asking them the direct question -- I think 25 it's in one of the Federation documents -- saying "So 50 1 given all this, would you want us to roll it out 2 further?" With a resounding yes. 3 So that isn't to say we were complacent. 4 I wouldn't want to give that impression. We weren't. 5 But the whole purpose of doing the live trial was, as 6 I said earlier, was to try to get those experiences from 7 those who were actually using the system and get some 8 feedback into that. 9 Q. Could I ask for the witness statement to be brought back 10 on screen on the same page if possible. Thank you very 11 much. Can we look at paragraph 69. There you say: 12 "... the natural forum to raise these difficulties 13 would have been at the meetings between the NFSP and the 14 network director." 15 Who was the network director? Was that Jonathan 16 Evans? 17 A. Yes, it was. 18 Q. So you would have expected problems to have been raised 19 by the NFSP at that stage; is that right? 20 A. The NFSP, as John said yesterday, certainly weren't in 21 the Post Office's pocket. They were vociferous if they 22 thought there was an issue and if there had been 23 substantive -- well I imagine, I'm speculating -- 24 I imagine if there had been substantive complaints by 25 those subpostmasters who had used the system because of 51 1 these sort of issues. I'm sure they would have been 2 alerted to it and they would have been not slow in 3 coming forward with those. 4 Q. But the number of post offices at that stage was very 5 small. 6 A. As I was saying, yes. 7 Q. What did you see as the role of the NFSP in that regard 8 then? 9 A. Clearly they -- they were with us bringing together 10 subpostmasters to get their reaction. I think the 11 meeting I alluded to before, which I read about when 12 Dave Miller went, I think there was an NFSP executive 13 member there for that. The NFSP have been involved 14 early on in the genesis of the Horizon project against 15 threat of ACT and therefore the threat to the national 16 network. They clearly would have been representing 17 their members there, fiercely, to ministers. 18 Q. Do you think the NFSP were given much of a role at that 19 stage? 20 A. Given? 21 Q. Well, did they have much of a role, much of a say, in 22 what was going on? 23 A. They were kept informed. They went to meetings with 24 Pathway, I think, during the bidding process. I think 25 all three bidders set out stalls at a National 52 1 Federation of SubPostmasters conference to show these 2 sort of things. We hadn't gone and we didn't think 3 about that until I was in the process of setting up Post 4 Office Counters Limited we didn't think about putting 5 them on the board, for example, or giving them 6 a non-exec role. I think that would have been a step 7 too far for us at that point. 8 Q. Can we look at POL00028137, please. Now, this is a very 9 early meeting, so some years before the period that I'm 10 talking about? 11 A. Yes, this is in the very genesis. This is just 12 before -- just to give context to this, the MOU you 13 showed me yesterday which was '95 there were earlier 14 non-legally -- non-legal versions of that to which this 15 was a contributor. 16 Q. Can we look at the penultimate paragraph of that and the 17 final sentence of that penultimate paragraph: 18 "NFSP involvement would be kept to a minimum." 19 Do you think that there was an attempt to keep 20 NFSP involvement to a minimum? 21 A. No. I don't. I'm surprised about that. Did I write 22 those notes? 23 Q. That's a minute of the -- 24 A. Yeah, I wonder who wrote them. For example, Andrew -- 25 no, that's not true. I was going to say he was the guy 53 1 who dealt with the Fed but probably wasn't at that 2 point. No, I don't -- we did engage with the Federation 3 around those issues but there were boundaries. You 4 know, there had to be boundaries. They were a trade 5 organisation representing people and we took them 6 seriously because they were our people too. 7 Q. Do you think some people took a view that they shouldn't 8 be getting involved in the technical side of things? 9 A. Well, I don't think they had technical capability but 10 certainly we were keen to include subpostmasters and 11 tell the Federation about the user testing. 12 Q. I'm going to move on to November 1997 and that's a month 13 after the PA Consulting report. Can we look at 14 POL00028599. This is an interim business continuity 15 status report for the period 20 to 26 November 1997 and 16 you were a recipient of this report. 17 A. Was this one of the new ones, Mr Blake, or one of the 18 old ones? 19 Q. I believe it's an old one. 20 A. Okay, fair enough. 21 Q. I'm only going to take you to a paragraph of it. Can we 22 look at page 2, paragraph 1, halfway down that paragraph 23 it says: 24 "The primary purpose is to identify any issues 25 (actual or potential) that might give rise to 54 1 considering a suspension of the Congo 4 roll-out or 2 regression from ICL Pathway services." 3 Do you remember the Congo 4 roll-out? 4 A. I will probably get this wrong. My memory is failing me 5 now from however long ago. Congo -- I can't -- 6 I wouldn't be able to give you the detail of what 7 Congo 4 roll-out. There was Congo 4, Congo 5, 8 Congo 4 plus, we knew it also as 1C which I think it 9 said before but I think these were releases relating to 10 other functionality other than the benefit encashment 11 service. 12 Q. At the bottom of that page it makes clear that there are 13 no matters that need escalating for consideration of 14 a suspending of the Congo 4 roll-out but it's 15 paragraph 3.2.1 that I'd like you to look at and that's 16 on page 3. Can we possibly blow up that paragraph. 17 Thank you very much. 18 "After assurances following a previous duplicate 19 payment ... a new duplicate payment situation has 20 occurred at Bath Road SPSO. This was due to the Post 21 Office not being able to 'poll' for 8 or 9 days, 22 therefore the system was unable to identify that the 23 original payment had already been made." 24 Now, we heard from Mr Roberts yesterday that he 25 wasn't really involved -- or his evidence was that he 55 1 wasn't significantly involved in this kind of level of 2 detail, individual branch level of detail -- but this is 3 something that would have been brought to your attention 4 at the time, that level of detail? 5 A. What date was this, please? 6 Q. It is November 1997. 7 A. '97, okay. Yes, it would have been then, yes. 8 Q. What do you understand by that paragraph? 9 A. I understand that there seemed to be an issue, as it 10 says, not being able to "poll" that ICL would need to 11 fix. 12 Q. Can we go over the page please to 3.4.1. Can we look 13 at 3.4.1 -- thank you very much: 14 "One of the new PMSR reports introduced at the 15 beginning of release 1c, does not appear to be working 16 correctly. It did not pick up the Bath Road duplicate 17 payment and report it as an unmatched encashment. 18 Pathway have stated that they will seek an urgent fix to 19 this." 20 Can we look at the paragraph below the two 21 paragraphs below: 22 "Another concern is that one of the transactions 23 involved in this incident didn't come through to ABED 24 and wasn't reported on the CBoS report. Pathway aim to 25 address this as part of the fix being applied to the 56 1 above. 2 "This has a double impact on transaction 3 processing: 4 "[first] it creates an error against cash account. 5 "[Second] settlement with BA is based on an 6 incorrect sum -- Pathway are looking to manually amend 7 the CBoS report as an interim measure. This has an 8 impact on POCL accounting." 9 Pausing there, do you understand that second 10 bullet? 11 A. I don't know what a CBoS report is, honestly I don't. 12 Q. Were you aware that Pathway was able to manually amend 13 certain things in branch -- in the branch accounts? 14 A. Well, from this -- not that I remember -- but from this, 15 it's in black and white, so I must have read it at the 16 time. But I don't remember that and I don't -- and it 17 would have been with Pathway and the PDA to fix. 18 Q. It says, "This has an impact on POCL accounting"; do you 19 know what that meant there? 20 A. Well, presumably the accounting would have been -- had 21 to have been adjusted to correct the error. 22 Q. So if there's a manual amendment it would have an impact 23 on Post Office accounting? 24 A. Yes. I don't know if -- it says, "Pathway are looking 25 to amend". I don't know if they did, so I can't comment 57 1 I'm afraid. 2 Q. The paragraph below: 3 "The above has been registered with Operational 4 Service Management, but there is a growing concern given 5 the limited functionality and few on-line offices." 6 A. Yes. 7 Q. So in that period were you concerned about errors during 8 the "limited functionality and few on-line office" 9 period? 10 A. I don't remember it as -- as I said earlier -- as 11 a large issue or a big material issue that was being 12 brought to our attention all the time. Clearly this is 13 an example. It's an operational service management 14 report which goes to me and others as part of the PDA 15 board and when the PDA board met we would have asked 16 what had been done about it and decided. I don't think 17 it's like -- it's not sent to me to action, as it were, 18 myself. 19 Q. Did somebody in particular action that? 20 A. Can you show me who the -- 21 Q. Can we look at the first page -- 22 A. Because I don't remember this report. Yes, it would 23 have been -- yes it would have been -- the person who 24 would have been tasked with fixing it with ICL would 25 have been Peter Crahan. He's the guy in charge of the 58 1 PDA at this point bearing in mind the PDA existed and 2 Horizon programme or -- it hadn't been moved yet to 3 Horizon or it was in the process of doing, so I think 4 (unclear). 5 Q. I'm going to move on to 1998, the spring to autumn 6 of 1998. Can you briefly tell us how your role changed 7 in that period? 8 A. In the spring/autumn '98? 9 Q. Yes. I think after 1997 did you leave the PDA board? 10 A. No. I think what happened was the PDA board -- the PDA 11 started to dissolve and -- that was part of the 12 recommendations, as John said yesterday -- we brought 13 more things back in-house. There was still a residual 14 PDA dealing with contractual matters and that sort of 15 things, where there was a joint contract, but we brought 16 a lot of it back into a new Horizon programme 17 director -- which was Dave Miller. So my role changed, 18 really, to more like a commercial strategic, rather than 19 delivery. Of course I needed to be kept appraised of 20 where we were because clearly things are iterative to 21 some extent. 22 Q. So at that stage was David Miller more focused on the 23 technical matters and you were -- 24 A. On the delivery. 25 Q. -- on the delivery. 59 1 A. -- on the delivery, including the technical matters. He 2 would have been resourced up. We resourced him up. He 3 would have reported directly on those matters to the 4 managing director. 5 Q. And your role at that stage, you saw as more strategic? 6 A. More strategic commercial I suppose if there had been 7 major contract re-negotiations that came out but 8 everything, then of course, was in hiatus during, as 9 John -- I hadn't heard him use that expression before, 10 "the year lost" -- in '98 when the Benefits Agency and 11 DSS basically called time and the Treasury working group 12 was set up et cetera, et cetera. 13 So Dave Miller was trying to keep the programme 14 going, with his Benefits Agency, under a guy called 15 Vince Gaskell, who was on there, was doing that for the 16 Benefits Agency, if you like, and we were more involved, 17 really, in trying to cope with the various reports, 18 consultancies and people sent to us by the Treasury 19 working group and give what we thought was good advice. 20 Q. Can we look at POL00038828, please and this is moving to 21 March 1998, March and April. 22 A. Ah yes. This was the first Bird & Bird report. 23 Q. Do you know whose handwriting that is on the front page? 24 It doesn't matter if you don't recognise it. 25 A. I would have thought it's Dave Miller's but it says 60 1 "Dave" so it's confused me. I don't therefore. 2 Q. Can we look at page 5 in paragraph 310. Again is this 3 the same -- I don't if this is same handwriting or if 4 this is handwriting that you recognise at all? 5 A. No, you gave me a document yesterday which had 6 annotations on it which looked similar to this. 7 Q. You don't need to be a handwriting expert. If it's not 8 yours, then that's sufficient. 9 A. -- I don't know -- Dave would have had a number, you 10 know, people working for him and I imagine it's one of 11 those. 12 Q. So there are some concerns set out there; I'm going to 13 very briefly summarise them. There were some concerns 14 about paperwork things or agreements to agree, 15 contracting authority responsibilities et cetera. The 16 second bullet point security controls, security 17 requirements. The third paragraph, training 18 requirements and solutions. That seems to be crossed 19 out, I'm not sure if you are able to assist as to why 20 that would be crossed out? 21 A. No. 22 Q. The fourth, problems with the Post Office estate, 23 availability of ISDN connections. 24 A. I've mentioned some of those before haven't it? 25 Q. Yes. Then it's the next paragraph that I'd like to look 61 1 at: 2 "The Pathway systems interface to BA and POCL 3 systems. Some of these are new systems, being developed 4 in parallel with Pathway, others are being modified to 5 include the Interfaces. There have been a number of 6 issues with the interface systems, particularly with the 7 BA CAPS programme and the POCL reference data system." 8 Does it say -- if may say "infer culpability" or 9 something on the right-hand side, but again if it's not 10 your -- 11 A. Honestly, I couldn't tell you who that was. 12 Q. Were you aware of POCL reference data issues at that 13 stage? 14 A. I know we had to get our reference data more systematic 15 in order to be able to cope with an automated world. 16 I wasn't aware of, at that time -- I don't remember at 17 any rate -- reference data issues being a particular 18 interface problem at that time. I've read subsequent 19 reports that you have sent me, or the Inquiry sent me 20 rather, sorry, that talks about reference data issues -- 21 Q. Would you have received these Bird & Bird documents? 22 A. I think I would have received this Bird & Bird document 23 at that time. I think I referred to it in my witness 24 statement, so I might have done. It's confusing because 25 this is Project Mentors and they did a separate one, out 62 1 of the blue really, at the end of '99 as well. 2 Q. So who were Project Mentors, very briefly? 3 A. I think they were a consultancy, an IT consultancy or 4 claim they were. I think they were run by 5 a professor -- 6 Q. We may see that -- 7 A. -- that obviously Bird & Bird, who were the joint 8 contract solicitors for the PDA, knew. 9 Q. Can we look at POL00069096. That's a meeting of the 10 Counter Automation Steering Group on 27 March 1998. 11 I think you'll need to go over the page but it's clear 12 that was sent to you, the first page that we skipped 13 over, and you would have received that because you're 14 named as being present at the meeting as well? 15 A. This is the minutes of this meeting I see, yes. Thank 16 you. 17 Q. Can we look at page 3 and the top two paragraphs there. 18 It says, about halfway down the first one: 19 "POCL would not seek to delay Pathway's April 1999 20 roll-out date, but before accepting the system would 21 want to be certain ... it was working correctly; work on 22 EPOSS was continuing and Pathway had indicated that 23 while it could provide a system which met the contract, 24 its lack of robustness could generate high level of 25 errors within POCL. This was being investigated 63 1 although it was difficult to quantify how the system 2 would work until after it had been installed and was 3 operational." 4 That all sounds quite serious at that stage, 5 doesn't it, especially in relation to EPOSS? 6 A. Yes, I mean, that doesn't say, though, that we would 7 have wanted to have let this system roll out, not in my 8 view anyhow, without the Acceptance Incidence, including 9 on EPOSS, being cleared. 10 Q. So it's similar to the evidence that you gave earlier 11 that, although you knew there were problems it would be 12 in the roll-out that that kind of thing might -- 13 A. Well, the roll-out -- before going into roll out, Dave 14 Miller I think had a system of acceptance instances 15 which he categorised high, medium and low and there were 16 a certain number that could be allowed but none that 17 were high, and I would have imagined the EPOSS one was 18 high. But that acceptance happened after my time, so 19 I can't really comment on that. 20 Q. The reference to high level of errors there sounds 21 concerning. 22 A. Well, it says "could generate high levels of errors". 23 It sounds like the way -- this is a report by Dave, by 24 the sound of it, and it said Pathway itself had -- might 25 have thought that the way it was doing it could generate 64 1 high levels of error. So that is a cause for concern. 2 Q. If we look back at the first page -- 3 A. I mean, clearly, that wasn't acceptable. 4 Q. -- sorry, the second page. The attendees of that 5 meeting: John Roberts was Chairman of that committee. 6 A. He was. 7 Q. Would you have expected him to take that kind of 8 information to the board level? 9 A. You need to ask John that. 10 Q. Was your expectation at that time that those kinds of 11 details would have been raised at board level? 12 A. I would have respected John's judgement about what he 13 took or didn't take to the board. This was, as he said 14 yesterday, an extra piece of governance on the board 15 that he included, he and his -- some of his colleagues, 16 the executive colleagues, Richard Close is the finance 17 director, Jerry Cope is the group strategy director, 18 Stuart Sweetman, the managing director for counters and 19 me and Dave from within Counters -- in order to more 20 closely monitor the project. 21 Q. I think you have said that acceptance criteria would be 22 an important factor in dealing with those kinds of 23 problems. 24 A. Yes. 25 Q. Who would you see as responsible for that? 65 1 A. Well, the acceptance criteria, from what I've read 2 recently, ICL was proposing different acceptance 3 criteria towards the end of '98/'99, as I understand it, 4 and that was then -- that would have fallen to Dave 5 Miller primarily to deal with at that point, not me, so 6 I can't comment on those and didn't comment on those and 7 those acceptance criteria, which would have then folded 8 in, in the second half of '99, after the ministerial 9 decision would have led, I imagine, to the acceptance 10 criteria being modified, as part of the re-negotiation 11 of the heads of terms between ICL and Post Office. 12 I don't know who had signatory authority within the Post 13 Office for that, I'm afraid. 14 Q. You don't know who had signatory authority but who would 15 you have expected to have taken responsibility -- 16 A. For the acceptance criteria? 17 Q. -- for the acceptance criteria? 18 A. I would have imagined Dave with taking legal advice and 19 surrounding himself with people from -- who had 20 experience of the programme and taking IT advice and 21 others. I'm sure he would have done this, I trust Dave 22 would have. 23 Q. That's David Miller? 24 A. David Miller, yes. 25 Q. I'm going to move on to the Treasury working group. 66 1 Now, you weren't part of that group. I think it was 2 Jonathan Evans. 3 A. Jonathan Evans, Jonathan represented Counters although I 4 gave inputs to it. 5 Q. Do you think that the Post Office provided enough 6 technical expertise to those kinds of groups? 7 A. Well, there was a panel of technical experts working for 8 the Treasury working group, alongside KPMG, who were 9 doing the overall evaluation, and they came and talked 10 to various people within the business. So it was really 11 responding to them, rather than being -- us saying 12 "Here's our technical people, could you do it?" They 13 were asking us a lot of questions about the impact of 14 cancellation or termination of the contract, in part or 15 in full, and, as you heard from John yesterday, the 16 variations to that theme got wilder and wilder and more 17 and more radical at times. 