POL00334118
POL00334118
In Confidence
PIR(02)08 POST IMPLEMENTATION REVIEW BOARD
Horizon Implementation
Introduction
This project started in May 1996 as a joint project with the Benefit Agency with the
primary objective of introducing a Benefit Payment card. After significant delays and
an extended review the Government decided in May 1999 that this card would not be
introduced and that the Benefit Agency would migrate benefit payments to Automated
Credit Transfer and withdraw from the project. The Post Office Board endorsed the
transfer of the project from the Benefits Agency in May 1999, and the primary aim of
the project became the automation of the counter network.
This paper covers the period from May 1999 until the successful completion of National
Roll Out in June 2001. However, the financial data in Annex A does include the direct
costs and sunk cost incurred since the start of the project in April 1996. The period prior
to May 1999 was subject to a National Audit Office report, published 18 August 2000
(www.nao. gov.uk/publications/9900857.pdf) and is not covered by this PIR unless it aids
the understanding of the post May 1999 situation.
Background
The project started as a joint project with the Benefits Agency with the primary aim to
introduce an automated benefit payment card. When the Benefit Agency withdrew, Post Office
Counters Limited took over the project and the primary aim was to automate all offices in
POCL retail network.
In 1996 Post Office Counters Ltd (POCL) and the Benefits Agency (BA) entered into a
Private Finance Initiative contract with ICL Pathway to provide automated services for
payment of benefits through post offices and to support other existing and new POCL
business. After December 1996 the programme experienced many slippages against its
plans and was subject to several reviews.
In 1998 a further review was commissioned by the new Labour Government to look
into the future of the programme. This review, involving all 3 parties, took far longer
than expected and concluded in May 1999 when Government announced that:
e the Benefit Payment Card would not be implemented;
¢ BA would migrate payment of benefits to Automatic Credit Transfer during
the period 2003-05;
¢ POCL would continue to contract with ICL Pathway to automate the
network albeit under a different pricing regime.
1
In Confidence
POL00334118
POL00334118
In Confidence
5. The Post Office Board endorsed this proposal on 24'" May 1999 and POCL signed a
Letter of Agreement and started a live trial with 300 offices. This was followed by the
Codified Agreement on 28" July 1999. Contractual acceptance was scheduled for
August butthe existence of a number of high severity faults meant that this was
delayed. However, the delay in acceptance did not prevent the go ahead being given
for a further 1532 offices to join the trial. Acceptance was formally given on and the
final phase of implementation began on 24" January 2000.
6. The Automation Directorate was reorganised in January 2000. During much of 1999
management attention was strongly focused on live trial, acceptance and the early
phase of roll out. By contrast the development and implementation of Core System
Release Plus (CSR+), a software drop of a very significant scope and size, received
scant attention. Dedicated teams were created for both the National Roll Out and
CSR+ to ensure the necessary degree of focus on each activity. It is strongly believed
that this decision contributed to the successful delivery of National Roll Out and
CSR+.
7. The project gave rise to other related projects such as Transaction Information
Processing (TIP), Reference Data, Business Support Management. Each of these was
a significant project in its own right. A Programme Management Team, the
Transformation Management Team, and a governance process was established to
ensure effective integration.
8. The project also had a wide impact across the business. During national roll out the
retail line was extensively used to support roll out and Transaction Services had
changes in their work load with the level of errors increasing.
Strategic Implications
9. The project had its origins in the Post Office Counters Ltd. counter automation
strategy as part of business mission and vision to create an automated retail network.
Strategically the Horizon system provides a nationwide network of post offices with an
automated platform to gain and retain clients who require electronic transactions,
either now or in the future. It has the potential to provide a electronic gateway into
Banking, EFPOS, and Your Guide. It supports the strategic aim of being seen as a
business which harnesses technology.
Results
[ The aims and objectives of the project were achieved within the cost and time targets. I
10. The main objectives of Horizon implementation phase are defined in the Project
Initiation Document for Horizon NRO and performance against these is detailed in
Annex B.
2
In Confidence
POL00334118
POL00334118
In Confidence
1. Expenditure against authority: the total project has a maximum authorised sum of
£662.068m. The actual outturn is as follows:
£ millions Authorised sum I Actual I Variance
Capital payments to ICL Pathway 564.000 I 561.693 2.307
Sunk Cost 47.841 47.841 0.000
Non Recurring Revenue final phase (NRO/CSR+) 41.427 41.600 -0.173
Departure from estimate (final phase) 8.800 0.000 8.800
Total Authorised Sum 662.068 I 651.134 10.934
12. Capital payments were reduced by making two payments to ICL Pathway early in
order to obtain a discount.
