Examination of witnesses (Questions 120
- 139)
WEDNESDAY 1 APRIL 1998
MR J MORTIMER,
Treasury Officer of Accounts, further examined.
120. From what date were they involved?
(Mr Jenkins) They were involved from late 1994.
121. They were involved from 1994, through
1995 and at the point of decision to go live with the phase 3
system.
(Mr Jenkins) Yes; correct.
122. Then clearly it went horribly wrong
as we can see from the report. What was their advice then? What
was their reaction as things went wrong around you? What were
they saying to you? What was their involvement then?
(Mr Jenkins) The consultancy period had in fact
concluded by go live.[7]
The tender only related to the period up to the time of operation.
123. So they were not involved after the
decision was made.
(Mr Jenkins) That is correct.
124. Were any other consultants involved?
(Mr Jenkins) In so far as the issues of main concern
were with the software we clearly were in touch with Oracle. We
consulted them on the need to make improvements. As indicated
in paragraph 24, there were these two software bugs for example
and we were in very close contact with Oracle to resolve these
issues.
125. You were without independent advice
apart from Oracle who were actually writing the software and advising
you on it after you made the decision to go live. I am confused
about how consultants fit into this whole picture and I am just
trying to get a picture over time of the role consultants were
playing at each stage.
(Mr Jenkins) We employed the consultants for assistance
with the development of the system. That was KPMG for phase 3.
Following the go live situation in May, we retained some technical
consultants and we still retain them, IBM, for the routine maintenance
and correction of upgrades as you always require in the computer
systems. We are also consulting Oracle Financial as the suppliers
of the software. We are in regular touch with them.
126. The cost overruns for implementation,
including phase 1 and 2 and including phase 3 amount to some £4.7
million, that is the overrun on the estimates leading to the total
of around £11 million on implementation to date. Is that
about right?
(Mr Jenkins) Without quickly looking at the figures,
that sounds about right generally, yes.
127. That is the difference between the
estimate for phase 1 and 2 of £2.65 million and therefore
it came out at £6.6 million and between £2.62 million
and it came out at £3.46 million so that cost overrun is
around £4.5 million or thereabouts.
(Mr Jenkins) Yes.
128. The consultants who were involved up
to and including the decision to go live were a part of that decision
making and therefore could be expected to bear some liability
for that cost overrun. Would that be true or not true?
(Mr Trevelyan) We felt the main failure in the
system arose from the software and therefore from the software
supplier and therefore from Oracle. The finger in industry has
pointed in that direction and has led Oracle to issue a number
of statements in explaining their position. More importantly,
the contractual arrangement we had, was with Oracle as supplier
and that is why they have been collaborating with us under the
terms of the original contract in effectively repairing or rebuilding
the system in the areas where we found it not to be functioning
properly.
129. What was your contractual arrangement
with KPMG? When you called them independent advisers, what was
the nature of that contract? What role were they playing? If Oracle
were doing all the work and doing all the software and getting
all the blame, where were KPMG in all of this? What were they
taking their money home for doing?
(Mr Bryant) We had a fixed price contract and
we set it out as a project with specified deliverables over specified
periods of time. Each time a deliverable was set up by KPMG it
was tested by the agency and what we would technically describe
as signed off. In other words we would say we accepted that part
of the delivery in accordance with the terms of their contract.
Therefore, having accepted it, it would in essence relieve them
of any responsibility which we had not detected in our testing.
130. They had a get-out clause essentially
in the contract with you that if something went pear shaped, as
it did, they did not have to bear any of the consequences.
(Mr Bryant) We have never described it as that
at all. What we said was that it was up to the agency to take
the decision on whether what had been delivered was operable.
So we tested it and took that decision. It was not KPMG's decision
as to whether I ran the software.
131. I am glad it was not. It is your job
to make the decisions and theirs to give advice but whether advisers
give good advice or bad advice they have to stand by that in some
way. Was one of the deliverables then an assessment of the wisdom
of going live or not? Was that part of the contract they were
asked to deliver, which you signed off, an assessment of whether
to go ahead?
(Mr Bryant) It was more in the negative. The decision
was the Board's but they saw no reason why we should not proceed.