18 Q. We heard earlier about the PA report which said that 19 there was at least some lack of expertise within the 20 Post Office when it came to those technical matters. Do 21 you agree with that? 22 A. I think we needed to improve our core competence in 23 that. I would agree with that. You could never say you 24 have enough capability, I don't think. I think that 25 would be arrogant to say so. I think we had sufficient 67 1 expertise to manage the programme, given the PFI nature 2 of it. 3 Q. Do you think Jonathan Evans going into that Treasury 4 working group had enough understanding of the technical 5 details? 6 A. He would have come back and asked us about that or had 7 provided briefings if he needed to but he was there 8 primarily around -- well, he's a good guy but also he 9 was the network director and a lot of the modelling 10 being done was on the impact on the network. 11 Q. Do you think that those with technical expertise were 12 given enough say on whether the Post Office should 13 remain committed to that project at that time? 14 A. Ah, I see. I don't know. John answered yesterday, 15 I think, that we talked about the options and Jonathan 16 did a working group about re-tendering, if necessary, 17 and that sounds easy, but isn't. You know, given the 18 situation, you had to postulate different scenarios and, 19 indeed, I think there's a report in the bundle where we 20 asked -- as well as asking for an external view, we 21 asked for a peer review run by our own finance director, 22 who was pretty fiercely independent, to say if Benefits 23 Agency withdrew from this what was the best option. His 24 conclusion was the least worst option was to carry on 25 with the project. 68 1 Q. Let's look at a letter or a note from Mena Rego. That's 2 POL00028649. It's a document of 8 July 1998. Can you 3 just tell us who was Mena Rego? 4 A. Mena was one of my direct reports and she was working 5 on -- she was the Horizon development manager, I think 6 her title was. She had a commercial background and 7 a general management -- 8 Q. The penultimate paragraph there -- 9 A. Oh, this is deep pink, yes. 10 Q. "... we have to get the message across very clearly to 11 officials/Ministers that we are not a pawn in the game 12 between DSS desire to exit and Treasury/DTI desire to 13 prop up ICL and that if BA cease the payment Card we 14 reserve our right to make our own decision on 15 continuation/extension of the contract or termination 16 and this would depend on our satisfaction on the 17 2 points above." 18 A. Can you remind what the two points above were? 19 Q. Yes, absolutely. Can we just have a quick look. 20 A. Yes, I see. 21 Q. What did you understand by "pawn in the game"? 22 A. You haven't met Mena but she's a very direct person and 23 it's probably not the language I would use. But we 24 can't -- what she was trying to say there is that the 25 Post Office and POCL couldn't be the fall guy for DSS's 69 1 desire to exit and the Treasury/DTI desire to do so but 2 put any of the losses that might result on that on to 3 Post Office rather than ICL. 4 I mean, I think the background to this, which 5 I must just mention, if I may, this was 8 July '98, so 6 this is going into the Treasury working group and you 7 will know that the KPMG report, eventually published, 8 showed that the clear best value for money was to 9 continue with the benefit payment card but on 10 an extended roll-out. The only loser in that scenario 11 was Benefits Agency and, therefore, they resisted it 12 tooth and nail. 13 Q. Did the Post Office come up with a credible alternative 14 strategy to the payment card? 15 A. We talked and considered and agreed that we would 16 migrate to a smartcard, which you will recall was in our 17 minds at the outset and, indeed, even appears in the 18 1995 MOU as the system would be able to migrate to it. 19 But it depends what you do with that smartcard and it 20 depends on the rate of business you lose at the same 21 time. You have to think about our post offices and our 22 customers here. 23 If, as John said yesterday, it wasn't just 24 a technical decision it was a rounded decision, at 25 '98/'99 less than 25 per cent of DSS claimants chose to 70 1 have their payments made by ACT so, even when it was 2 available to them, they were still choosing to come to 3 post offices. We were determined that that customer 4 choice should remain in one way or another. 5 So the credible alternative we tried to put up, 6 under one of the option 2 variants, when DSS had dug 7 their heels in with Treasury backing, to some extent, 8 saying they wanted to terminate their part of the deal, 9 was that we were prepared to go to a smartcard that 10 would enable banking services in some way, shape or 11 form, provided that BA continued its managed transition 12 on ACT over a number of years and continue to pay Post 13 Office and, therefore, we could pay subpostmasters and 14 our people some amount for that card. 15 Q. Do you think that the Post Office weren't thinking of 16 a simpler plan, though? 17 A. Such as? 18 Q. There are some criticisms certainly in some internal 19 government reports -- I don't need to take you to them 20 because you won't have seen them at the time -- but they 21 were critical of the Post Office's lack of a robust 22 business plan at that stage? 23 A. I don't -- disagree with that. I mean, I take issue 24 with that. It's very easy for people who don't run 25 things to make up comment on those things. The enablers 71 1 that would have had to be in place then, as now, for 2 that network to survive was that a universal banking 3 product would have had to be developed. One suggestion 4 from within the Treasury was the Post Office could 5 become a bank to do that, which was unacceptable because 6 of all the regulatory risks. We sort of sold our bank 7 off ten years earlier with Girobank privatisation. 8 We would have had to make enabling arrangements 9 with banks to do that and the government would have had 10 to do that because the banks wouldn't have been rushing 11 to take on a lot of unbanked people who they wouldn't 12 have seen. 13 We had a plan also to provide what we called 14 network banking services, which we then went on to 15 develop for banks as they closed branches to come and do 16 stuff. 17 That would have also required an automated 18 network. It would have required further releases to 19 enable things like debit cards to be done, probably 20 an ATM-type network. So it wasn't a quick fix, is what 21 I would say. 22 Q. Can we look at POL00028644, and that's a letter to you 23 from David Sibbick who was the Director of Posts at the 24 DTI at the time. 25 A. Yes. 72 1 Q. That's August 1998. 2 A. Yes. 3 Q. At the bottom of that first page, he's asking if 4 Benefits Payment Card were dropped, what technology 5 would the Post Office want and could simpler technology 6 be used? It seems there that the Post Office are being 7 given an off-ramp to the Horizon project if they wanted 8 it. 9 A. He's asking for a scenario and we gave him that and 10 I think it's in the bundle. 11 Q. Did the Post Office ever come up with a simpler system 12 as an alternative? 13 A. I'd like to know what he meant by that. Such as? You 14 know, what is -- the simpler system -- we had a simple 15 system that was called payment method -- paper-based 16 payment methods, and that clearly was unacceptable and 17 not modern. The type of technology to enable us to be 18 competitive in the future would have had to -- should 19 build on the infrastructure that was being developed and 20 being rolled out at that point and included the option 21 then to upgrade it to more of a banking-type solution. 22 Q. I think a witness has said that, effectively, a tank was 23 built when all you needed was a car, or something along 24 those lines. 25 A. Did they? 73 1 Q. What would your position on that be? 2 A. Well, if we'd have started again without the benefits 3 card at that time and re-tendered we would have 4 obviously had a different type of specification, at that 5 point, that would have taken advantage of the latest 6 technology, without all the work that had been done on 7 very infrastructure to start with. It wouldn't have 8 taken away some of the earlier ones. 9 It was also a quite serious point about timing, 10 which the Treasury tried to ignore, which was -- there 11 were serious procurement law issues. 12 Q. I'm going to take you to a document very briefly, it's 13 POL00038842. It's again about a meeting that took place 14 with David Sibbick and it's a question that I'm asked to 15 ask you and I will just ask it very quickly. It relates 16 to paragraph 2. It seems as though, from paragraph 2, 17 there's a proposal from the DTI for a joint spokesperson 18 and it says "This was rejected by Paul Rich". Do you 19 remember that and, if so -- 20 A. I do remember it. I remember the background to it. 21 Q. Are you briefly able to explain? 22 A. Hamish Sandison was the Bird & Bird lawyer who had acted 23 for the BA and POCL in the PDA and he had -- for 24 example, we think he had commissioned that last Project 25 Mentors report in '99, which remarkably and 74 1 coincidentally came at the same time as DSS were 2 considering whether they withdraw or not, and we thought 3 he was in conflict. We had advice from -- because we 4 didn't agree on the way forward and we had advice from 5 Slaughters to Treasury solicitors and to our own 6 solicitor that -- I can remember the partner at 7 Slaughters giving a very clear indication that Hamish 8 would have been in conflict. 9 Q. So was your concern with Bird & Bird and the Project 10 Mentors you just described -- were you concerned that 11 they were adopting a DSS line? 12 A. The second -- I'll be honest about this and if 13 I misremember you will have to forgive me, and you can 14 ask Sarah Brown, but I don't remember commissioning -- 15 jointly commissioning the second Bird & Bird report 16 which, when I saw it, was, I think, on New Year's Eve 17 was sent to George McCorkell with me as a copy, which 18 I then passed on to Dave Miller, because clearly it was 19 technical issue, where they claimed there was 20 insufficient requirements analysis of the Benefits 21 Agency solution, which was a fundamental point. 22 But I thought it was interesting that they never 23 raised that in their report a year before. 24 Q. We'll get to that document shortly. Just in relation to 25 this document, did you suggest an alternative spokesman? 75 1 A. I can't remember. 2 Q. Moving on to late 1998, we're in autumn but let's move 3 to November, it's POL00028421. 4 Mr Rich, I should ask are you okay to continue? 5 A. It depends for how long because, I'm afraid, I'm of that 6 age. 7 Q. I have about -- I imagine I'm going to finish at half 8 past or thereabouts? 9 A. I don't know if there will be follow-up questions 10 though. 11 Q. It's unlikely. There maybe a very short follow-up from 12 Mr Stein. 13 A. If we can say -- I'll be fine until about 20 to/quarter 14 to. 15 Q. Let's see how we do and if we need a short break, 16 perhaps we can have a short break? 17 SIR WYN WILLIAMS: Mr Rich, I want to repeat that. At any 18 moment when you feel the need for a break, you say so 19 and we'll have one. 20 A. That's very kind thank you. I'd rather not lose the 21 flow if I can help it. 22 That's an unfortunate phrase, sorry! 23 MR BLAKE: So the document in front of us is from David 24 Miller about Horizon testing -- 25 A. Yes, I remember that one, yes. 76 1 Q. -- and it's the first substantive paragraph that I want 2 to ask you about. He highlights there that: 3 "My present assessment is that there are some 4 significant problems with the way Horizon passes 5 information through to TIP. These relate to the 6 provision of balanced outlet cash accounts and the 7 processing of the ensuing information via TIP. Whilst 8 we allocated some extra time during the Corbett review 9 to sort out any outstanding issues we need to be aware 10 of a potential threat to 14 December date." 11 A. Yes, I'm glad he wrote that and I'm glad he wrote that 12 there will be a potential threat to the 14 December 13 date, rather than saying we're going ahead with them. 14 Q. I'm going to move on because there's a theme developing 15 in November 1998. Let's look at POL00028320. This is 16 the "Transformation Steering Group Progress Report to 17 23 November 1998" and can we look at page 6. There's 18 what is called "Red Light Issues", which -- 19 A. This is another document that I only saw at 2.00 20 yesterday. 21 Q. Okay. If you need more time to consider it -- 22 A. This is the most substantive one, or one of the most 23 substantive ones, so my -- this is the annotations that 24 I said I didn't recognise. So I don't know who's 25 writing this. It also said, I think at the beginning of 77 1 it, that the meeting didn't take place. 2 Q. So is this a document that you think you would have 3 received at the time? 4 A. I can't tell. I would have received it if it was going 5 ahead because I would have been chairing the meeting. 6 Q. If we look at the first paragraph, and that's the only 7 paragraph I want to take you to, "Red Light Issues, 8 Horizon System": 9 "There are major concerns about the test results 10 emanating from Model Office and End to End. 11 "The results indicate that cash accounts and 12 transaction data delivered to POCL's downstream systems 13 lack accounting integrity, all of which raises serious 14 doubt about Pathway's ability to enter into the next 15 phase of Model Office and End to End testing without 16 some form of remedial action." 17 A. That's basically reflecting what Dave Miller said 18 before, isn't it? 19 Q. I was going to say, even if you didn't necessarily see 20 this at the time, were those issues that you would have 21 been aware of? 22 A. Yes, because Dave told us. What does whoever wrote it, 23 said about it, may I just ... 24 Q. If we could scroll down and highlight that? 25 A. "Remedial analysis has taken place [following] meeting 78 1 TIP and Reference Data personnel. Remedial action is 2 now underway. The point must be made that we will not 3 enter the final phase of testing until we are content 4 that we have a robust set of code." 5 Q. So somebody's written that at the bottom? 6 A. I think that's Dave Miller's number 2. 7 Q. Who was that? 8 A. Could have been one of two people. My guess is -- 9 there's a lady called Janet Topham at the time. 10 Q. That paragraph on the Horizon system, though, that 11 background 23 November 1998, I'm going to now look at 12 POL00038829 and I think this is the controversial 13 document, perhaps, that you were talking about from Bird 14 & Bird? 15 A. Yes. 16 Q. Can you assist us with that handwritten note or -- 17 A. That handwritten note is from Mena's secretary. 18 Q. "Dave" being? 19 A. Miller. 20 Q. Thank you. Can we look at the report itself, 21 paragraph 1 -- 22 Sorry, there are a few different versions of this 23 document. 24 MR BLAKE: Sir, perhaps we could take a five-minute break 25 now for everybody's convenience and we can go back on in 79 1 five minutes. Thank you very much. 2 A. Thank you. 3 (12.17 pm) 4 (A short break) 5 (12.22 pm) 6 MR BLAKE: Thank you very much, sir, we are back and I have 7 found the relevant page, page 3 of that document. 8 POL00038829. This is a letter from Bird & Bird, 9 December 1998, can we look at that first paragraph, 10 please. It's difficult to read: 11 "As you will see, [Andrew Davies'] team have 12 documented a further specific failure by ICL Pathway to 13 follow good industry practice in meeting the 14 Authorities' requirements." 15 Can we go to page 5. There's a letter there to 16 yourself and George McCorkell and Pat Kelsey from Bird & 17 Bird, and that first paragraph summarises the view of 18 Andrew Davies of Project Mentors. Can we just have 19 a look at that first paragraph, sorry. The quote there 20 is quoted from Andrew's letter: 21 "... 'deeply concerned that their findings show 22 a serious problem with the way in which ICL Pathway have 23 developed the system. The impact of this is likely to 24 be that there will be failures to meet essential user 25 requirements, causing the need for extensive rework 80 1 before the system can be accepted and, potentially, 2 operational problems if the system is rolled out." 3 That's quite a concerning statement, isn't it? 4 A. It's his opinion, yes. 5 Q. The impression that it gives is that there are real 6 concerns about Horizon at the time. 7 A. But if you look at -- no, not but. I think this is 8 commissioned on the way, if I'm correct, the benefit 9 encashment service was developed and the focus was on 10 that, and the claim then by Mr Davies was that the same 11 must apply to all other components. I think it also -- 12 I'm not a -- you know, I'm not a technical person, I'm 13 not an IT expert and wouldn't ever aspire to be but 14 I think it completely ignores the fact it was procured 15 under private finance. 16 Q. Can we go to page 6, which is the letter from Project 17 Mentors to Hamish Sandison and over the page to page 7, 18 and it's the second paragraph there. He says: 19 "Our experience of systems where requirements have 20 not been analysed satisfactorily is that the system 21 fails to meet the users' needs. An effective acceptance 22 test will identify many such failings necessitating 23 considerable rework. The result is a significant 24 extension of time and cost required to complete the 25 system and roll-out it out. The alternative is to allow 81 1 unacceptable processing in the operational environment, 2 with unpredictable and potentially damaging results." 3 A. Is there a question? 4 Q. I'm going to move on to show you the substantive report. 5 That's at page 8. Can we go to page 11. Can we look at 6 paragraph 1.3, "Scope", the second paragraph: 7 "We have to date only considered the [Benefits 8 Payment System]. Further work has recently started to 9 perform a similar assessment of the approach adopted for 10 other elements of the system, such as EPOSS. 11 Nevertheless our findings are, in our view, sufficiently 12 serious to bring into question the whole of Pathway's 13 design process." 14 Moving on to page 14, paragraph 2.3.4. Again, at 15 the bottom: 16 "Of particular concern is the EPOSS system [that's 17 the second paragraph]. We are informed that at 18 a relatively early stage Pathway wanted the Authorities, 19 principally POCL, to be involved with the design of this 20 element. The plan was to use the Rapid Application 21 Development ... methodology to design the system. This 22 approach was started, but discontinued after some 23 months, when the Pathway staff member involved left the 24 project. The suggestion to use RAD leads us to believe 25 that more traditional methods have not been used, and 82 1 since the RAD experiment was abandoned, we have doubts 2 whether any proper requirements analysis has been 3 performed." 4 On the same page. Can we go to page 14, it's 5 2.4.3. Sorry, that's the part we just read. Then over 6 the page again. You have addressed this in your witness 7 statement and you say that you were only copied in to 8 this document and that you would have passed it to 9 others. 10 A. Yes. 11 Q. You say: 12 "My recollection is that POCL did not necessarily 13 agree with the report as proving the system was 14 fundamentally, technically flawed." 15 So, again, that's the use of the term 16 "fundamentally flawed". Were these issues raised here 17 serious issues? 18 A. Clearly, they were serious -- seriously phrased. 