Examination of Witnesses (Questions 880-899)
MR TONY
COOPER, MR
SIMON VRY
AND MR
IAN HEWETT
27 NOVEMBER 2006
Q880 Lynne Jones: You were not involved
at that point?
Mr Vry: I was not at the Agency
at that time.
Q881 Lynne Jones: Do you know who
was?
Mr Vry: No. I do not know the
details of who accepted that business case.
Q882 Lynne Jones: When you took over,
you thought it was a very reasonable proposition and were happy
to go along with it?
Mr Vry: When I joined, in November
2001, I looked at the business case as part of my review before
joining the organisation to see what was proposed and the business
case itself seemed reasonable and sound. It was recognised even
then as being high risk, and the original proposal, to implement
it inside of 12 months, from PWC was rejected and a longer timeframe
was set for implementing the changes that were necessary because
it was thought that the timeframe was going to be too aggressive.
We then went into a phase, at the same time as I joined, of recruiting
a supplier to help RPA, which did not have that experience, to
develop an IT system and to help with the business process re-engineering,
to change our business processes to map onto that new way of working.
That procedure went through and culminated in the appointment
of Accenture at the end of January 2003. Then we went forward
with Accenture starting to work on business process re-engineering
to look at all of the business processes that RPA had got across
all of its schemes to see how they could be re-engineered to fit
into a task-based system. RPA had already done some work around
specifications for that and shared those with the suppliers who
had been short-listed and asked them to look at that, and there
was opportunity there for parties to view what was the best approach.
In the end, we concluded a contract with Accenture which was based
around a model which was task-based processing. That was a period
whereby we had the opportunity to review that. Of course, as we
moved forward, and then during the course of 2003, there were
difficulties in the business process re-engineering work which
was kicked off with Accenture, which we did jointly with Accenture.
RPA was not happy with the outcome from that work and we did not
think, at the time, that it reflected necessarily the requirements
of the business or, in particular, how we would need to comply
with EC regulations in operating the schemes. That piece of work
was rejected by RPA and at that point we decided not to proceed
with that work with Accenture. In the same timeframe began to
emerge the changes of the CAP reform and so there were detailed
discussions held between RPA, Accenture and Defra in terms of
what would be the appropriate way forward, given the fact that
the whole locus of the Change Programme was about to change. This
being from one driven by benefits, in terms of implementing a
new IT system to put all of the schemes onto one IT system to
reduce the number of staff needed, to deliver a lower-cost operation,
and the underpinning system therefore was to deliver that flexibility
that we needed, to one which was driven more to do with delivering
a new scheme, the requirements of which were not known. Therefore
it being inherently more risky, in terms of specifying the requirements,
than one where we had run, in particular, the 11 schemes which
were replaced by SPS, and the whole profile of the Change Programme
changed at that point. We went through the process with Defra
and Accenture of looking at what the options were, and decided
that the focus of the Change Programme should be to deliver the
Single Payment Scheme and that we would continue developing the
system which was already in stages of development to deliver Single
Payment Scheme as the priority. There was still the element of
bringing the other schemes on board, but at a much later stage.