132. That sounds like a non-denial denial
so I will just try to press it again. They did not say you should
but they did not say you should not.
(Mr Bryant) It was not their decision.
133. It was not their decision but advice
I am getting at. Was it their advice that it was a wise move to
go ahead on 1 May? Did they give judgement? Was it part of their
contract to give a judgement? What were they there for with regard
to that decision? Were they giving advice or not? If the answer
is no, just say no. I am happy to accept no as an answer. I am
just trying to find out what the answer is.
(Mr Trevelyan) Mr Jenkins was the chairman of
the project board and is probably the best placed.
(Mr Jenkins) The precise terms of the contract
did not say "Thou shalt give advice as to whether or not
we go live". They were party to both the project and also
another committee which I set up, an executive committee of the
key players, where a partner from KPMG was present. We discussed
regularly the question of go live. As part of the combined agreement
they were party to that expression of agreement. All parties there,
including KPMG and the Intervention Board, agreed it was time
to go live. The final decision clearly rested with me on the advice
I received.
134. Would it be too much to ask for a brief
note about the role KPMG were playing? I do not want to see the
full contract but I should like to have some clarification of
exactly what role that consultancy group were playing in this
process. That would help me.[8]
(Mr Trevelyan) Certainly.
135. Another consultant is mentioned. Forgive
me but I am intrigued by the whole area of how consultants work
and how we manage and relate to them as agencies. Paragraph 27
talks about another consultant being brought in to do an assessment
of the system and the problems you were having in October 1997.
Who was that? It says, "The consultant's preliminary observations"
etcetera.
(Mr Trevelyan) That was KPMG as an independent
contract consultant who then supplied us with a team of contract
consultants with whom we have been working since then.
136. Up to and including the decision to
go live in May they were there. It went pear shaped and badly
wrong and they were at the committee where decisions were being
made and where things went wrong and you have taken full responsibility
for that, taken any burden away from them. Then at the very point
that you realise it has gone badly wrong and you think you had
better get this sorted out you bring the same consultants back
in who have been sitting at the table whilst you were making decisions
which clearly went wrong. Was that appointment put out to competitive
tender?
(Mr Trevelyan) No. We wanted immediate assistance.
The urgency was quite clear.
137. Why did you not go to competitive tender?
Why did you appoint the same consultants you had been sitting
round the table with for the previous 15 months who failed to
spot the problem you were about to hit? You then hit the problem
and you appoint the same consultants to tell you that you have
hit a problem. Do you find that bizarre, strange?
(Mr Trevelyan) No, because what the Committee
is tending to obscure is the distinction between the software
supplier who supplies you with the kit and the contractor who
gives you advice on how to implement that. The implementation
in this particular case and the problems we were facing was a
matter of accountancy. It was a matter of how, given the software
you had in place, an accountant draws the books to account. That
is not Oracle's job. Oracle provided a piece of software which
was not providing satisfaction and we were pursuing them with
our own resources. There is quite clearly a balance between our
own resources, our own internal accountants, any advice we need
on the operation of the system, which would be KPMG or another,
and the software supplier which is Oracle. It may seem a complex
arrangement and I am certainly very happy to explain it further
in the note which you have asked for.
138. I still do not understand how you made
a decision with a group of consultants present who do not bear
any responsibility and then they go away and then when it has
gone completely wrong with a catalogue of problems which have
been sorted out you appoint the same consultants to come in to
tell you something which I then understand, if I have read this
right, paragraph 27, "The consultant's preliminary observations
confirmed those already identified by my staff", that is
the Comptroller and Auditor General's staff, "and by the
Internal Audit report". If you had already been told what
was going wrong, why did you appoint consultants to tell you again?
(Mr Jenkins) May I clarify? I believe you referred
to paragraphs 27 and 28.
139. Twenty-eight is the IBM consultants,
is it not?
(Mr Jenkins) In fact it is the other way round.
Paragraph 27 is a very short term review by IBM to assess where
the main problems were. That is the paragraph in which the NAO
comments that confirmed the views expressed earlier.
7 Note by witness: In fact, the consultancy
did not include the support of live running." Back
8
Note: See Evidence, Appendix 3, p. 25 (PAC 268). Back
| 