19 I passed this on to Dave Miller at the time because he 20 was clearly in the process. As I said, this is December 21 '98 when he was considering whether or not to authorise 22 the further release. As this says, it doesn't actually 23 analyse the EPOSS system. It talks about things had 24 started to go on the EPOSS system, it talks about rapid 25 application development in the past, which ICL did want 83 1 to use -- RAD, it's called -- and we didn't particularly 2 want them to do that unless it had quality of outputs to 3 do so. 4 As I said before and I stick to that, the previous 5 Project Mentors report a year or so earlier on the whole 6 thing seemed not to address the issue of insufficient 7 requirements analysis at all, which I find surprising, 8 and I just note that it came out as the DSS were 9 considering praying in aid whether or not to continue 10 with the benefit card or not. 11 Q. Coming in a year later, presumably that's even more 12 concerning because you are further down the line and 13 these issues are being experienced? 14 A. But -- yes, but if it was a fundamental design flaw 15 because of insufficient requirements analysis by ICL, 16 that would have been apparent earlier on -- 17 Q. Now that -- 18 A. -- logically. 19 Q. Now that they have found these issues, whose 20 responsibility within the Post Office would it have been 21 to take those forward? 22 A. To consider it? 23 Q. Yes. 24 A. Dave Miller. 25 Q. You said earlier that, at this stage, I think you were 84 1 dealing with strategic matters and David Miller was 2 dealing with the more technical matters? 3 A. He was dealing with delivery. I was in my last -- as 4 I said earlier on, I was appointed to a role completely 5 outside of the Post Office Network when the entire group 6 reorganised itself, as a managing director of a new unit 7 in March. 8 Q. Who would you have expected David Miller to have brought 9 this to their attention? 10 A. Maybe Stuart Sweetman. I don't know. 11 Q. Do you think it was sufficiently serious to bring to the 12 managing director's attention? 13 A. You need to ask Dave that. 14 Q. If you had received it at the time, would you have 15 brought it to the managing director's attention? 16 A. Depends in what context. As I said, I would have 17 probably talked at length to the person authoring the 18 report first to understand it because, as I said, I had 19 reservations on it and I seriously don't remember 20 commissioning, even though, no doubt, we'd have paid 21 half of it. 22 Q. Do you think that by the end of 1998 and the beginning 23 of 1999 technical issues were being taken seriously 24 enough within the Post Office? 25 A. They would have been taken seriously. I maintain, as 85 1 I said, we would not have been expedient for the sake of 2 speed or trying to shoehorn a solution that suited us 3 strategically if we didn't think it could work. We 4 relied on -- you'd have seen Dave Miller's notes to us. 5 So we knew it, "no denial", as I said. I know I'm 6 sounding like I'm in denial about this report. It's the 7 first time probably today but I'm irritated by it. 8 We then -- you will, no doubt -- well, maybe you 9 won't, there's another document that Dave Miller wrote 10 in April to Vince Gaskell. That's POL00028407, where he 11 says the team -- following these issues, the team now 12 believe they're ready to be able to roll out, and 13 I would have relied on Dave's judgement about that, 14 standing from afar. As I said, I was in the process -- 15 I had a new job but I was hanging on for a couple of 16 months because the ministerial decision didn't come 17 until May, so I was still giving input at that point. 18 Then the next I saw was something in a note from 19 Keith Hardie, where I was a copy amongst many others, as 20 I said in my witness statement, that the Post Office was 21 beginning to roll-out Horizon. That's POL00028463. 22 Q. Let's stay in January 1999 for now and just look at 23 POL00031230. This was a report by Post Office's POCL's 24 finance director. 25 A. This is the one I mentioned earlier. 86 1 Q. Yes. What was your involvement in this particular 2 document? 3 A. Well, there was an earlier -- we called these peer 4 reviews. There was an earlier one done in '97 by the 5 strategy director of Post Office. This was done by the 6 finance director -- sorry, Post Office Counters, 7 I should say -- and we asked him, as a contingency, to 8 look at, as I said earlier, whether or not -- without me 9 getting -- I mean, he'd have talked to me and given 10 input but without me trying to influence his decision or 11 his analysis, in any way, shape or form, really, to give 12 an opinion on what the best way forward for POCL would 13 have been if this scenario occurred. 14 Q. Can we look at page 2, paragraph 2.5 and 2.6: 15 "Even on the basis of protecting benefit payments, 16 the go/no go decision is finely balanced, with neither 17 option being fully satisfactory for POCL. Proceeding 18 means full commitment to an automation route and 19 a partner, neither of which are ideal in the context of 20 realising the new Counters vision. But not proceeding 21 would so delay the building of automated capability, and 22 undermine the business' credibility internally and 23 externally, as to put the vision at significant risk of 24 becoming undeliverable." 25 A. I think I said earlier it was the least worst option. 87 1 Q. "Several senior managers, close to the project, but no 2 principal negotiators, whose judgement I respect, 3 express significant reservations about the risks of 4 proceeding. These centre of their continuing doubt 5 about the ability of ICL to deliver a satisfactory 6 product; the absence of transparency in the PFI 7 contract; the risk that ICL's financial fragility will 8 endure throughout the project, with the possibility of 9 repeated claims on The Post Office for extra 10 contributions (which, by then having no alternative, it 11 will be unable to resist); and doubts about POCL's own 12 ability to give it the focus essential for success." 13 A. Yes, none of that is new news, though, is it, really? 14 I mean, the track record demonstrates that. 15 Q. Over the page, the decision is: 16 "On balance, I agree that it remains right to 17 press ahead with Horizon, despite the extra costs 18 involved." 19 Did the commercial importance of seeing Horizon 20 through at that stage outweigh the kinds of technical 21 concerns that we've talked about this morning? 22 A. No. I mean, as I said, the decision was in a round and 23 a business decision always is surrounded by both 24 technical, commercial, operational and financial 25 aspects. That's a matter of judgement for general 88 1 managers but we would not -- I don't believe -- well, 2 I find it really hard to believe that anyone from that 3 culture, at that time, would have compromised quality 4 knowingly, in order to be expedient to suit strategic or 5 financial matters. 6 Q. I'm only going to ask about a couple more documents. 7 The first is a return to your witness statement, 8 WITN04030100 and it's page 33, paragraph 94. 9 A. Yes. 94? Okay. 10 Q. Yes, I'm just going to -- can we bring that on the 11 screen? Thanks. 12 So what you have said there is you were less 13 involved in the Horizon project as 1999 proceeded and, 14 by May 1999, you were no longer involved? 15 A. At that time, yes. 16 Q. Can we look at POL00021469. 17 A. You are going to show me a document where I was, okay. 18 Q. Well, it's a board meeting. 19 A. Is this the one in 2000? 20 Q. Yes. 21 A. That's me going, though, in my new role as MD customer 22 management and the board asking -- I'm an attendee, 23 aren't I, along with other people like, I don't know, 24 someone else from -- yeah, Basil Larkins, the managing 25 director of Network Banking. You see, we all had 89 1 different titles then because we are one of the 2 17 managing directors under it. 3 Basil was the Network Banking person. I was the 4 managing director of customer management and, within 5 that customer management, there was a peripheral -- 6 well, not peripheral, but a smaller part of the job was 7 to look at the opportunities for electronic government 8 with the Post Office in general, including perhaps at 9 counters. 10 Q. You would have been present for the whole meeting there? 11 A. I doubt it. 12 Q. Would you have stayed for the part that addresses 13 Horizon? 14 A. I doubt it. I mean, the normal way -- I don't know. 15 I can't remember but the normal way of board meetings, 16 if you weren't a board member, was that you attended for 17 the item in question that you were asked to present. 18 Later on, when I was doing the six-month job as 19 the acting MD and I attended board meetings, I would 20 have been present at those board meetings. But that's 21 later than that. 22 Q. Can we look at page 7, the bottom of page 7. I think 23 this is a document that we went through yesterday and 24 I'm not going to spend any time on it but this is the 25 part where discussions turned to the commercial 90 1 development of Horizon and things had moved on. Were 2 you aware, at least at that time, that things had moved 3 on to commercial exploitation? 4 A. Yes. I mean, I'd have been away from Post Office 5 Counters Limited entirely, or the three business units 6 that it had been split into by then, including Post 7 Office Network, which Dave Miller was the managing 8 director of, and he would have -- you know, we'd have 9 met from time to time and he'd have said, "Well, we're 10 now rolled out to 4,000 offices". But clearly at that 11 point, I certainly was completely unaware of any 12 material accounting or balancing-type issues at that 13 point and, therefore, my brief was, as part of my new 14 job, to have a team looking at Government Gateway 15 opportunities. 16 Q. Who, if anybody, would you have passed on your knowledge 17 to about those technical issues that you gathered -- 18 A. Oh, Dave Miller, obviously. We had that large overlap, 19 didn't we, and there would have been quite a lot of 20 continuity in the team from the team that were brought 21 back from the PDA in to work for Dave. I mean, clearly 22 Stuart Sweetman. You'll be talking to Stuart, I think, 23 later on. Stuart was still the group MD for those 24 matters that included Post Office Network and Network 25 Banking so he will have had a rounded view about 91 1 Horizon. 2 Q. If there were concerns about Horizon at that stage -- so 3 March 2000 -- who do you think would have or should have 4 raised them at the board level? 5 A. Well, I imagine the way it works, certainly when I was 6 managing director of customer management, Stuart was 7 also my group MD then, who sat on the executive board 8 alongside John, you know, as the chief executive, and we 9 would have had regular contacts on how the big issues in 10 my patch were going and if Stuart thought they had been 11 serious enough or if I'd have proposed that I wanted 12 board support, I would have expected the channel to go 13 through him. 14 Q. Finally from me, I think you have asked to very briefly 15 address the Chair in respect of your overall 16 reflections. 17 A. Yes, Chair. All I wanted to say was, really in line 18 with my witness statement, that I fully respect this 19 Inquiry and really hope that you get to the bottom of 20 it. I have been reflecting hard ever since I've been 21 asked to come here as a witness, back in May, with all 22 the documents and I want to express my own sadness about 23 the impact this has put on so many lives. 24 In some of the stuff I've seen, either through the 25 press or through recent documents that I've seen from 92 1 later periods, I'm frankly baffled and shocked by how 2 some of these people were treated and I'm completely 3 baffled and it's not a culture I personally recognise in 4 terms of the ethos of what must have been happening at 5 that time or the governance that allowed it. 6 So that's all I wanted to say. 7 Q. Thank you very much, Mr Rich. 8 I've just been given one very brief question that 9 touches on earlier matters and it relates to PFI. You 10 have, throughout your evidence, referred to the 11 difference that PFI made to the process. Can you very 12 briefly just tell us what you mean by that. 13 A. Well, contractually, it meant, as I said, the risk 14 was -- for design, operate, build and run was with the 15 supplier, not with the procurer. In addition, because 16 of this particular PFI, that also included the risk 17 around benefit fraud occurring at post offices, which 18 I'm not aware of any other PFI-type project in the world 19 that did that. 20 It also meant that the supplier only got paid on 21 outcomes and outputs. So until -- for example, as 22 I understand it, until the system was accepted before 23 roll-out, there were no substantial real payments to ICL 24 so, clearly, their cash flow projections from the start 25 of it would have been completely up the wall, as we 93 1 heard earlier from their parent company to John 2 yesterday. 3 The only other thing -- two other things I'd say. 4 One is that, therefore, I think it was a learning 5 experience because the nature of this PFI project was 6 unusual. It was usually used by Government for capital 7 projects that were properties or leases, or something 8 like that, where there's some more secure -- it wasn't 9 really operational. So we were learning about the 10 boundaries between being able to go in and assure 11 ourselves of the details of the design, where ICL would 12 have said, "No, under PFI you don't do that, wait for 13 the outputs and test them". 14 The final thing I'd say is that, of course, this 15 project straddled two governments. So PFI was a Tory 16 policy, you know Treasury guidance in the late '80s. 17 New Labour came in, middle of '97, and so there was some 18 concern, certainly among subpostmasters and others and 19 no doubt ICL, about what the attitude to a PFI project 20 would be, whether it would be within policy for them to 21 do it and I think it morphed into what New Labour called 22 a public sector/private sector partnership. 23 So I think it got re-badged, really, and they 24 early on committed -- I can remember reading the 25 document in the bundle with a meeting with 94 1 subpostmasters, with the Federation actually, with 2 a minister assuring them that it wouldn't be pulled on 3 the grounds that it was a PFI project. 4 MR BLAKE: Sir, I believe there may be a question or a short 5 series of questions from Mr Stein; is that correct? 6 MR STEIN: Sir, yes. 7 SIR WYN WILLIAMS: All right. Yes, ask the questions 8 please. 9 Questioned by MR STEIN 10 MR STEIN: Thank you, sir. 11 Mr Rich, my name is Sam Stein, I represent 12 a number of postmasters, mistresses and managers in 13 relation to what happened through the use of the Horizon 14 System. 15 A. Yes, Mr Stein. 16 Q. You have just provided some answers to Mr Blake that 17 touch upon the question of PFI? 18 A. Yes. 19 Q. When you were giving your evidence earlier today, you 20 spoke about the Bird & Bird document you were being 21 referred to and you made this comment that it completely 22 ignores the fact that it was under a PFI. Does that 23 mean that the burden of governance running and general 24 maintenance of the Horizon System was placed upon 25 Fujitsu? 95 1 A. Primarily, yes, but I was also saying it in relation to 2 the supposed innovation that they could bring in 3 designing and developing the system. So a way of -- the 4 report talked about a requirements analysis that was 5 allegedly standard industry practice, which may or may 6 not have been true, I don't know. I don't know if that 7 was an outdated one but my guess is that ICL would 8 argue -- I'll let them argue for themselves -- that they 9 did do sufficient requirements analysis once they did it 10 and they applied an innovative way of developing. 11 Q. You also make some general comments that you set out, 12 regarding your bafflement regarding what happened to 13 subpostmasters. I'm just going to remind you of a part 14 of your own statement at paragraph 116, page 39 of your 15 statement. You say this: 16 "I am truly baffled by the apparent later 17 professional advice, investigative processes and 18 governance that appears have led to so many unjust 19 prosecutions." 20 Can we unpick that please. What the later 21 professional advice that you're referring to? 22 A. I'm only reading what I've read in the media and in the 23 judgment that was -- that was in the opening counsel's 24 statement, that it would seem -- I don't know. I mean, 25 Post Office Limited will now speak for itself but it 96 1 would seem that it was ill-advised in going ahead. The 2 investigative processes seemed to be ones that I didn't 3 recognise from my days. Certainly, non-disclosure would 4 have been a real issue to me with any investigative 5 processes around fraud. The governance processes, I'll 6 leave that to Post Office Limited and perhaps UKGI 7 because I can't see how a board would have knowingly -- 8 knowingly -- not noticed that 700 people had been 9 prosecuted cumulatively, roughly. 10 Q. Regarding you mention in your statement and just now of 11 the investigative processes, and also mention of 12 governance, whilst you were at the Post Office and 13 dealing with matters up until, I think, 2002, where you 14 moved over to the RM group more generally until 2005, 15 what control was put in place in relation to the 16 investigative processes as regards subpostmasters, 17 mistresses and managers? 18 A. Well, the investigative -- by the way, I wasn't involved 19 continually until 2002, just to be clear on the records. 20 The investigative processes were largely as John 21 described yesterday, really. You know, the 22 investigative processes were done separately and at 23 arm's length but if there'd have been a high -- a sudden 24 hiatus or peak of them I'm sure it would have been 25 brought to the attention through the relevant senior 97 1 manager who would have been in control of that. 2 Q. You say that they were done at arm's length. What does 3 that mean to the disclosure process of bugs and errors 4 within the Horizon System? Were those bugs and errors 5 brought to the attention of the arm's length body 6 dealing with the investigation of subpostmasters and, if 7 so, how? 8 A. I don't know because, at the time I was there, I don't 9 think there were bugs and errors -- I don't know. 10 I absolutely don't -- I can't be absolutely certain but, 11 by the time I left in this period, in this phase, in 12 March '99 I don't think -- you might correct me -- 13 I don't think in the live trial offices there were any 14 prosecutions brought. 15 Q. Well, I could correct you. There were bugs and errors 16 in place -- 17 A. No, I didn't say that. I didn't say that. I didn't say 18 there weren't bugs and errors. I said bugs and errors 19 that led to investigation and prosecution. 20 Q. Mr Rich, my question was in relation to what controls 21 were put in place to make sure that bugs and errors were 22 brought to the attention of the investigation processes. 23 What can you help regarding that? 24 A. Well, we wouldn't have -- during that time if there were 25 bugs and investigations that we thought had technical 98 1 issues related to them, which was the point of the live 2 trial, as I said, and therefore needed fixing, we 3 wouldn't have put those to the investigation department. 4 Q. How did the investigation department get to learn about 5 bugs, problems, issues with the Horizon System? 6 A. I don't know. 7 Q. What's the system? 8 A. I don't know. 9 Q. Who's in charge of it? 10 A. I don't know. 11 Q. Who should have been in charge of it? 12 A. I don't know. I can't help you. 13 MR STEIN: Thank you. Thank you, sir. 14 SIR WYN WILLIAMS: That concludes the questioning, I take 15 it, and, assuming that's the case, thank you very much, 16 Mr Rich, for making a detailed written statement and 17 also for coming to give oral evidence. I'm grateful to 18 you. 19 A. Thank you. Thank you for the opportunity. 20 MR BLAKE: Thank you, sir. It's now lunchtime. Could we 21 come back at -- would it cause anybody inconvenience, 22 including yourself, if we came back slightly earlier, 23 perhaps at 1.50? 24 SIR WYN WILLIAMS: No, I was going to suggest that we (a) 25 make a determined attempt to finish Mr Copping this 99 1 afternoon and, therefore, (b) if it helps to have 2 a shorter lunchtime we should do so. 3 MR BLAKE: Excellent. Thank you very much, sir. We'll come 4 back at 1.50. 