That pushed back the length of the Change Programme; at that stage,
it was estimated that an additional year would be required. We
then worked through a new schedule to deliver that IT and all
the supporting business processes which would be needed, and training,
etc., to support the implementation of the scheme, and settled
on a set of dates which we thought were achievable and realistic
at that time. As we went through successive months and years,
obviously some of those dates were not met; as it talks about
in the Report, there were delays in the delivery of policy and
there were subsequent changes in the European policy, which meant
that as we progressed with the design of the system the requirements
which that system was to meet had changed and we had to go back
in and redesign elements of the system and retest them. That is
why, in the Report, where it talks about the timeframes for the
IT being compressed and there being limited amount of testing
and no end-to-end testing, it is that picture. The original locus
of the Change Programme was focused for a very different thing,
and implementation of SPS changed the direction of the Change
Programme. Unfortunately, as we moved through to the later stages
of the programme in delivering SPS, it became increasingly apparent
that a task-based approach alone was not adequate to give us proper
control of knowledge of how quickly we were getting through individual
payments. Our difficulties were that, as part of the necessary
descoping to get the system delivered on time, one of the key
components of that, the management information, was not in scope,
it had been removed from scope. Therefore we had information which
gave us a view of how many tasks were outstanding but actually
we had problems with being able to manage through to complete
the payments in the timeframes that we had expected. As I said,
when we started actually changing the scope of the programme,
we had a timeframe, that timeframe changed over a period of time
and became fore-shortened by delays in realising policy. That,
in turn, impacted upon the testing timeframes, which meant that
the testing was not completed to the level of detail that we would
have wished, and the processes suffered similarly from compacting
on the timeframe. We spent a significant amount of time looking
at alternative options, including the development of a contingency
solution as an alternative to the RITA processing function, and
also looking at potential options for partial payments or advance
payments. Significant amounts of RPA and Defra management time
were spent looking at those options and how best we could deliver
that. Also we used the Office of Government Commerce to provide
regular reviews of progress, to give us an indication of whether
our own views were realistic or whether they were either too optimistic
or too pessimistic, to try to give us that balance. Clearly, at
the end of the day, we were not able to get the volume of payments
out to the customers in the timeframes that we expected, and we
got that badly wrong, and I do not think anybody is arguing that
is not the case. We did set a forecast, in January 2005, that
the earliest we could make payments would be February 2006 and
we met that objective, so we managed that process through. Where
we got it badly wrong was that actually, the process of finishing
off the validation around these payments and getting the payments
out of the door, we got that wrong in the sense that we predicted
we would do it sooner than we did. That was a difficulty of running
a new IT system, with new business processes, for the very first
time. That was why, when I was talking about the elements of risk,
the risk profile was increasing, because things that ideally we
would have liked to do, like completing end-to-end testing, were
not practical in the remaining timeframe.
Q883 Lynne Jones: Were you quite
happy to go along with the optimistic scenarios which were being
put forward, the bullish reports that Johnston McNeill was putting
forward, you felt that they were realistic, knowing, as you have
just described, all the examinations of the systems which took
place, looking at alternatives, which all appear to have been
rejected?
Mr Vry: The benefit of hindsight
is a wonderful thing. Working as a team, we sat down ourselves
in RPA and with Defra and with external advice, we looked at the
options, we looked at what were the realistic options at the time
and what we thought was the viable way forward and, based on that,
we felt that it was deliverable. Indeed, the initial headline
of getting payments out in February we met, having predicted that
a year ahead. What we badly underjudged was actually the ability
to process the payments and get the bulk of payments out in a
reasonable timeframe thereafter.
Q884 Lynne Jones: You were not put
under any pressure, either by senior management in your team or
politically, which prevented you from expressing any concerns
that you might have had?
Mr Vry: I think it is fair to
say that everybody felt pressure, not that they were put under
pressure but felt pressure, that we wanted to provide good, early
payments to our customers, which were accurate, and that, at the
same time, we wanted to provide value for money for the taxpayer
and not encounter significant sums of disallowance.
Q885 Chairman: Can I be clear on
one thing. Earlier on, you were describing the process of the
changing demands that were being put upon the system that you
were building as part of the Change Programme, and, if you like,
the pressure accelerated as new elements of the Single Farm Payment
policy emerged. Did you ever say, at one time, "We need more
time in total to deliver the package" to the standards that
you wanted and did you ever flag that up to ministers with a request
that the process should be slowed down, that the introduction
of the Single Farm Payment should be delayed?
Mr Vry: There were discussions
around the risks of implementing the IT system in constrained
timeframes and the risks that could pose and what our alternative
options were, and issues including postponing the implementation
of it were raised and discussed. One of the issues was that in
England we had committed to the European Commission, I believe,
I think it was in August of 2004, if my memory serves, that we
would be implementing the scheme in the 2005 scheme year. Therefore,
it was felt that had to be met and we had to deliver that, therefore
we were looking for alternative contingencies to mitigate the
risks around the IT delays, but all options were considered.
Q886 Lynne Jones: Who would have
told Lord Bach that the impact of reform would not be significant
on the overall IT solution?
Mr Vry: I am sorry, I do not know.
Q887 Lynne Jones: You do not recognise
that phraseology: would you say that it was not an accurate point
of view? Mr Hewett implied earlier that it was not an accurate
point of view because he was talking about major changes as a
result of CAP reform.