5 SIR WYN WILLIAMS: Fine. See you all then. 6 (12.55 pm) 7 (Luncheon Adjournment) 8 (1.50 pm) 9 MS HODGE: Good afternoon, sir, can you see and hear me? We 10 can't hear you. You appear to be on mute. 11 SIR WYN WILLIAMS: Well, I didn't think I was on mute. Am 12 I on mute now? 13 MS HODGE: No we can hear you perfectly, thank you. 14 SIR WYN WILLIAMS: Fine. Our next witness is Mr Copping. 15 PETER JAMES COPPING (sworn) 16 Questioned by MS HODGE 17 MS HODGE: Please give your full name. 18 A. Peter James Copping. 19 Q. You should have in front of you a witness statement, 20 dated 2 September of this year. 21 A. Yes. 22 Q. Could I ask you please to turn to page 18 of your 23 statement. Do you see your signature there at the end 24 of the statement? 25 A. Yes. 100 1 Q. Is the content of the statement true to the best of your 2 knowledge and belief? 3 A. Yes. 4 Q. Mr Copping, your statement and its exhibits are now in 5 evidence before the Inquiry. I would like to begin by 6 asking you a few questions about your professional 7 background. You qualified as a chartered engineer and 8 a Fellow of the Institute of Engineering and Technology; 9 is that correct? 10 A. That's correct. 11 Q. What competencies were you required to demonstrate to 12 qualify as a chartered engineer? 13 A. It's a long process but, in essence, you have to display 14 technical competencies, managerial competencies, in 15 quite a wide range of topics. 16 Q. You've explained that you worked in the electronics and 17 telecommunications industry for approximately ten years 18 before joining PA Consulting; is that right? 19 A. That's correct. 20 Q. PA Consulting being a management information and 21 technology consultancy? 22 A. That's correct. 23 Q. You joined that organisation in 1976? 24 A. Yes. 25 Q. Before being appointed a director of PA Consulting in 101 1 1990, you worked on a variety of telecommunications and 2 information technology projects; is that right? 3 A. Yes. 4 Q. Did these projects require you, at any stage, to 5 undertake software design and development? 6 A. Personally, no, but I did lead teams that were doing 7 that. 8 Q. How would you characterise your area of expertise in 9 engineering? 10 A. Broadly speaking, I would characterise it as in the 11 telecommunications area, networking and IT. 12 Q. You first became involved in Horizon when you were 13 commissioned in the summer of 1997 to leader of the 14 PA Consulting of what was known at the time as the 15 Benefits Agency and Post Office Counters programme; is 16 that correct? 17 A. That's correct. 18 Q. Had you ever previously worked on a project of the scale 19 and complexity of Horizon? 20 A. Not quite the same. I certainly worked on large 21 projects of similar scale in the mobile 22 telecommunications area in particular. 23 Q. Did you have any prior experience of working on an IT 24 system developed by ICL? 25 A. No. 102 1 Q. I'd like to, if I can, briefly explore what you 2 understood at the time about the broader context of the 3 review that you were asked to undertake. Why had that 4 review been commissioned? 5 A. Sorry, could you repeat that? 6 Q. Why had your review, in the summer of '97, to your 7 understanding, been commissioned? 8 A. Primarily because of delays to the project. 9 Q. What had arisen as a result of those delays? 10 A. There were concerns about the possibility of future 11 delays, there were concerns about Pathway's ability to 12 deliver and there were concerns about Post Office 13 readiness to accept Horizon. 14 Q. In your statement you describe the purpose of the review 15 as being to identify the reasons for the delay to the 16 project and to recommend actions to de-risk the project 17 to bring it back on track; is that correct? 18 A. That's correct. 19 Q. You were also required, were you not, to make 20 an assessment of the programme's future delivery 21 capability? 22 A. That's correct. 23 Q. That assessment involved examining not only management 24 and resourcing issues but also the technical aspects of 25 the project, which had a bearing on the programme's 103 1 ability to deliver its end-to-end delivery obligations? 2 A. That's correct. 3 Q. You've explained in your statement your review focused 4 on four principal areas. These were the business 5 objectives of each stakeholder; the contractual 6 arrangements between the parties; thirdly, the programme 7 management processes; and, finally, the technical 8 infrastructure proposed for Horizon by ICL Pathway. Is 9 that right? 10 A. Yes. 11 Q. You use a term "technical infrastructure" in your 12 statement. Can you explain what you mean by that? 13 A. Essentially, that is the hardware platform on which the 14 software services reside from the counter back into 15 various back-end systems. 16 Q. Are you suggesting, therefore, that you were only asked 17 to consider the hardware, as opposed to the software? 18 A. No, no. 19 Q. Elsewhere in the documents we see the term "technical 20 architecture" used. So far as you're concerned, are 21 they one and the same: architecture and infrastructure? 22 A. The architecture defines the various layers in the 23 system that worked together to make up the 24 infrastructure. 25 Q. So slightly different nuances then? 104 1 A. Yes. 2 Q. Do you recall how far the project had progressed at the 3 point at which you undertook your review in the summer 4 of 1997? 5 A. When we started work there was a -- I think we were 6 presented with a programme rework which was titled 7 "Version 3", and I think all of our work was based on 8 that particular document. 9 Q. It might assist if we bring that up. That's 10 POL00028186, please. Is this the document to which you 11 were referring? 12 A. Yes. 13 Q. The "Programme Delivery Authority Master Plan 14 Version 3" -- 15 A. Yes. 16 Q. -- dated 8 April 1997. Could we turn to page 8, please. 17 You see there a number of strategic milestones in the 18 project. Can you see that in front of you? 19 A. Yes. 20 Q. The first of which was the initial Go Live implemented 21 in one post office on 23 September '96 and then rolled 22 out to ten post offices on 23 October '96? 23 A. Yes. 24 Q. That was followed by the roll-out of Pathway 25 infrastructure on 7 March '97, so we see next to B1? 105 1 A. Yes. 2 Q. And B2, the release of what became known as software 3 release 1b, which, as we can see, implemented OBCS 4 functionality. Can you describe what OBCS was? 5 A. It was order book CS, something. I don't remember. 6 Q. The control service? 7 A. Control service, that's right, yes. 8 Q. Were you aware of what its function was? 9 A. This was the service that was used to confirm that the 10 person in the Post Office was entitled to the benefit 11 that was on the order book, as I understood it at the 12 time. 13 Q. We can see a further date of 30 June '97 about midway 14 down the page and that was the planned release of 15 Pathway Release 1c, which was due to contain further 16 OBCS -- so the order book control service -- and BPS, 17 which was the Benefit Payment Service, functionality. 18 We know, however, that milestone had been missed 19 because, at the point at which you conducted your 20 review, development work on Release 1c was ongoing. Is 21 that consistent with your recollection? 22 A. That is correct. 23 Q. What did you understand at the time about the state of 24 development of the Post Office counters functionality, 25 by which I mean EPOSS, the Electronic Point of Sales 106 1 Service, and APS, the Automated Payment Service? 2 A. My goodness, I really don't remember. 3 Q. In your statement you explained you adopted two 4 principal methods of assessment, the first conducting 5 a series of in-depth interviews and follow up 6 investigative meetings with senior figures in each of 7 the stakeholders, those being the Programme Delivery 8 Authority, Post Office Counters, the Benefits Agency, 9 Pathway and ICL; is that right? 10 A. That's correct. 11 Q. The other aspect of your review or your assessment was 12 a document review essentially; is that right? 13 A. Yes. 14 Q. In that you reviewed a significant amount of 15 documentation relating to technology status and plans 16 for Horizon? 17 A. Yes. 18 Q. If we could start by addressing the second of these 19 methods, your document review, could you please describe 20 the types of technical documentation to which you were 21 granted access to ICL Pathway? 22 A. First of all, we would have started with a demonstration 23 of the model office system. We would have taken 24 presentations from ICL on the software status, status of 25 development, the overall architecture, the way the 107 1 system was supposed to work and their view of the 2 current issues in the program. 3 Q. What you've just described there, it sounds mostly like 4 a practical demonstration and oral presentations, rather 5 than an analysis or a review of documents. Did you 6 carry out such an analysis? 7 A. Yes, there were analyses undertaken by members of the 8 team on technical documentation, mostly in the software 9 area, also with Escher. 10 Q. So it was other members of your team, employees of 11 PA Consulting, who looked at the more technical aspects? 12 A. Yes. 13 Q. Do you recall what, if anything, they told you about the 14 completeness or quality of the design documentation that 15 was shown to them? 16 A. Sorry, I don't recall that level of detail, I'm afraid. 17 Q. If we turn back to your first method of assessment to 18 which you referred in your statement, the interviews 19 which you conducted, you described carrying out more 20 than 30 face-to-face meetings and interviews; is that 21 right? 22 A. Yes. 23 Q. Do you recall the names of those whom you interviewed? 24 A. They are all listed in the report. 25 Q. Could we possibly bring that back up POL00028092. I say 108 1 "bring back up", this is for the first time. Thank you. 2 At the conclusion of your review, you prepared 3 a written report; is that right? 4 A. Yes. 5 Q. The third and final version of that is dated 1 October 6 1997. 7 A. That's correct. 8 Q. That's the report to which you just referred? 9 A. Yes. 10 Q. If we could turn to page 48, please, of the report, we 11 can see here Appendix A, a list of those whom either you 12 or your colleagues interviewed in connection with this 13 review; is that right? 14 A. I probably met most of the people on that list myself at 15 some stage, either in individual meetings or group 16 meetings with ICL Pathway particularly. 17 Q. You've explained in your statement that some of the 18 in-depth technical interviews were attended by 19 specialists employed by PA Consulting; is that right? 20 A. Yes. 21 Q. Why did you consider it necessary to bring in 22 specialists to conduct those technical interviews? 23 A. It's a way of working to ensure that we cover the ground 24 appropriately. 25 Q. Did you yourself have the necessary expertise to deal 109 1 with the more technical aspects of the project? 2 A. Did I? 3 Q. Did you have the necessary expertise to deal with the 4 more technical aspects of the project? 5 A. On the networking and architectural issues, yes, but on 6 the software aspects and, particularly in regard of 7 Escher and the processes ICL were using for development, 8 no. 9 Q. You were therefore reliant upon your colleagues? 10 A. Correct. 11 Q. In your report, you identify a number of concerns about 12 technical issues with Horizon, which were raised by 13 senior figures in the Benefits Agency, in Post Office 14 Counters and Pathway. Do you recall the nature of those 15 concerns? 16 A. In the report? 17 Q. Yes. 18 A. I'd have to read the report again. 19 Q. If it assists, at page 28, please. Thank you. 20 At the bottom of page 28 there's a paragraph 3.3.5 21 entitled "Technical issues". It records: 22 "Concerns have been expressed to us about the 23 ability of the solution to meet the security 24 requirements, whether it is scalable to support a 40,000 25 terminal network and what performance will result. 110 1 Concern has been increasing with failures in test and by 2 regular requests by Pathway for exclusions to key 3 releases, mainly concerned with security features." 4 Is that consistent with your recollection? 5 A. Yes. The security issue was a particularly difficult 6 one, I think, for Pathway because I would say there were 7 so many moving parts. My recollection is that the 8 security requirement was made increasingly more 9 demanding as it became aware of the risks and the risk 10 transfer arrangements in the PFI contract to ICL. 11 Q. Could we please turn back to page 8 of the report, where 12 we see a part of your management summary. In relation 13 to Pathway, at M3.4 -- excuse me, if we can scroll down 14 a little bit -- you've, observed here, in the bottom 15 paragraph: 16 "We believe the current status of the Programme is 17 surrounded by considerable contractual ambiguity, 18 Pathway are, in essence, proceeding on an 'own risk' 19 basis to deliver Release 1c with a 'known problems 20 register' and its proposal is to address the 'known 21 problems' in Release 2." 22 Were you shown a copy of the known problems 23 register? 24 A. I'm sure we were. 25 Q. What did you understand the purpose of the register to 111 1 be? 2 A. This would have been issues that were expected to take 3 longer to resolve than the plan allowed and, therefore, 4 they'd be deferred into a subsequent release. 5 Q. When you say "issues", what types of issues did you 6 understand? 7 A. Development releases, essentially, that would take 8 longer to work through. 9 Q. Are we talking about problems in the software, bugs and 10 errors and defects, things of that nature? 11 A. Not necessarily problems. More likely 12 an underestimation of the effort and time required 13 against what I recall was, in some instances a moving 14 requirement over time, and I would put the security 15 requirement into that category, for example. 16 Q. So your understanding, essentially, was these were 17 generic problems with the software release, rather than 18 specific issues that had been identified? 19 A. I think that's a fair description, yes. 20 Q. In your report, you identified another significant 21 concern on the part not only of the sponsors but also of 22 Pathway, which related to the robustness of the 23 technical architecture; is that fair? 24 A. Yes. 25 Q. Can you describe the nature of the concerns which were 112 1 articulated to you? 2 A. At the time we did our review, my recollection is that 3 Pathway's approach to testing was to test individual 4 components and then fit them together and retest and it 5 was at that stage that I think we detected there were 6 certain concerns that, when everything was put together, 7 it might not be as robust as perhaps was expected. 8 Q. Do you recall whether the concerns expressed to you 9 related to any particular component or whether it was 10 a more general concern about the overall architecture? 11 A. There was a particular concern about Escher and, if 12 I recall correctly, Pathway did institute a rework of 13 Escher's software, as a result of the issues they were 14 experiencing there. 15 Q. What steps did you take to investigate these particular 16 concerns that had been articulated? 17 A. We would have pursued them with further investigations 18 of any documentation that was available and further 19 face-to-face interviews. 20 Q. What conclusions did you ultimately reach at this stage 21 about the robustness of the architecture? 22 A. That it would be -- I think we took the view that the 23 overall system was achievable in development terms. The 24 question was how long it was going to take before it 25 became completely reliable and robust. 113 1 Q. I'd like to turn now to your findings about the causes 2 of the chronic delays to the programme. In your 3 statement you describe the contractual arrangements 4 between the parties as being a significant cause of 5 initial delays to the programme. Can you explain the 6 basis of that conclusion? 7 A. The contract was let under a private finance initiative 8 which is where the risk of delivery is transferred to 9 the supplier and this particular contract was unusual, 10 in that there were differing business objectives between 11 the sponsors, and that created a lot of tension between 12 the parties -- and I think I've lost the thread of where 13 I was going on that. 14 Q. That's okay. So I think you've explained that, 15 essentially, much of the detailed specification for the 16 contract hadn't been agreed at the point at which -- 17 A. Well, there were a lot of agreements to agree and 18 I never was really sure why that was allowed to happen 19 and, normally, one would expect, in a PFI contract, that 20 the supplier would be allowed to work up, once the 21 contract has been let, the proposition to deliver for 22 the outline requirement. 23 But because of the agreements to agree 24 arrangement, it took a long time for ICL to work through 25 each individual part of the system to decide what 114 1 actually needed to be delivered and, at one point, 2 I recall they proposed to use a rapid application 3 development method, which was becoming -- just coming 4 into favour around that time, which allowed the 5 developer to work closely with the sponsoring party to 6 explore how an application might work and I think ICL 7 were unsuccessful in pursuing that particular approach 8 because of the reluctance of sponsors to become engaged. 9 Q. Did you consider the rapid application development 10 technique to be suitable for a project of this scale and 11 complexity? 12 A. I couldn't see a reason why it wouldn't be, provided the 13 parties were happy to pursue that particular approach. 14 It's probably worth saying that there were many 15 occasions where we were coming to a view that this was 16 being treated by the sponsors and particularly the 17 Benefits Agency as a supply and build contract rather 18 than a PFI contract where there was a lot of 19 intervention from the sponsors because they weren't 20 necessarily happy, for reasons of their own, with what 21 was going on in the development activity and that in 22 itself caused delays. 23 Q. Another factor which you identified as causative of 24 delay and which related to the parties' contractual 25 arrangements, concerned ICL's original assessment of the 115 1 development work and resources required to deliver the 2 system. What conclusions did you reach in that regard? 3 A. Well, ICL told us themselves that they had seriously 4 underestimated the amount of work required, despite 5 quite a long and protracted selection process as 6 I understood it and, again, I think that was partly 7 because of the complexity of the system. I think at 8 this stage no-one really fully understood the 9 implications of the end-to-end arrangements and the 10 necessary interfaces to all of the systems outside of 11 the Horizon project that needed to interface in order to 12 make it all work and I think that, together with the 13 agreements to agree issue, was a key cause for the 14 delays. 15 Q. I'd like to turn now to the findings you made at this 16 stage about the programme's management capability. In 17 your report you express concerns about the resourcing of 18 the programme and in particular about the level of 19 managerial expertise within Post Office Counters; is 20 that fair? 21 A. Yes. 22 Q. Would it be fair to say you also expressed some 23 criticism of the programme delivery authority? 24 A. Yes. 25 Q. In your report you describe the PDA as focusing almost 116 1 exclusively on achieving a high quality outcome, even 2 potentially at the expense of timeliness and cost 3 effectiveness. 4 A. Yes. 5 Q. That was one of the concerns you had at the time? 6 A. Yes. 7 Q. Was it your perception that a more pragmatic approach 8 needed to be adopted, with trade-off's being made 9 between the performance of the product on the one hand 10 and the business impact of delays -- 11 A. I think that's fair, yes. 12 Q. You ultimately concluded that there was no sensible way 13 of de-scoping or radically altering the plan and that it 14 was better to continue than to terminate; is that 15 correct? 16 A. Yes. 17 Q. Did you understand the parties to be contemplating 18 termination at this stage? 