Mr Vry: I can say, speaking with
hindsight, and hindsight is a wonderful thing, that it is quite
clear that the impact of CAP reform was a major issue on the overall
Change Programme and RPA's ability to commit the IT.
Q888 Lynne Jones: When do you think
it became known that it was a major issue?
Mr Vry: I think that was one of
the issues. It was not at any one point in time, there was a slow
accumulation of issues, and if the question is at what point did
the tipping-point become—
Q889 Lynne Jones: It was a slow accumulation
of issues and it just crept up on you and you did not notice it
happening?
Mr Vry: It crept up on us without
anybody noticing. If you look at the reports, the reports do state
the key issues and there is a lot of reporting around red, amber,
green status of the key issues the organisation was facing, and
it was recognised that it was an increasing risk profile.
Q890 Lynne Jones: Can you tell me
what the difference is between the Change Programme and the CAP
reform and the Single Payment Scheme? Is developing the SPS part
of the Change Programme, because I was led to believe that they
were two separate issues, that you were embarking on a Change
Programme and then you had to implement this new scheme?
Mr Vry: That was the discussion
which went on, around the middle of 2003, when the CAP reform
issues began to emerge, in terms of being crystallised, that CAP
reform was going to have to be sooner rather than later. The Change
Programme existed already and it had a clear set of requirements
around reducing the organisation, in terms of the number of offices
from which it operated, the number of staff which operated it
and improving its efficiency and placing the lion's share of the
IT onto one single IT system, from what was previously a plethora
of IT systems, legacy systems, which were reaching the end of
their supporting timeframes. When CAP reform began to emerge as
a current issue and looked likely to happen in the near future,
there were discussions around how it would be best to implement
that. Various different options were considered and the ultimate
decision was that it was best that the Single Payment Scheme and
the implementation of CAP reform was part of the wider Change
Programme and became the key driver for the Change Programme,
going forward, as the number one priority for that Change Programme
and for RPA itself to focus upon.
Q891 Lynne Jones: It was sensible
to let go all those experienced staff, as part of the Change Programme?
Mr Vry: Again, it is an issue
of timing. When it was envisaged originally, with the way in which
the organisation was designed to move, it was sensible that there
was going to be a downsizing in the organisation, with 11 schemes
being simplified into one scheme, and the various different aspects
of CAP reform was anticipated in the very early days, that there
were opportunities to continue with that downsizing. There was
a requirement, which was discussed, about having to give notification
to staff about that, so it was not that you could turn around
on the Friday and make a decision and then remove people on the
Monday, you would have to give them adequate notice, so there
was a period in which we had to give significant, I think it was
in the order of six months, notice of the intention to do that.
Unfortunately, as the complexity of what the organisation was
seeking to deliver increased, some of those decisions had already
been taken and commitments had already been made to staff to allow
them to leave. It was those two issues, pulling in opposite directions
but in the same timeframes, which were one of the problems that
the organisation faced.
Q892 Lynne Jones: Since you have
moved to operating on a claims-based business process, is RITA,
which was designed for a task-based process, now fundamentally
unsuitable?
Mr Vry: Currently we are having
a review to look at that situation and we are using an external
party to have a look at the applications that RPA operates and
the infrastructure on which they are provided, to ensure that
both the applications and the infrastructure are suitable for
RPA's needs going forward. We do not have the review outcome from
that yet. I spend some time talking to staff and getting their
views and ask them what they think about the system, and there
is frustration about some aspects of the system, about the design
and how it is not, as Tony Cooper was saying, as user-friendly
as it could be, and there are certainly opportunities for us to
improve that. The system continues to process claims from receipt
through to payment, so it does work and it continues to work.
Therefore it is not possible to say that it is not fit for purpose,
because it does the job, as things stand. Does it do the job well
enough: no, I think we could do more.
Q893 Chairman: Throughout your response
to Lynne Jones's questions, you have used the word `risk'. Risk
is a concept about probabilities of either achieving or not achieving
something. Did you quantify the mounting riskiness, as you went
on, numerically; in other words, as we go through, the percentage,
i.e. the chance, if you were a gambler, of not achieving the results
you have set out to? Obviously, in gambling terms, the odds would
be shortening all the time. Did you quantify risk at all as you
went on?