19 A. There were veiled implications of termination, yes, at 20 that stage for default against the contract by ICL, as 21 I recall it. 22 Q. On 24 September '97 you presented a summary of your 23 findings to the PDA board -- 24 A. Yes. 25 Q. -- which they accepted, I believe -- 117 1 A. Yes. 2 Q. -- and the recommendations which you have made to 3 minimise further delays to the programme. I think that 4 was followed up by a meeting, a special meeting, of the 5 PDA board on 2 October '97; do you recall that -- 6 A. Yes. 7 Q. -- where it was agreed that you would lead a series of 8 workshops to examine the strategic risks to the 9 programme? 10 A. That's right. 11 Q. As well as the root causes of delay. 12 A. Yes, that was essentially to ensure that there was 13 a common agreement on the issues before the parties 14 decided to proceed and investigate what they needed to 15 do. 16 Q. Did you regard those workshops as a success? 17 A. They were, yes. 18 Q. If I can move on now to the next significant stage of 19 your involvement which came in the spring of 1998, you 20 became involved in the programme again, on this 21 occasion, at the behest of Her Majesty's Treasury; is 22 that right? 23 A. Yes. 24 Q. You were appointed to act as a consultant to an expert 25 panel, chaired by Adrian Montague, the head of the 118 1 Treasury private task-force on private finance. The 2 panel, as I understand it, had been established to 3 review the deliverability of the Horizon project 4 together with the risks associated with the estimated 5 timescales and cost of the programme. Is that a fair 6 characterisation of their function? 7 A. Yes. 8 Q. So this is spring '98. Do you recall the extent of 9 progress which had been made on the programme by that 10 stage? 11 A. No. 12 Q. I think you might be assisted if I were to refer you to 13 the written report which was produced by the expert 14 panel at the conclusion of their review. I think it's 15 correct that you weren't the author of that report but 16 your findings contributed or were taken into account 17 when that report was written; is that right? 18 A. I didn't know that there had been a report written at 19 the time but I have now seen it and it certainly concurs 20 with my understanding of what the Panel was going to say 21 at the conclusion of our work. 22 Q. Please could we pull up POL00028094. This is a copy of 23 the report. We can see the composition of the panel 24 there, Adrian Montague, Bill Robins and Alec Wylie, with 25 whom you collaborated in the review. 119 1 A. Yes. 2 Q. Can we please turn to page 11. At point A, at the top 3 of the page, there's a heading "Current Status of the 4 Programme". Paragraph 21 reads: 5 "The programme has moved on since PA reviewed it 6 towards the end of 1997", and cites the following 7 indicators of progress: 8 "Release 1c ... a partial solution providing the 9 benefit payment card and [order book control service] 10 has been working satisfactorily in just over 200 offices 11 since November 1997. Pathway has brought in new 12 technical skills and management resources, increasing 13 headcount to around 270 staff and introduced new 14 procedures to support the high level of software 15 development needed; 16 "BA has increased its resources on the programme 17 and Release 3.0 of its key feeder systems (CAPS) has 18 been given DSS [the Department of Social Security] Seals 19 of Approval; 20 "[Post Office Counters Limited] has also increased 21 is resources on the programme, establishing a pilot 22 service management function and a National 23 Implementation organisation, to support Pathway in 24 preparing outlets and training; the Horizon Programme 25 Office [referred to as the HPO] also started work on 120 1 1 April 1998." 2 Does that reasonably encapsulate where things had 3 got to by spring of '98? 4 A. It's a good reminder. 5 Q. You were instructed by the panel to undertake a number 6 of investigations into issues which you've identified in 7 your statement. These were the extent to which Horizon 8 was future proofed; whether it had the capability to 9 support the Electronic Point of Sales Service; whether 10 it could be developed to support simple banking 11 applications; the likely lifetime of the technology; and 12 whether the technology was suitable for long-term 13 government infrastructure. Does that encapsulate the 14 areas? 15 A. That's correct. 16 Q. You have explained that, in carrying out these 17 investigations, you held a series of meetings and one to 18 one discussions with ICL -- 19 A. Yes. 20 Q. -- sorry, with ICL Pathway, as it was, and had extensive 21 engagement with each of the parties over several weeks. 22 A. Correct. 23 Q. On this occasion, did you carry out any extensive 24 analysis of the underlying documents? 25 A. No. 121 1 Q. Does it follow that your findings then were based, to 2 a very great extent, on the information that you were 3 given by the participants? 4 A. Yes, that's correct, together with external research 5 that we would have done as to the state of play in the 6 deployment of EPOSS, simple banking, and so on, and the 7 comparison of that with what the system was capable of 8 doing. If I recall correctly, we said we felt that the 9 system could be developed to support those applications; 10 the only question was how long it was going to take. 11 Q. So if I can just clarify, I what I understand you to 12 have explained is that, essentially, you weren't looking 13 at the parties' underlying documents, you weren't 14 carrying out analysis of design documentation, and so 15 forth, but you were carrying out some external research, 16 in order to, I suppose, analyse what you were being 17 told? 18 A. Yes, we were calibrating the art of the possible against 19 what we were being told by ICL Pathway. 20 Q. To what extent did your discussions with ICL Pathway 21 touch upon problems that had arisen during the 22 development of the EPOSS application? 23 A. Not at all. In fact, I don't think the Montague review 24 looked at issues at that time, problems, at all. 25 Q. In your statement you say you have no recollection of 122 1 being asked to enquire about the parties' knowledge of 2 technical faults and defects in Horizon during the 3 review; is that right? 4 A. Correct. 5 Q. Albeit you weren't asked, did you yourself make any 6 enquiries into those matters? 7 A. No. 8 Q. Did you not consider the existence of known software 9 problems might be relevant to Pathway's capability to 10 deliver the programme? 11 A. No. We were looking at -- well, as I understood it, the 12 panel was looking at the possibility to reconstruct the 13 programme in a way that would make it more deliverable 14 and remove some of the risks that weren't associated 15 with the technical issues, on the basis that, if the 16 programme could be reconstructed to achieve that, it was 17 then a matter of time for ICL to deliver. 18 Q. But if there were very serious technical issues, would 19 that not be very relevant to whether they would 20 ultimately be able to deliver? 21 A. I guess it's fair to say it could have been relevant. 22 Q. In your statement, you explain that your overall view of 23 the Horizon technology was positive -- 24 A. Yes. 25 Q. -- and that you believe the assessment you made at the 123 1 time was accurate. 2 A. Yes. 3 Q. Would it be fair to say, though, that whether or not 4 your assessment was accurate would depend on whether you 5 were asking the right questions and you were receiving 6 accurate answers to those questions? 7 A. Well, inevitably, but the work that we did was guided by 8 the panel and we were not asked to investigate any 9 technical issues at that time. 10 Q. If you weren't making enquiry into the parties' 11 knowledge of faults and defects, how could you or the 12 panel make an accurate assessment of Pathway's ability 13 to deliver this solution? 14 A. I don't think -- that's a difficult question to answer. 15 Q. The panel reached a number of conclusions which appear 16 to have been informed by your own investigations and 17 findings and I'd like to explore some of those with you 18 now, if I may. Could we pull back up POL00028094. 19 Please could we turn to page 12. I think it might be 20 internal page 12. Thank you. 21 Forgive me, I think maybe it's page 11. It must 22 be written down wrong. Could we go back one page, 23 please. Thank you. 24 We can see at the bottom of this page, heading B, 25 "Solution Design and Fitness for Purpose", and if we 124 1 could go to the following page please, page 11, at 2 paragraph 25 this records that: 3 "The main architectural issues are scalability and 4 robustness. We are advised [the panel said] that 5 a solution of this scale and scope with so many 6 different platforms and products has, as far as PA is 7 aware [PA Consulting, I assume], no precedent. We are 8 satisfied that Pathway's approach to design, development 9 and performance testing is sufficiently rigorous for 10 such a major undertaking." 11 Was that a reflection of the finding that you had 12 made that end of your review? 13 A. I think it's probably more correct to say that was the 14 panel's conclusion. The lead technical person on the 15 panel was an MOD man, Bill Robins and I think he 16 probably did more than we did, in terms of investigating 17 the technical aspects. 18 I think it's probably fair to say and clarify we 19 weren't a member of the panel. We were there being 20 asked to investigate specific aspects, I would imagine, 21 in order to clarify their own view or to concur with 22 their own view. In other words, we weren't given free 23 rein. 24 Q. No, your role was to assist, essentially, and to follow 25 the specific investigations that you were asked to 125 1 follow. 2 A. Yes. Yes, it was quite a different arrangement to the 3 first review. 4 Q. At paragraph 29, so on the same page we can see just 5 over halfway down -- thank you: 6 "The project is probably the biggest of its kind 7 and many of the component parts, although sourced from 8 industry strength products and companies, are being used 9 towards their current limits and scale. Pathway has 10 recognised the risks and has in place the controls we 11 would expect to see in a development project of this 12 scale." 13 Again, was that based on your own findings or, as 14 you recollect, was that a conclusion which the panel 15 reached? 16 A. I think that's a conclusion the panel reached. 17 Q. Turning then to future proofing, which was one of the 18 aspects you were asked specifically to look at, at 19 page 13, please, of POL00028094. At paragraph 33, so 20 the very first paragraph, it records that: 21 "There is good evidence of future proofing at all 22 levels. We have been satisfied that all reasonable 23 steps have been taken to ensure robust sources of supply 24 and compliance with industry standards in designing the 25 architecture. Upgrades to software platforms and 126 1 individual components are provided for, should they be 2 necessary." 3 Bearing in mind this was an aspect that you had 4 looked at, as I understand, was that your finding or was 5 this informed by your findings? 6 A. That was our finding. 7 Q. That was your finding. Insofar as you found that there 8 had been compliance with industry standards in designing 9 the architecture, that was a finding based on your 10 external research of what those industry standards ought 11 to be -- 12 A. Correct. 13 Q. -- and what you had been told by Pathway as to what they 14 were doing? 15 A. Yes, correct. 16 Q. The report of the panel states that you carried out 17 a critical path analysis to establish the risk of 18 further delay to the programme; is that right? 19 A. Yes. 20 Q. One of the factors which you identified as being 21 a likely cause of further delay was the absence of 22 an agreement, any agreement, between the parties 23 concerning the criteria and procedure for acceptance of 24 the system; is that right? 25 A. Correct. 127 1 Q. What did you understand to be the areas of disagreement 2 between the parties at this stage on the subject of 3 acceptance? 4 A. The disagreement was essentially about the specification 5 and criteria for acceptance and, as I recall it, the 6 basis on which acceptance would be given and whether 7 acceptance should be on a model office or a live trial 8 end-to-end under live conditions. That's essentially 9 what I recall. 10 Q. To what extent had the earlier concerns that you had 11 about the resourcing of the programme in the summer and 12 autumn of '97 been addressed by the time of this review 13 in July 1998? 14 A. I think all parties had resourced up and my recollection 15 is that there was still a concern that POCL were not 16 ready to accept a system of this complexity, and that's 17 readiness in terms of preparing the network to live in 18 a very highly structured environment, as opposed to 19 a very unstructured environment at the branch using 20 paper; a question about whether they were ready to 21 receive a system in terms of the help facilities and 22 helpdesk facilities that were outside of the technical 23 helpdesk. 24 Q. Did you have ongoing concerns about the competence or 25 the expertise of the staff managing the post office 128 1 counters aspect of the project? 2 A. I think there were concerns in terms of the number of 3 people involved and their technical competencies and 4 their understanding of business process transformation 5 that would be necessary to accept the system. 6 Q. We know from the report that we've seen produced by the 7 expert panel that one of the proposals which they made 8 was the appointment of a neutral trouble shooter to 9 facilitate negotiations between the parties over the 10 future of the project? 11 A. Yes. 12 Q. Is that your recollection? 13 A. Yes. 14 Q. The individual appointed to carry out that role was 15 Graham Corbett, Deputy Chairman of the Monopolies and 16 Mergers Commission; is that correct? 17 A. Yes. 18 Q. You explain in your statement that Mr Corbett was tasked 19 with advising ministers -- that is ministers in 20 government -- on whether the framework suggested by the 21 Treasury task-force would provide a commercial basis for 22 continuing and whether the parties could develop 23 a robust implementation plan to complete the project. 24 Is that your recollection? 25 A. Yes. 129 1 Q. Because albeit, as you say, you didn't author the 2 report, you were aware that the expert panel had made 3 a number of recommendations at the conclusion of their 4 report; is that right? 5 A. Yes. 6 Q. That they were not in favour of terminating the project 7 at that stage; is that right? 8 A. That was my understanding. 9 Q. What they proposed had been either a full restructuring 10 or partial restructuring of the programme? 11 A. Yes, yes. 12 Q. So you became involved again in October 1998; is that 13 right? 14 A. Yes. 15 Q. And at this stage, to test this feasibility of that 16 restructuring exercise? 17 A. Feasibility in the sense of the programme itself. 18 Q. You have explained that you were asked to join a working 19 group established by Mr Corbett; is that right? 20 A. Yes. 21 Q. Which was chaired by the director of the Horizon 22 programme office. Do you recall who that was? 23 A. No. 24 Q. Does David Miller sound like the correct -- 25 A. David Miller was chair of? 130 1 Q. Sorry, the director of the Horizon programme? 2 A. Oh, yes of course, yes. 3 Q. The purpose of the working group -- sorry, your task was 4 to liaise with the parties as they answered questions 5 posed by Mr Corbett and to provide an assessment of any 6 risks arising from the reconstruction; is that right? 7 A. Yes. 8 Q. We can see an agenda for one of those workshops at 9 POL00090010, please. This is obviously a fax header 10 sheet from you, Peter Copping to Peter Crahan, who was 11 a senior figure in the Benefits Agency and Mr David 12 Miller at the Horizon programme office and Mr Mike 13 Coombs at ICL Pathway. On the following page, please, 14 your letter confirming that you've made some proposals 15 for the workshop and on the third page, please, we have 16 here a list of a number of issues for resolution. The 17 first of these is E2E and model office testing. That's 18 end-to-end; is that right? 19 A. Correct. 20 Q. And model office testing. You were looking for the 21 Benefits Agency to explain the current concerns about 22 the testing philosophy; is that right? 23 A. Yes. 24 Q. Do you recall what the nature of their concerns were at 25 this stage? 131 1 A. No. 2 Q. You were looking to the Horizon programme office to 3 describe the current process as well as the proposals 4 for management of reporting of progress to sponsors, and 5 it was your function, I believe we can see, PA to 6 identify any remaining disagreements, issues, concerns 7 and sensitivities on that subject. 8 A. Yes. 9 Q. So we also see, under topics for discussion, the service 10 management product set. What does that mean? 11 A. This would have been the construct around service 12 management, who did what, where they would reside and so 13 on. 14 Q. The next section is the service management environment 15 and the interim arrangements in place. Can you describe 16 what those were, please. 17 A. I really don't recollect that level of detail I'm 18 afraid. 19 Q. At D we have multi benefit with soft EVP, that's 20 a reference to the security, the extended verification 21 procedure; is that right? 22 A. Yes. 23 Q. I think you are seeking there for Pathway to describe 24 the plans to realise NR2+. Do you recall, in terms of 25 the future software development, what the plans were at 132 1 this stage in relation to the new release? 2 A. No. 3 Q. It's called NR2+ ... 4 A. I recognise NR2 but I couldn't tell you what was in it. 5 Q. Thank you. Then I think there's one more page beginning 6 E if we could zoom in. Thank you very much. Consistent 7 and complete technical design, "BA to list areas where 8 assurance is needed". What were the BA's concerns at 9 this stage about the consistency and completeness of 10 Pathway's technical design? 11 A. I'm sorry but I do not recollect. 12 Q. We can see then that the final topics were acceptance 13 testing and release authorisation and the Horizon 14 programme office. Then, under item 2, the programme 15 critical path and dependencies. Is this an accurate 16 reflection of the types of issues that you were dealing, 17 in the workshops that you were having -- 18 A. Yes, this is essentially a process that we will have 19 taken the responses through, in order to flesh out areas 20 of disagreement which would then be documented for 21 someone to go away and work on and decide how to take 22 those forward. 23 Q. Your assessment of the programme and project management 24 issues, which were prominent in autumn 1998, are 25 summarised in an annex to Graham Corbett's report; is 133 1 that right? 2 A. Yes. 3 Q. Can we please show POL00028098, could we scroll down to 4 page 32 please. We can see here at the top, "Management 5 Summary -- Key Programme Risks". Is this a table that 6 you produced or is it simply summarising your findings? 7 A. I think it's summarising our findings. 8 Q. So in terms of the risks that you had identified, the 9 first of those we see under the heading "Critical" is 10 the speed of acceptance process. Can you explain, 11 please, the nature of your concern at that stage about 12 the speed of the acceptance process? 13 A. My recollection is that things on -- acceptance process 14 got clogged up over disagreements on what the criteria 15 were and how those criteria should be -- differences 16 should be resolved. I really can't recollect any more 17 than that. 18 Q. In terms of the impact that this was likely to have on 19 the programme, it records that: 20 "[A] Failure to complete acceptance in planned 21 timescales could cause one or more of the parties to 22 resort to legal action and program could stop at end of 23 '98 or before." 24 So was the essence of the concern that unless the 25 acceptance process could be agreed and implemented it 134 1 was likely to lead to litigation? 2 A. Correct. 3 Q. What did you understand Pathway's position to be on 4 acceptance at this stage? 5 A. I think they were seeking to base acceptance on 6 a self-certification process and of course no-one on the 7 sponsor side was particularly happy with that. They 8 also, I recall, were seeking to have acceptance on 9 a model office as opposed to end-to-end acceptance, 10 ie in a live system. 11 Q. I will come back to the point about the model office and 12 end-to-end testing shortly but, before I do, just scroll 13 down please to the page, I think it will be 33, where we 14 see what were described as the minor risks recorded. 15 I wonder if we could zoom in please. Thank you. 16 So point 4, risk number 4, under the heading 17 "Minor" is the "Consistent and complete technical design 18 for key products". The assessed impact of that on the 19 programme is that it's likely to impact mainly on the 20 speed of testing and the acceptance process. 21 A. Yes. 22 Q. Did you not consider that the consistency and 23 completeness of the technical design was relevant to 24 assurance of the quality of the programme? 25 A. I think it amounts to the same thing, doesn't it? 135 1 Q. Well, your focus here is on speed of testing and 2 acceptance? 3 A. If the criteria for acceptance are all agreed and the 4 system is submitted against those criteria and there are 5 no issues, then speed will be fairly quick. I'm not 6 sure I'm following your point. 7 Q. I think my point is this, that you seem here to be 8 attributing the significance of the completeness and 9 consistency of the technical design, its overall 10 significance, to the programme is its likely impact on 11 testing and the acceptance process. What I'm saying is 12 that does it not also have a function in ensuring the 13 quality of the solution that's being put in place? 14 A. Yes. 15 Q. On the fifth point, also a minor risk we have 16 "scalability of Pathway design". Scalability was 17 something that had been regarded as quite a significant 18 issue in your earlier reviews in 1997 and 1998. Why is 19 it here characterised as a minor risk to the programme? 20 A. Well, because, at that stage, I think there was 21 beginning to be a better understanding of how the system 22 would be rolled out and scaled up, where scalability 23 relates to the number of offices connected. So it was 24 seen as less of an issue at this stage of the 25 development. But, of course, there could always be 136 1 issues. If you move from 10,000 to 20,000 offices, 2 there might -- suddenly an issue might arise. 3 Q. Were both of these issues, that is the consistency and 4 completeness of the technical design and the scalability 5 of it, were they both not factors that were likely to 6 affect, ultimately, the robustness of the system? 7 A. Not necessarily but possibly. 8 Q. There's one topic I would like to deal with, please, 9 before we have a short break. This comes back to the 10 question of acceptance and you've mentioned in your 11 evidence that you understood Pathway's position to be 12 that they were looking for acceptance to take place at 13 the end of model office testing, as opposed to a full 14 end-to-end test. Shortly after the negotiations that 15 were being facilitated by Mr Corbett concluded, you 16 wrote to David Miller, the Horizon programme director, 17 in order to set out some private thoughts you had about 18 how the parties might break the through the potential 19 impasse on acceptance; is that right? 20 A. Yes. 21 Q. I wonder please if we could pull up POL0009009. Thank 22 you. 23 So we can see here your letter of 19 October 1998, 24 addressed to Mr David Miller, the director of the 25 Horizon Programme, and, as I've just said, your proposal 137 1 to set out some of your private thoughts on the issue of 2 acceptance. That letter enclosed a short paper entitled 3 "Acceptance testing: a framework for developing a new 4 paradigm". We can see that on page 2, please. 5 You observe in that paper, under the heading 6 "Problem definition" that: 7 "Sponsors and Pathway have agreed to de-risk the 8 programme by decoupling card roll-out from NR2", which 9 I understand is New Release 2; is that correct? 10 A. Yes. 11 Q. "... and to base NRO ..." 12 Is that national roll-out? 13 A. Yes. 14 Q. "... on child benefit and EPOSS only until NR2+ [this is 15 New Release 2 Plus, further functionality] is available 16 when multi-benefit roll-out starts. This new sequence 17 raises a legitimate question whether an alternative 18 acceptance process can be designed that protects the 19 commercial objectives of the parties and which at the 20 same time reflects the status of the revised programme 21 at completion of [Model Office Testing] and at 22 completion of Live Trial. Additionally, the acceptance 23 process for any requirement to be delivered during NRO 24 [National roll-out] would need to be included in any new 25 approach." 138 1 Your paper essentially proposed a new paradigm for 2 acceptance and we can see the essence of that distilled, 3 please, on page 3. At the end of the second paragraph 4 you say: 5 "Simply put [following your new paradigm], 6 sponsors would give up termination rights on acceptance 7 following [Model Office Testing] in exchange for the 8 option of being able to have more punitive SLAs ..." 9 Is that service level agreements? 10 A. Correct. 11 Q. "... following the start of [National roll-out], should 12 the system fail to meet acceptance criteria in Live 13 Trial. Similar arrangements could be put in place for 14 future releases of functionality/services." 15 Would it be fair to say that, boiled down to its 16 core, your proposal envisaged the sponsors forfeiting 17 their right to reject the system, even if it failed to 18 meet the criteria, which the sponsors deemed to be 19 necessary for acceptance? 20 A. Yes, and perhaps I should put this in context. The 21 clues in this letter are it was a private thoughts 22 letter and, in the last paragraph, "Next steps", 23 bluntly, this an unsolicited proposal for more work from 24 PA, and it was rejected. 25 Q. It was indeed. Viewed from the perspective of the 139 1 sponsors, this is an approach which would have been 2 fraught with risk, would it not? 3 A. Possibly. 4 Q. By their very nature, model office tests tended to be 5 carried out under optimal circumstances; that's right, 6 isn't it? 7 A. Yes. 8 Q. Because they didn't accurately replicate the real-life 9 environment in which the system would actually operate, 10 these tests were very unlikely to identify the full 11 breadth of usability and performance issues -- 12 A. That's correct. 13 Q. -- which would only become apparent, ultimately, in live 14 operation of the system -- 15 A. Yes. 16 Q. -- by which point, the termination rights would have 17 been lost? 18 A. Yes. 19 Q. So there was a risk in adopting this approach that the 20 sponsors might find themselves bound to accept and roll 21 out a system that later didn't prove to be fit for the 22 purpose for which it was -- 23 A. Which is why it was rejected. 24 Q. Bearing in mind those risks inherent in the approach, 25 why did you consider this to be a suitable paradigm for 140 1 acceptance? 2 A. We were trying to be creative to find a way through the 3 acceptance block. 4 Q. Was this reflective of the pragmatism which you felt was 5 earlier lacking in the programme and which had 6 contributed to significant delays? 7 A. I don't think we saw it in the broadest light. We saw 8 that as a possible opening of a discussion that could 9 help solve the problem. It was a pragmatic approach. 10 MS HODGE: Thank you. 11 Sir, that brings me to the end of that topic. 12 I wonder if now would be a convenient time to take 13 a short break? 14 SIR WYN WILLIAMS: Yes. 15 MS HODGE: We're making good progress. 16 SIR WYN WILLIAMS: Good. So what time shall we start again 17 Ms Hodge? 18 MS HODGE: Shall we resume at 3.10? 19 SIR WYN WILLIAMS: Okay, fine. See you then. 20 (2.55 pm) 21 (A short break) 22 (3.09 pm) 23 MS HODGE: Good afternoon, sir. Can you hear, and see me? 24 SIR WYN WILLIAMS: Yes. 25 MS HODGE: Thank you. We can hear you. 141 1 Mr Copping, picking up where we left off, which 2 was the new paradigm for acceptance which you had 3 proposed (which, as you say, did not find favour with 4 sponsors), you've explained in your statement that you 5 continue to have some involvement in the public sector 6 negotiations over the future of Horizon in the early 7 part of 1999; is that right? 8 A. Yes. 9 Q. But the last significant engagement which you had 10 related to the acceptance of the system; is that 11 correct? 12 A. Yes. 13 Q. Contrary to the proposal which you had made which 14 envisaged acceptance at the end of model office testing, 15 we know that what Post Office Counters and Pathway 16 agreed upon cancellation of the Benefits Payment Card 17 was that an operational live trial would take place; is 18 that right? 19 A. Yes. 20 Q. By that stage, however, the thresholds for acceptance 21 had changed. Were you aware of that? 22 A. Yes. 23 Q. Could we please show POL00028208. Thank you. This is 24 a copy of "Schedule A11" to the codified agreement, 25 dated 28 July 1999, concluded between ICL Pathway and 142 1 Post Office Counters. I suspect you won't have seen 2 a copy of this contract at the time. 3 A. Correct. 4 Q. Have you read this document since? 5 A. Yes. 6 Q. We can see, if we turn to the second page, please, at 7 paragraph 2.2 -- so about a third of the way down the 8 page -- a reference to the thresholds for acceptance of 9 the CSR. Are you aware of what the CSR was or what it 10 signifies? 11 A. I don't recall what CSR stands for, I'm afraid. 12 Q. I believe it's the Core Systems Release -- 13 A. Right. 14 Q. -- the name given to the package of software tested 15 during the operational trial and ultimately rolled out. 16 It comprised the EPOSS, Electronic Point of Sales 17 Service, the order book control service, which was still 18 in operation, and the Automated Payment Service. Does 19 that sound broadly correct, in terms of what you 20 understood -- 21 A. That sounds familiar, yes. 22 Q. What this provision provides, we can see it's framed in 23 the negative. It effectively says: 24 "The thresholds will not be met if in respect of 25 CSR Acceptance there are: 143 1 "[First condition] one or more high severity 2 deficiencies as categorised in paragraph 7.1(a) of this 3 Schedule ('category (a) faults'); 4 "[Alternatively] more than 20 category (b) faults; 5 "[Finally] more than 10 category (b) faults in 6 respect of any one CSR Acceptance Specification." 7 Is that broadly consistent with what you 8 understood at the time to be the overarching criteria 9 that the system had to meet in order to be accepted? 10 A. Yes, that's correct. 11 Q. So if there was one or more high severity deficiency, it 12 wouldn't be eligible for acceptance? 13 A. Correct. 14 Q. Likewise, if there were more than 20 of a medium 15 severity, it wouldn't be eligible? 16 A. Correct. 17 Q. We can see on page 3 at point 5, under the heading 18 "Appointment of Expert", that contract made provision 19 for you to be appointed as an expert to assist in 20 resolving any disputes relating to CSR acceptance. Is 21 that how you understood your role at the time? 22 A. I didn't know about this at the time, so I didn't 23 understand I was being proposed. 24 Q. When you did become involved, how would you characterise 25 your role? 144 1 A. It was explained to me that my role would be essentially 2 to facilitate the parties to come to an agreement. 3 There was a lot of allocation of blame from one party to 4 another about what actually was going on on acceptance 5 and it was explained to me that my role was to ensure 6 the parties worked together to resolve conflict and, 7 through that process, reach an agreement on the level of 8 severity of each incident and a resolution plan. 9 I subsequently found out that, apparently, I had 10 the opportunity to arbitrate but I don't think, to the 11 best of my recollection, that was ever exercised by 12 either party. In other words, I had the option to tell 13 them the way it was going to be on particular incidents. 14 Q. Do you recall being consulted by either of the parties 15 about the arrangements that were put in place in the 16 contract? 17 A. In this contract? No. 18 Q. It appears, from your statement and from the records 19 we've obtained, that you were first called upon to 20 provide assistance to the parties on completion of the 21 operational live trial. Is that consistent with your 22 recollection? 23 A. That resonates. 24 Q. In preparation for a meeting which appears to have taken 25 place on 16 August 1999, you were sent what was 145 1 described as a "hot list" of Acceptance Incidents; is 2 that right? 3 A. Yes. 4 Q. Please could we pull up POL00028355. This is an email 5 from Tony Houghton, dated 13 August 1999. We can see 6 you named as one of many recipients there, the fourth in 7 the list. Can we turn to the following page, please. 8 Here we have the "Acceptance Incident Hotlist". Could 9 you please explain what you understood the significance 10 of this list to be? 11 A. This was a list in which I identified all the incidents 12 that were outstanding, together with those where there 13 was a disagreement on severity and I think, from memory, 14 there were three, possibly four, where POCL and 15 ICL Pathway were in disagreement. 16 Q. So I think, looking at the list, it appears as though 17 there was disagreement in terms of severity on almost 18 every single -- 19 A. Sorry, I was looking through the lens of the medium to 20 high. 21 Q. Forgive me. There were three incidents categorised by 22 POCL as high. We can see the first of those in the list 23 being Acceptance Incident number 376, which is described 24 as the "Derived cash account not equal to the electronic 25 cash account". 146 1 The next high severity is in relation to training, 2 number 218. It's described as the "Training course Cash 3 Account module inadequate". POCL have assessed that as 4 high, whereas Pathway were treating that as closed at 5 this stage. 6 Then, thirdly, in the POCL infrastructure 7 Acceptance Incident number 298, described as "Counter 8 system subject to lockups & screen freezes requiring 9 reboots", assessed by Pathway as a low severity incident 10 but by Post Office Counters as high. 11 I think there's one in the category of medium to 12 high, which was number 369, also in the POCL 13 infrastructure, "Scanner reliability in relation to 14 [order book control service] transactions". 15 So that was the state of play at the end of the 16 operational live trial, so far as you were aware; is 17 that right? 18 A. Yes. 19 Q. Do you have any recollection of the meeting that took 20 place on 16 August 1999? 21 A. I might need reminding. 22 Q. I don't think we have any record, written record, of 23 that meeting but we know that the disputes between ICL 24 Pathway and Post Office Counters over the status and 25 severity of these incidents were not resolved at that 147 1 meeting. That much is clear from a supplemental 2 agreement reached between the parties on 20 August 1999. 3 I wonder if we could please pull up FUJ00000485. 4 So at the top we can see this described as 5 a "supplemental agreement" dated 20 August 1999 between 6 Post Office Counters and ICL Pathway. If we could 7 please scroll to page 3 -- thank you -- under the 8 heading "It is Agreed as follows", we can see in 9 relation to "CSR Acceptance", paragraph 1.1: 10 "The parties agree that CSR Acceptance was not 11 achieved as at the end of the CSR Operational Trial 12 Review Period." 13 If we go on, please, to page 4, there's proposed 14 here a remedy for the outstanding faults in the system. 15 It provides: 16 "With a view to facilitating the obtaining of CSR 17 Acceptance in the Second CSR Acceptance Test, the 18 parties agree as follows ..." 19 Paragraph 2.1: 20 "In the period between the date of this Agreement 21 and 17th September 1999 (the 'Limited Trial Period'), 22 the parties will set up and conduct a programme of joint 23 workshops for the purposes of agreeing (to the extent 24 not already agreed): 25 "resolution plans for the Agreed Category B 148 1 Faults, the Disputed Category A Faults, the Disputed 2 Category B Faults ... and (if appropriate) the Unagreed 3 Fault ..." 4 It goes on to say: 5 "a single timetable for resolution of outstanding 6 category (b) faults", would form part of those 7 workshops. 8 I think you recall participating in those 9 workshops; is that right? 10 A. Yes. 11 Q. We can see they were chaired by Keith Baines, the late 12 Keith Baines, of Post Office Counters Limited and Tony 13 Oppenheim of ICL Pathway? 14 A. Yes. 15 Q. On page 5 you're referenced again. On this occasion, 16 were you consulted about your involvement? I assume you 17 were on the 16th. 18 A. Not that I recall. 19 Q. So here at 2.4: 20 "The parties will involve Peter Copping as Expert 21 in the activities referred to in this paragraph 2 [to 22 which we've just referred]. There may be occasions on 23 which the Expert is asked to determine an issue as 24 between parties in accordance with the provisions of 25 Schedule A11 ... but otherwise ... his role will be as 149 1 facilitator and adviser to the parties in their efforts 2 to achieve successful resolution of outstanding issues." 3 That seems to tally with what you have described 4 earlier as your understanding, that is to say the latter 5 part, facilitating agreement. 6 A. I hadn't seen this document at the time. 7 Q. But it reflects, as far as you're -- 8 A. It reflects my understanding. 9 Q. You explained in your statement that, prior to each of 10 the joint workshops to which this supplementary 11 agreement makes reference, you received briefings from 12 each of the parties relating to the status of the 13 acceptance incidents; is that correct? 14 A. Yes. That's something I asked for, if I remember. 15 Q. Did the briefings you received include ICL Pathway's 16 proposed plans for resolving the outstanding acceptance 17 incidents? 18 A. I'm not sure whether all the proposed plans were 19 included. 20 Q. We'll take a look at some of them shortly. In the hot 21 list that we reviewed a short time ago we looked at 22 three acceptance incidents that were graded high by Post 23 Office Counters. I would like to explore with you now 24 what you recall about those. 25 If I could begin with Acceptance Incident 218, 150 1 what did you understand this particular incident to 2 entail? 3 A. This concerned a post office view that the training for 4 Horizon users was inadequate and I think, in response, 5 Pathway offered a number of initiatives, which involved 6 half day training for Post Office Counters staff in IT 7 usage, computing, and so on. 8 The underlying issue here, I think, was to do with 9 the situation that I mentioned earlier, that Pathway 10 hadn't, in completeness, considered the business 11 transformation that would be necessary to accept the 12 system and, with that, all the process changes that 13 would need to take place at the counter, and my 14 recollection is that there were extremes of counter 15 configuration from quite a number of counters to 16 a simple remote terminal in an outlying area, and users, 17 if I recall correctly, between 20 and 75 or 85 years 18 old. So there's a tremendous spread of capability that 19 needed to be trained. 20 In my understanding, that was why the Post Office 21 decided this should be categorised as high severity 22 impact. 23 Q. Do you recall whether the concerns about training were 24 focused on any particular aspect of the training 25 programme being offered? 151 1 A. My recollection is that the there were a lot of issues 2 around closing of accounts and it was unclear, at that 3 stage, whether that was a training issue or a systems 4 issue. 5 Q. So far as you're aware, this was an incident that was 6 resolved to the satisfaction of the Post Office during 7 your workshops; is that right? 8 A. Yes. 9 Q. I'd like to turn to another incident, this was AI298, 10 also categorised by the Post Office as being of high 11 severity. Could you describe your understanding of that 12 incident, please? 