Mr Vry: We tried to quantify risk
and actually set down some examples of how we thought that risk
profile was changing over time, but it is not a scientific exercise
where we can say we are 75% confident and be sure that is 100%
accurate. We tried to give some indication of the level of risk
and how that was changing over time, and indeed in the reports
that we provided there were graphs indicating risks for different
areas that we were operating under which showed the movement of
those risks over time.
Q894 Chairman: Looking at the summation
of the Gateway Review report, within the NAO report, it looked
to me as if you had managed to get nearly a complete line of red
traffic-lights on 18 August 2005. You have got only one subject
which is an amber light, which was `Scope' of the project. On
that particular assessment it is red lines, all bar one, right
across. On 22 September, you have got two ambers, the rest were
red. Then you were back to all, bar one, on 18 October. That says
to me that, in that period, the risk factors were increasing quite
substantially.
Mr Vry: Are you referring to Appendix
Six?
Q895 Chairman: Yes. This is actually
RPA's report, page 43.
Mr Vry: This is the report that
RPA did for the Executive Review Group; it is not OGC's report.
Q896 Chairman: Even better, if you
produce these reports yourself, there is this mounting tide of
red traffic-lights and yet nobody puts on the brakes, the car
continues to career towards the edge of the cliff?
Mr Vry: These red, amber, green
ratings were constructed in accordance with the OGC standards
for the red, amber, green, and red meaning that actions need to
be taken immediately to address the situation.
Q897 Chairman: The more reds you
get the more, as a layman to these matters, it says to me there
is a mounting tide of uncertainty about what we are going to be
able to do. In terms of the kind of decision-making process over,
for example, interim payments, it is just interesting to look
at the increasing occurrence of red traffic-lights against a background
of "It'll be all right on the night," amounting to Lord
Bach coming to this Committee saying everything would be okay
and then, within almost days, weeks, suddenly it was not going
to happen. I just do question the competency of the assessment
of risk, particularly during that second half of 2005, when your
own reasonably objective assessment shows a mounting tide of risk
increasing?
Mr Vry: There was a key issue
of risk, and we used OGC to help review constantly our progress
to make sure that we were not being overly optimistic, or indeed
overly pessimistic, to give us a way forward. Indeed, we even
brought in OGC in October of 2005, and at the end of 2005 they
advised us that they were convinced that the programme had a reasonable
chance of success of delivering SPS payments within the payment
window. We had independent confirmation that, yes, it was high
risk and, yes, there were challenges and it was difficult.
Q898 Chairman: Did Accenture ever
talk to you about their view of the project beyond the sort of
narrow little world in which they seem to operate? The message
that they gave to this Committee was "We did what we were
asked to do. It worked when we tested it; the system was stable.
No, we're not offering any observations about that which we weren't
responsible for. The end." The end was it did not work. Did
Accenture ever communicate beyond their little box, or did they
just do what you asked them to do?
Mr Vry: First of all, most obviously,
there was a contract in place with Accenture for what Accenture
was being asked to deliver and we had in place contract management
arrangements with them in order to ensure that they delivered
against that contract. Also, they did sit on the Change Programme
Board. There was a seat for a senior supplier, in fact we had
two senior suppliers on that Programme Board; one was external,
which was Accenture, our lead delivery partner, and one was internal,
which was the IT Director of RPA, to provide us with the ongoing
assurance and review as to how the programme was progressing.
They were involved in the overall programme, but obviously they
had a contract which they had to deliver, which was specifically
a part of the Change Programme.
Chairman: You did not actually answer
my question. My question was did Accenture, particularly if they
were sitting on this Board, comment on anything about what they
saw going on around them?
Lynne Jones: All the additional tasks
which suddenly came up?
Q899 Chairman: Here, you have got
a company with a world reputation, which was why you employed
them, because they were good at doing these complicated tasks.
They have to understand the totality of the universe in which
they are operating, otherwise they cannot plug in the bits, their
software, to make it work; they have got to understand externalities.
Therefore, if you are looking at what is going on around you and
you are a very experienced company and operating in this field,
and if you can see risks mounting and problems occurring, did
they just sit there, like Trappists, and say nothing, or did they
offer any guidance to the management of the RPA that, in their
professional opinion, as observers of a complicated system, they
might have one or two observations to make about what was going
on around them?
Mr Vry: As I say, they did take
part, on the RPA Change Programme Board, so they were party to
some of the discussions which were taking place there.
|