13 A. This was about instability in the system, the symptom 14 being lockouts at the terminal, crashes in the middle of 15 a process, system busy incidents, and so on, and I think 16 Pathway took the view that this was pretty normal for 17 IT: PCs crash, PCs lock up. The Post Office or POCL 18 considered this to have a high impact on the business 19 simply because, while the system was down, customers 20 couldn't be dealt with and, therefore, it had a high 21 impact and so there was a disagreement about the 22 severity. I think POCL classified it as high severity; 23 ICL Pathway, I think, as medium. 24 Q. Do you recall how this particular incident was resolved? 25 A. My recollection is that it wasn't resolved through the 152 1 series of seven workshops. It was escalated to the 2 management resolution meeting towards the end of 3 probably August/early September. 4 Q. Thank you. We'll return to that final workshop -- that 5 final meeting, sorry, a little later. Before we do, I'd 6 like to address with you the third incident categorised 7 by Post Office as high severity, and that is incident 8 AI376. 9 What was the nature of the problem that had been 10 identified, so far as you were -- 11 A. As I understood this, it was about intermittent failures 12 in the reconciliation process between the money in the 13 till and what the system had recorded and, at the time, 14 I think it was believed the root cause was about -- or 15 lack of integrity in transfers between Post Office and 16 ICL Pathway systems. 17 Q. Do you recall where in the system this problem had 18 arisen? 19 A. At the TIP interface, as I understood it. 20 Q. Can you explain what you understood TIP to be? 21 A. It's the interface between the Post Office system that 22 is responsible for counting and the ICL Pathway 23 interface and the database that recorded the 24 transactions in the system. 25 Q. You have described the root cause as being a lack of 153 1 integrity in the information passing. 2 A. That was my understanding at the time. 3 Q. Before we go to one of the documents I'd like to show 4 you, do you recall how the problem had come to light, 5 how it had been detected in the system? 6 A. Not specifically, I'm afraid, no. 7 Q. Please could we show POL00028332. This is another email 8 of 13 August 1999 from Andrew Simpkins, addressed to you 9 and to David Rees. Was that a colleague at 10 PA Consulting? 11 A. Correct. 12 Q. It says: 13 "Peter/David 14 "Following the Management Resolution meeting 15 yesterday I attach as agreed by [Post Office Counters] 16 and Pathway the minutes of this meeting and a summary of 17 the incidents that are in dispute. 18 "The minutes will give you an up-to-date position 19 on the high priority incidents in particular. 20 "We propose that the meeting with yourselves does 21 not start until 12.00. It will be in Gavrelle House 22 room 7." 23 If we could please turn to the minutes on the 24 following page, so of course not a meeting that you 25 yourself attended but minutes that were shown to you to 154 1 bring you up to speed for the meeting, to which we've 2 referred, on the 16th. 3 A. Yes. 4 Q. On page 3, please, we can see, just over halfway down 5 the page, at point 3, the heading "Review of High 6 Priority Incidents", the first of these being Acceptance 7 Incident 376. "JD", who I believe was John Dicks, 8 an employee of ICL Pathway, reported that: 9 "Pathway recognise that not all transactions had 10 been harvested and sent to TIP. A provisional fix went 11 in on 2nd August and this has worked satisfactorily so 12 far with the effect that all records had been sent. 13 A root cause analysis has been developed, identifying 14 8 contributory problems, and all but one has been 15 diagnosed and tested in Pathway to date. Pathway cannot 16 guarantee however that all problems have been trapped. 17 They will need to see evidence from the fix of the 18 8 known problems, and will continue to monitor the 19 problems for 8 [sic] months to be confident of its 20 resolution. 21 "The provisional fix and the control procedures 22 developed allow Pathway to identify any errors, to patch 23 the file, and to notify TIP in advance. Since 24 implementation there have been no errors to report and 25 hence Pathway contend that this action taken to date and 155 1 the result they have observed justify the downgrading of 2 this incident." 3 We know, of course, it wasn't ultimately agreed 4 that that incident be downgraded -- 5 A. Correct. 6 Q. -- hence why the workshops took place. So far as Post 7 Office Counters were concerned, what did you understand 8 their assessment of the business impact of this incident 9 to be? 10 A. An inability on a consistent basis to reconcile Horizon 11 data with cash data. 12 Q. Would it be fair to say it was an issue of fundamental 13 concern to Post Office Counters? 14 A. I would have said so, yes. 15 Q. The principal purpose of Horizon being to perform 16 an accounting function -- 17 A. Absolutely, yes. 18 Q. -- which would enable Post Office Counters to reconcile 19 the transaction performed by its agents at the branch 20 counter with its own records of cash and stock held, as 21 well as the transactions performed on behalf of its 22 clients. 23 A. Correct. 24 Q. If the system wasn't producing accurate cash accounts, 25 which appears to be what this incident was showing, did 156 1 this not call into question its very fitness for 2 purpose? 3 A. Yes. 4 Q. Do you recall how ICL Pathway proposed to rectify this 5 problem? 6 A. In broad terms, yes. There were a number of proposals 7 in the resolution plan. I think the most worrying was 8 that one of the fixes wouldn't be implemented until the 9 year 2000, either at the end of '99 or 2000, and it was 10 that that caused the incident to be escalated to the 11 management meeting at the end of the seven workshops, 12 not just because the final fix wouldn't be available 13 until the end of the year but also because regression 14 testing would therefore take place afterwards and, if 15 there were further issues to be found, that would not be 16 something that would rest easily with roll-out. 17 Q. If we could take a closer look at what was discussed at 18 these workshops, you attended one on 26 August 1999; is 19 that correct? 20 A. If my name's on the agenda, yes. 21 Q. Could we show POL00028472, please. This is an email 22 from Altea Walker to Graeme Seedall and others. You are 23 not a recipient of the email but if we could please turn 24 to page 6, we can see here a minute of the Acceptance 25 Workshop -- this is number 2 -- held on 26 August 1999 157 1 and under the attendees you're listed first, as the 2 expert, Peter Copping, and the first item that appears 3 to have been discussed is Acceptance Incident 376, the 4 issue of data integrity. 5 At point 1, it records that: 6 "[Post Office Counters] needs to be confident of 7 the root cause analysis and fixes, both applied and 8 planned to be applied." 9 The proposal was that a working group of Post 10 Office Counters, comprising a number of employees there 11 and Pathway, in brackets John Pope, an employee of 12 ICL Pathway, were to review the TIP incident status 13 report and report back progress and issues to the 14 workshop. 15 Under point 3, we can see Pathway proposed to 16 introduce a fix to ensure that the cash account does not 17 lose transactions and there's a reference to a PinICL. 18 Were you aware what a PinICL was? 19 As I understand it, it was a record of 20 an incident, the method by which Pathway recorded 21 incidents in the system. As part of 1 above, that is to 22 say the root cause analysis and fixes, a review was to 23 be taken to fix and confirm acceptability of the fix to 24 this group. 25 If we could turn the page, please, we can see at 158 1 point 5 there's a reference to Pathway proposing 2 a three-level data integrity check to be implemented in 3 December: 4 "This needs to be documented as a high level 5 design including failure state analysis." 6 We see there Post Office Counters Limited, 7 a number of employees identified in brackets, are to be 8 involved in interactive walk-throughs during the 9 development of the design to report progress and issues 10 to the group. 11 At point 6 it records that: 12 "[Post Office Counters'] position is that roll-out 13 should not commence until data integrity can be assured. 14 Ruth Holleran [an employee of Post Office Counters] to 15 consider with the Auditors, and report back to this 16 group, whether the current Pathway checks plus, 17 possibly, continuing [Post Office] checks, would be 18 adequate until Pathway's full data integrity checks are 19 in place." 20 Finally, at point 8, we see a reference to Pathway 21 preparing a rectification plan that will be presented to 22 the group. So this appears to be the state of play as 23 at 26 August 1999? 24 A. Yes. 25 Q. I understand you don't have a detailed recollection of 159 1 these events but does that broadly tally with what you 2 understood the position to be? 3 A. Yes. 4 Q. Point 8, to which we have just referred, mentions 5 a rectification plan that was produced in response to 6 Acceptance Incident number 376. I believe you were 7 shown a copy of that plan. 8 A. Yes. 9 Q. If we could bring that up, please, POL00028466. This 10 document is dated 4 September 1999, we see that at the 11 top, Version 0.3. Under the title it's the "Acceptance 12 Proposal for Acceptance Incident 376", and the abstract 13 records: 14 "This document contains ICL Pathway's proposal to 15 the independent Expert in respect of Acceptance 16 Incident 376." 17 Under distribution, you're named as the first to 18 receive it, "Expert: Peter Copping". Do you recall 19 seeing this document at the time? 20 A. I do. I don't know why they addressed it to me 21 specifically. 22 Q. If we could turn please to page 5, we can see here 23 summarised Pathway's position in relation to this 24 incident. At paragraph 5.1, they set out the 25 background: 160 1 "During the Live Trial and since, incidents have 2 occurred that, in [Post Office Counters'] view, 3 constitute a potential threat to the integrity of their 4 accounts. These can be categorised into three groups. 5 "1. Some outlet transactions were not sent to 6 TIP: 7 "because the harvester deliberately omitted 8 incomplete records, caused principally by missing modes, 9 and 10 "because, on one occasion, harvesting started 11 before replication between recovering correspondence 12 server nodes was complete." 13 The second principal cause was that: 14 "Not all transactions were [completed] in the 15 outlet cash account because of end-dating of Item 16 Reference Data." 17 Thirdly that: 18 "Some Cash Account records were sent to TIP 19 because the pointer used by the harvester was not 20 available: 21 "because a counter was rebooted before it could 22 write it; and 23 "on one occasion, because a second balance process 24 was allowed to run." 25 Pathway suggested: 161 1 "Important advances have been made since the above 2 incidents occurred, [which are] discussed below under 3 the same numbers ..." 4 We see here, I think, a list of fixes that have 5 been applied to address the causes identified at 1, 2, 3 6 above, the first of these being that: 7 "All instances of messages written without 8 harvester-sensitive fields have been fixed, except one 9 that will be fixed shortly. Accounting integrity has 10 been safeguarded by establishing routine examination of 11 Event Logs to detect and report daily to TIP any 12 harvester exceptions. 13 "The harvester has been enhanced to positively 14 check that the full message set for an outlet is present 15 on the correspondence server before initiating 16 harvesting for that outlet." 17 Secondly: 18 "The system is being modified so that the 19 balancing and Cash Account processes can continue ... if 20 an item is end-dated during a period for which there are 21 transactions." 22 Thirdly: 23 "The system has been made robust against 24 inopportune reboots by writing persistent objects to the 25 message store, enabling controlled restart of the office 162 1 balance process after power failure, etc. 2 "A change has been made to ensure that multiple 3 balance processes cannot run concurrently. In addition 4 a message will be displayed to inform the user that the 5 balance process has initiated." 6 So here we see Pathway essentially presenting 7 a picture of three principal causes for the end balances 8 having been identified. Would that be a fair 9 characterisation? 10 A. Yes. 11 Q. When I say "root causes", three overarching root causes 12 and the fixes that they have applied? 13 A. Or planned to apply. 14 Q. Or planned to apply to address them. Do you recall how 15 widespread these cash account discrepancies were at this 16 time. 17 A. In terms of quantity? No. 18 Q. Could we please turn to page 8 of this document. It 19 appears that this table was appended to ICL Pathway 20 acceptance's proposal. Do you recall being shown a copy 21 of it at the time? 22 A. I've certainly seen that before, yes. 23 Q. Could we zoom in a little bit, please. Thank you. 24 So we can see at the top it is entitled "Incident 25 analysis". At the very bottom it confirms that these 163 1 are the figures recorded as at 5 pm on Friday, 2 3 September, presumably 1999, bearing in mind the date 3 of the document. At the very top we see "Number of 4 outlets affected by cash account week". 5 The top row appears to record the cash account 6 weeks numbered 8 to 27, and the left-hand side column, 7 the root causes. Do you agree with that broad analysis 8 of the table? 9 A. Yes. 10 Q. So I think what we can see here is that there are two 11 root causes, number 9 and 10 -- so missing mode scales, 12 replication recovery -- that in the weeks 16 to 19 have 13 caused quite a substantial number of outlets to be 14 affected. In relation to 9, there are 22; in relation 15 to 10 there are 37. Following the application of a fix, 16 what this appears to record is that no further outlets 17 have been affected. 18 But, in total over that period, we see 80 outlets 19 affected by one of the 12 root causes of this problem; 20 is that a fair picture? 21 A. Yes. 22 Q. It's not necessarily the case that this equates to 23 80 different outlets, I think, because one possible 24 interpretation of the table is that a single outlet was 25 affected in more than one week but this is quite a high 164 1 figure, is it not? 2 A. Absolutely, yes. 3 Q. Particularly when we bear in mind the relatively small 4 number of outlets that were, in fact, operating the 5 system at the time; is that right? 6 A. Yes. 7 Q. Do you recall how many branches were operating Horizon 8 at this stage? 9 A. At this stage, would it be about 200? 10 Q. I think it was approximately 300, in the region of 299. 11 When viewed in that context, 80 branches being affected 12 by cash account discrepancies is very significant 13 indeed. 14 A. Absolutely. 15 Q. We can also see from this table -- if we can zoom back 16 in, please -- in the penultimate row it starts with 17 a question mark "under investigation", which appears to 18 suggest that there were 36 branches affected by cash 19 account discrepancies for which no root cause had been 20 identified. 21 A. Correct. 22 Q. So, at this stage in early September 1999, would it be 23 fair to say cash account balances remained a very 24 serious problem? 25 A. Yes. 165 1 Q. In addition to the fixes that Pathway had identified in 2 their resolution plan, they had also proposed 3 introducing what was called a three-level data integrity 4 check. We can see reference to that at paragraph 5.2 on 5 page 6 of this document, please. At 5.2, the heading 6 "Maturity of plan", says: 7 "The Pathway proposal in this area has now been 8 expanded into the High Level Design document Logical 9 Design for EPOSS/TIP Reconciliation Controls." 10 It goes on to review that was a document being 11 reviewed by the working group in detail. Do you recall 12 seeing a copy of that document? 13 A. I think I've seen it somewhere. It might have been just 14 recently, it might have been quite a long time ago. 15 Q. Before we go to it, can you explain what you understood 16 this three level data integrity check to entail? 17 A. I thought you might ask me that and the answer is no. 18 I've seen a very complicated description from ICL 19 Pathway in a letter written by Tony Oppenheim, I think, 20 in the pack. 21 Q. If we could turn up, please, POL00090428. This contains 22 a copy of the second supplemental agreement concluded 23 between ICL Pathway and Post Office Counters on 24 24 September 1999, so at or around the time that your 25 involvement came to an end. I think that's right. If 166 1 we could turn, please, to page 135 this is a copy of the 2 "Logical design for EPOSS and TIP Reconciliation 3 Controls". It's dated 20 September 1999, Version 0.7. 4 It's quite a lengthy document, I don't intend to take 5 you through all of its detail but if we could turn, 6 please, to page 6 -- I apologise, it was 135. It's 7 internal page 6. Forgive me, so that should be 141. 8 Under the heading 3, "Overview", there's quite 9 a helpful summary of what the process was intended to 10 entail. It records: 11 "The reconciliation processes will be split into 12 two separate sets of activity. Daily reconciliation 13 tasks and Weekly (or more accurately at the end of each 14 [Cash Accounting Period] CAP) reconciliation tasks. 15 "The daily tasks will ensure that the base 16 transaction data recorded at the counter matches the 17 base transaction data transferred to TIP for that day. 18 At the same time, the transactions will be used to 19 generate total control totals for the Cash Account 20 tables to which the transactions will report at the end 21 of the [Cash Account Period]. 22 "At the end of the [Cash Account Period], the 23 daily control totals generated for each Cash Account 24 table will be accumulated and the resulting value 25 calculated for the Payments and Receipts table will be 167 1 compared with the Cash Account line records generated by 2 the Cash Account production process. If there is 3 a discrepancy in this comparison, then the system will 4 validate each of the accumulated daily control totals 5 with the corresponding Cash Account line records to 6 identify the table which does not reconcile and record 7 an error message in the Riposte message store. 8 "The existing functions in the system which create 9 the outlet stock holding records and the Cash Account 10 Line records will also be amended to accumulate 11 a control total for each set of records which will be 12 written into the message store at the end of each set. 13 These control total records will be harvested and 14 inserted into the TPS Host database. The TPS Host 15 system will compare the Stock Holding records ... and 16 the Cash Account Line records ... output to the TIP Cash 17 Account subfile with the control totals received from 18 the OPS system. In the event that the TPS harvester 19 fails to locate either the Stock Holding ... records or 20 the Cash Account Line records ... or the control records 21 calculated by the TPS Host system differ from the 22 control totals received from the OPS, then 23 a reconciliation error report will be produced." 24 Does that assist you at all in relation to how 25 this piece of software was intended to function? 168 1 A. It resonates and of course the big question is what 2 happens with the error report data? 3 Q. If I've understood it correctly -- and that is a big 4 if -- these new reconciliation controls, I think, were 5 intended to automate a task that Post Office Counters 6 had been performing temporarily in TIP -- is that 7 right -- in that they had been seeking to verify whether 8 the base transaction data recorded at the counter was 9 consistent with the transaction data being transferred 10 to TIP? 11 A. Yes, I think the aim was to have a completely automatic 12 reconciliation process, which corrected, as a result of 13 the process the checking process. Whether that happened 14 in practice, I can't say. 15 Q. I think, as you have said, it was a system which was 16 designed to generate a report to verify that the error 17 had taken place -- 18 A. Mmm. 19 Q. -- forgive me, to verify that in imbalance had been 20 detected but what the reconciliation control didn't do, 21 did it, was identify the root cause of the discrepancy 22 in the first place? 23 A. I think that's right. 24 Q. As you have explained, you attended a series of 25 workshops, the last of which, I believe, took place on 169 1 17 September 1999; is that right? 2 A. I think so. 3 Q. Do you recall what progress had been made in relation to 4 resolving IA376 by that stage? 5 A. It was still, as I recall it, categorised as high by 6 POCL and medium by ICL. 7 Q. Please could FUJ00079716 be shown on the screen? 8 We can see your name's recorded under attendees at 9 this meeting on Friday, 17 September 1999. This was the 10 last of the seven workshops that had been arranged. 11 A. Correct. 12 Q. If we could turn to page 6, please, so at the bottom is 13 the heading "AI376 Data Integrity", the substance of 14 which we can see on the following page. If we could 15 scroll down, please. So here is, essentially, an update 16 on where things are in relation to the data integrity 17 checks. It records, as we've seen before, that Pathway 18 were proposing a three-level data integrity check to be 19 implemented in December, with the relevant design 20 documentation to be considered. 21 At the fourth workshop, the update was that Post 22 Office Counters had considered the high level design to 23 be generally good but wanted further checks to be 24 undertaken in relation to failure scenarios and 25 operating procedures. 170 1 We can see then under workshop number 5 that Post 2 Office Counters had reviewed identified failure 3 scenarios and some issues with the high level design, 4 which needed clarification. It was proposed that 5 a meeting would take place to discuss Post Office 6 Counters' paper, the EPOSS/TIP reconciliation controls, 7 summarising the failure scenarios and the design issues. 8 A further update at the sixth workshop was that 9 progress was good and that Pathway were to issue a paper 10 for Post Office Counters to review. 11 The final update is at the bottom there for 12 workshop 7. Pathway had issued the high level design -- 13 I assume HLD means "high level design" -- paper for Post 14 Office Counters' review and Post Office Counters to 15 provide their written comments to John Pope. 16 So that's essentially where we were with the high 17 level design. Could we go to the next page, please. 18 In relation to whether or not to accept and 19 roll-out out the system, this confirms Pathway's 20 position, as stated previously, was that roll-out should 21 not commence until data integrity should be assured. At 22 workshop number 4, the update was to the effect that 23 Post Office Counters and Pathway needed to develop 24 a contingent approach, possibly including indemnities, 25 suggesting that Keith Baines and Tony Oppenheim would 171 1 meet with the lawyers to initiate that process. 2 At workshop number 5, Post Office Counters' 3 position remained that the incident should be classified 4 high until the data integrity fix was in place. Further 5 internal meetings were proposed to further confirm the 6 position. 7 Workshop number 6, "This will now be part of the 8 contractual discussions being held" between Post Office 9 Counters and Pathway. 10 At workshop number 7, the one that was held on 11 17 September, it said: 12 "This issue is now focused on the success criteria 13 for [national roll-out] resumption." 14 It confirms at a review in November -- which may 15 be an error because, of course, these minutes are dated 16 September -- Pathway had previously proposed four weeks' 17 operation with a less than 1.5 error rate. Keith Baines 18 and Ruth Holleran proposed an error rate of 19 0.6 per cent, the current average being 1.2 per cent, 20 together with six other conditions, five of which were 21 listed in a paper that "RH" -- presumably Ruth 22 Holleran -- had produced, and the sixth being a further 23 two-week period of live running of the permanent Cash 24 Account fix, prior to the actual recommencement of 25 national roll-out in January. 172 1 The penultimate paragraph records Tony Oppenheim 2 responding as follows: a 0.6 error rate agreed subject 3 to this being measured as the average of six weeks from 4 4 October to mid November, with a maximum of ten working 5 days to analyse each TIP fault, comprising a root cause 6 analysis, diagnosis and agreed resolution and that was 7 agreed, except for faults requiring diagnostics. 8 A further two-week period agreed, subject to the 9 agreement of logistics of the plan. On the basis of the 10 current plan, this condition would lead to a two week 11 delay in the planned date for recommencement of national 12 roll-out and this was agreed: 13 "Re the error rate criterion, the Cash Account 14 does not reconcile and is attributable to an error in 15 the POCL domain. The error rate is to be calculated as 16 the ratio if the number of incidents and the total 17 number of cash accounts during the six weeks period." 18 So this is where we were on 17 September. What do 19 you understand these discussions to relate to? 20 A. Two things. There was still a problem with 376 and 21 there was some negotiation beginning to start about what 22 the acceptance criteria might be for that particular 23 incident. 24 Q. You use the term "acceptance criteria". Would it be 25 right to refer to it maybe as "conditions", the 173 1 conditions upon which the system might be accepted? 2 A. Yes. 3 Q. So what we have, effectively, here is evidence of 4 discussions taking place concerning the conditions on 5 which Post Office Counters might be prepared to accept 6 the system, notwithstanding that ongoing cash account 7 balances were being affected? 8 A. Yes. 9 Q. The proposal from Keith Baines and Ruth Holleran, 10 employees of Post Office Counters, was that an error 11 rate of 0.6 per cent, together with a number of other 12 conditions, would be acceptable. Was that your 13 understanding at the time? 14 A. It resonates with me, yes. 15 Q. Do you recall the advice that you gave to the parties 16 concerning the conditions upon which the system might be 17 accepted in late September 1999? 18 A. I don't recall giving advice on 376. 19 Q. Were you in favour of Post Office granting conditional 20 acceptance to the system at this stage? 21 A. I don't think I indicated that, no. 22 Q. Are you essentially saying that you played no part in 23 facilitating the resolution of this particular incident 24 by this late stage in September? 25 A. By implication, I suppose I had an impact in the parties 174 1 getting to that position but there was a subsequent 2 meeting between the two senior people, which continued 3 to debate what those conditions should be and that ended 4 in an agreement that further staff work was necessary to 5 understand the implications of those agreements and 6 I didn't play a part in that. My understanding was that 7 there was some further negotiation which resulted in 8 an agreement for somewhat different conditions but 9 I don't know what they were. 10 Q. Just dealing, first, with the level of your involvement 11 with this AI, you were of course present at the meeting 12 on the 17th, when these conditions were being mooted as 13 a possible condition for accepting the system and 14 rolling it out? 15 A. Yes. 16 Q. Even if you didn't give specific advice on it, you were 17 aware that this was what was under discussion by the 18 parties; that's right? 19 A. Yes. 20 Q. To accept the system with ongoing cash account 21 imbalances, did you not consider at the time that this 22 represented quite a significant risk to Post Office 23 Counters? 24 A. And I think we discussed that. 25 Q. Because an error rate of 0.6 per cent in any given week 175 1 would have equated to more than 100 Post Office branches 2 when spread out to the national level. 3 A. No, it's quite significant. 4 Q. Now, I think you say you don't recall having any part in 5 the final resolution that was reached. I wonder if we 6 could look at POL00083907. This is an email from Keith 7 Baines on 22 September 1999 to a number of employees of 8 Post Office Counters, Andrew Simpkins, John Meagher, 9 David Miller, David Smith and Ruth Holleran. It 10 records -- the subject of it is "The AB and RAB on 11 Friday". Do you know what that is a reference to? 12 A. Acceptance Board and Release Acceptance Board. 13 Q. I think it's the "Acceptance Board" and the "Release 14 Authorisation Board", possibly? 15 A. Right. 16 Q. It says: 17 "John/Andrew 18 "At this morning's briefing session with Stuart 19 Sweetman on the acceptance position, there were some 20 discussions about the role and empowerment of the AB and 21 RAB. I have since spoken to Jeff Triggs and obtained 22 his view on this and then discussed with David Miller. 23 The position is as follows: 24 "Post Office Counters will not be accepting the 25 service against the existing contract and therefore the 176 1 nature of the decision at the acceptance board is 2 different to that originally intended. The board should 3 make a recommendation as to whether or not the second 4 supplementary agreement which has been negotiated with 5 Pathway over the last few weeks should be signed. The 6 supplementary agreement then states that acceptance is 7 deemed to have taken place and the various contractual 8 consequences of that, such as payment to Pathway, will 9 follow. 10 "The same applies to the RAB since the 11 supplementary agreement says that Post Office Counters 12 Limited has authorised roll-out. The supplementary 13 agreement is formally a change control note to the 14 contract and, therefore, can be signed by David 15 Miller -- it doesn't need Stuart's signature. Can you 16 please make this rather subtle change in the roles of 17 the meetings apparent in their agendas, please?" 18 That's from Keith at 13.27. 19 Can we turn to the next page please. There's 20 a further email on the same date at 13.51, the subject 21 being the "Supplementary Agreement": 22 "The enclosed is my understanding of the position 23 we agreed at the end of yesterday's meeting with 24 Pathway. There is one area not yet agreed -- namely the 25 question of how to count incidents under AI298. The 177 1 wording in the enclosure is that suggested by Pathway 2 and recommended by Peter Copping. We were not able to 3 agree it yesterday because we don't have the right 4 people available to review it. Can John and Ruth look 5 and comment to me? Copied to Jeff Triggs, please." 6 So can we turn to the following page, please. So 7 here we have the position reached in negotiations on the 8 21st, relating to Acceptance Incident 298. What the 9 previous email suggests is you had some input on the 10 wording of this particular -- 11 A. On 298 -- 12 Q. Do you recall that? 13 A. -- yes. 14 Q. That provided that: 15 "The occurrence of operational incidents in 16 connection with this AI should have been reduced below 17 a target threshold as measured over the four weeks. 18 Measurement will be based on all outlets installed 19 before or on 1 October 1999, provided there are at least 20 750 such outlets." 21 If we scroll down to the penultimate paragraph: 22 "The target to be met is that the rate of 23 occurrence measured over the four-week period to 24 mid-November should average no more than one unit per 25 counter position per three months." 178 1 So that essentially was the target set for 2 resolution of AI 298; is that correct? 3 A. Correct. 4 Q. And on the following page, please, we have Acceptance 5 Incident 376. This refers to the arrangements for the 6 integrity control to be implemented by Pathway by 7 31 December 1999 and that those will be as previously 8 required by Post Office Counters, apart from the 9 following amendments listed below. So is it right to 10 say your evidence is you had no involvement in the 11 detail of -- 12 A. I think this was an agreement that was struck between 13 POCL and Pathway without my involvement. 14 Q. After the workshop on 17 September 1999, do you recall 15 having any further involvement in the Horizon System? 16 A. My involvement ceased after 24 September '99. 17 Q. Forgive me, the meeting to which -- the further 18 resolution meeting to which we've referred? 19 A. Sorry, I'm not following. 20 Q. Sorry, my question was whether your involvement ended on 21 the 17th but you attended a further meeting as we've 22 just -- 23 A. I attended one meeting which was between Richard 24 Christou and David Miller, I think, which was the first 25 stage of the escalation process defined in the 179 1 acceptance documentation. That meeting ended with an 2 agreement between the two that further staff work would 3 be necessary in order to understand whether or not there 4 could be an agreement on 376. I did not take part in 5 that process and I'm not familiar with the output. 6 Q. At the point at which you ceased to be involved in 7 Horizon, what was your professional assessment of the 8 robustness of the system? 9 A. I think my overall assessment was that the Post Office 10 had, in accepting the system -- and this is 11 a benefit-of-hindsight judgement -- had accepted further 12 risk in agreeing to accept the system and release for 13 roll-out with the proposals from ICL, particularly on 14 376. We certainly talked about what needed to be put in 15 place in order to monitor and mitigate any risk arising, 16 but I really don't know what happened after I left the 17 project in terms of risk mitigation and further testing 18 of the bug fixes that were being put in place beyond the 19 acceptance timescale. 20 MS HODGE: Thank you, Mr Copping. I've no further questions 21 for you. There may be some questions from the 22 representatives of the Core Participants. 23 Questioned by MR STEIN 24 Q. Sir, yes, I have just one short area of questions that 25 have been permitted by the Inquiry team. 180 1 Mr Copping, my name is Sam Stein. I represent 2 a number of -- a large group of subpostmasters, 3 mistresses and managers. 4 I'm going to take you to your report that you 5 dealt with already with my learned friend at 6 POL00028092. Paul, if you're handling -- thank you very 7 much -- if you're handling what we see on the screen, 8 could you go to internal pagination on Relativity page 7 9 of 132 and roughly two-thirds of the way down you will 10 see the paragraph, Paul, that starts: 11 "Our key concern ..." 12 Could you highlight that paragraph. Thank you. 13 Now, Mr Copping, I am just going to remind you of 14 what is being said here within this report: 15 "Our key concern is that the skills required for 16 many of the new senior post are, in our opinion, not 17 those we would have expected to find as part of POCL 18 core competencies. This is especially true in relation 19 to implementation management and contract and service 20 management. There seems, however, to be no evidence of 21 external recruitment activity or robust plans to create 22 the competence internally." 23 Mr Copping, that seems to foreshadow a fairly bad 24 problem within POCL; do you agree? 25 A. Yes. 181 1 Q. The reason for that is that you are talking about major 2 parts of the future planning. That's implementation 3 management; that's putting it into place? 4 A. Yes. 5 Q. Contract -- that's oversight presumably of the 6 contract -- to get Horizon working? 7 A. Yes. 8 Q. And then oversight of the service which is then being 9 provided by Horizon? 10 A. Yes. 11 Q. These are core competencies? 12 A. Yes. 13 Q. When subpostmasters and mistresses started to use the 14 Horizon System, they found, as their evidence has set 15 out, that the training was inadequate, that there were 16 difficulties with the operation of the Horizon System, 17 and, in particular, difficulties with trying to make 18 sure that they could achieve balance. 19 A. Yes. 20 Q. Achieving balance in relation to their accounts is an 21 important part of their process; do you agree? 22 A. Yes. 23 Q. If there is insufficient oversight and knowledge of the 24 Horizon System within the Post Office, within POCL, does 25 that mean that these particular difficulties that 182 1 postmasters and mistresses were suffering from might not 2 be remedied by the Post Office? 3 A. I don't think I can answer that question. I think at 4 the time we did this review in '97 there was no doubt in 5 our minds that the Post Office had a shortfall in 6 competent resources in the areas we discussed. There is 7 no doubt in my mind that the Post Office did resource 8 up, and my recollection would be that David Miller was 9 the first significant appointment that was made in that 10 resourcing up process. 11 I would maintain at later stages of my involvement 12 that the Post Office had a shortfall of what I'd 13 describe as general technical competence that was 14 capable of properly interrogating the Pathway personnel 15 as to exactly what was going on in the development 16 process and everything that flows from that. 17 Q. So, in other words, your concern then, and concern 18 remains towards the end of your time working on this 19 project, that the Post Office might not have the 20 technical ability to understand what's actually 21 happening within the system; is that fair? 22 A. I think that's fair comment and it's broader than that 23 as well. There was also a readiness for acceptance of 24 the system within the Post Office POCL organisation and 25 all that implies in terms of service management on the 183 1 Post Office side, as opposed to the technical side, 2 which was ICL's responsibility, and the need for process 3 change in order to support new ways of working. 4 Q. And finally, it's stating perhaps the blindingly obvious 5 but Horizon was a new way of working for the Post 6 Office? 7 A. Absolutely. 8 MR STEIN: Thank you, sir. 9 SIR WYN WILLIAMS: All right. Well, thank you very much, 10 Mr Copping, for providing written evidence to the 11 Inquiry and for coming to give oral evidence. I'm 12 grateful to you. 13 MS HODGE: Sir, I believe Ms Page has some questions for the 14 witness as well. 15 SIR WYN WILLIAMS: Oh, I see. Sorry, I hadn't appreciated 16 that. 17 Questioned by MS PAGE 18 Q. Just one, in fact, from the questions we were permitted 19 and it's on Acceptance Incident 376, which you've just 20 described, as you did in your statement, as something 21 which you understood POCL to be taking on more risk as 22 a result of the position as it was left when you no 23 longer were involved. 24 A. Correct. 25 Q. Would you therefore have expected those risks that they 184 1 were taking on, those additional risks that they were 2 taking on, to be registered in some way perhaps by the 3 board or by some management level and monitored until 4 they were satisfied that those risks were no longer 5 significant? 6 A. Yes, and I don't know whether this was put in place. 7 What would normally happen on a project of that sort 8 would be a full risk assessment process which is updated 9 on a regular basis within the programme, and if 10 insufficient progress is being made (for example, on bug 11 fixes and regression testing), then those risks would 12 begin to replicate themselves over time and that, in 13 turn, one would expect would escalate the issue to 14 a higher level through the organisation. Whether or not 15 that was put in place, I can't say. 16 Q. But that's what you would have expected? 17 A. That's what I would expect from a management point of 18 view, yes. 19 MS PAGE: Thank you. 20 SIR WYN WILLIAMS: Right. Well, I won't repeat my thanks 21 but thanks again, Mr Copping. And I take it that now is 22 the end of this session? 23 MS HODGE: Yes, sir. Thank you that concludes the evidence 24 of today. 25 SIR WYN WILLIAMS: All right. Well, we will meet again on 185 1 Tuesday morning at 10.00. Thank you all very much. 2 MS HODGE: Thank you. 3 (4.27 pm) 4 (Adjourned until 10.00 am on Tuesday, 25 October 2022) 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 186 1 I N D E X 2 PAUL RICH ........................................1 3 Questioned by MR BLAKE (continued) ...............1 4 Questioned by MR STEIN ..........................95 5 PETER JAMES COPPING (sworn) ....................100 6 Questioned by MS HODGE .........................100 7 Questioned by MR STEIN .........................180 8 Questioned by MS PAGE ..........................184 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 187