13. The Non Recurring Revenue spending (NRO/CSR+ ) was broadly in line with original
projections. The savings made were due to a reduction in number of offices to be
implemented, and to lower than expected costs of office modifications; these were
sufficient to cover the increased costs of POiT and PO Consulting arising from the
new charging regime. The majority of the Departure From Estimate was not required,
the programme roll out rate proved sustainable and on occasions was exceed. The
potential problems around training, and go-live office support, which were of concern
to the capital board, proved to be manageable and extra costs were not incurred.
14. The authorisation included costs £1.2m related to supporting system costs, namely
Interim TIP and Reference Data. These eventually became projects in their own right
and these initial costs £1.7m have also been reported against these projects, but for
ease of comparison these costs have been retained
15. The value of the assets acquired by the project were impaired and written down by
£571m in the 1999-2000 Accounts in order to comply with Financial Reporting
Standard FRS 11.
16. Implementation against original timetable: performance against key milestones is
detailed in Annex C.
Achievement of Planned Benefits
The project enabled POCL to bind the BA into a long term relationship and the project has
exceed the financial benefits as defined in the original plan.
Financial Benefits
17. The business case used a discount rate of 12% and had an overall a negative Pre Tax
NPV of £181m. Using the same basis and the actual out turn it has a negative Pre Tax
NPV of £96.2m. This £84.8m substantial improvement has been generated both in
terms of reduced project spending and in substantially improved projections of the
business benefits, mainly income. Explanations of these improvements are detailed
later in paragraph 22. However, this NPV calculation includes avoided income loss not
in itself an incremental cash flow
3
In Confidence
19.
20.
21.
(ii)
(iii)
22.
POL00334118
POL00334118
In Confidence
The business case Post Tax NPV was negative £97.8m.The actual out turn Post Tax
NPV improves by £103.6m and is positive £5.8m. This out turn is subject to the
agreement of the Capital Allowances by the Inland Revenue.
When the project was initiated the primary purpose for POCL was to bind the BA into
a long term relationship, with a secondary purpose of providing an automated platform
to all offices in the network.
The financial savings attributed to switching to Horizon identified in the business case
were limited to the avoided cost of five legacy systems. These systems were Electronic
Cash Registers at Counters (ECCO+), Automated Payment Terminal (APT),
Automation of London Post Offices (ALPS), and CAPTURE, and the central support
Host Automated Payment System (HAPS). Savings of £24.6m (business case £21.8m)
have been achieved.
The major area of financial benefit arises from gaining and retaining client income and
the business case assumed that Automated Credit Transfer (ACT) was to be heavily
promoted from 2003, but not compulsory, and that the network continued to contracted
at the then current rate. The benefit can be broken down into three areas:
Benefit Distribution Incremental Income. Implementing Horizon enables the
business to encourage transaction substitution and thereby mitigate the impact of
promoting ACT from 2003. The business case had a NPV benefit of £116.9m, the
revised business case has improved to a NPV benefit of £153.5m. The £36.6m
improvement is due to actual traffic volume being higher than in the business case
projection. This has then followed through to future years projections where a more
optimistic view of traffic volumes has been taken.
Order Book Control System Income. Previously provided by ALPS but only within
the M25. Implementing Horizon enables this service to be extended to the whole of the
country and gain additional income, the business case had a NPV benefit of £101.8m,
the revised business case has increased to £118.9m. The position improved because
contract negotiations were still taking place when the business case was being
prepared and a more favourable outcome than expected was achieved. Renegotiation
of the contract was seen as a major risk in the business case projections. £91.1m has
already been received and included in the accounts.
Other POCL contribution. By agreeing to implement Horizon enabled the business
to prevent the Benefit Agency immediately switching to compulsory ACT in 2000.
Had this occurred there would have been a dramatic drop in income but the network
fixed costs would still have been in place. Therefore, the business case had a NPV
benefit of £148.9m which was an assessment of the costs which would not be
recovered had this occurred. This was a one off calculation which has not been revised
since the submission.
Operational Benefits: although outlet operations as measured by quality of service
and quality of performance, were expected to be adversely affected in the 12 weeks
following an office implementation; there was to be no long term impact. Operations
4
In Confidence
23.
24,
0)
25.
POL00334118
POL00334118
In Confidence
were expected to return to pre implementation levels following this learning curve
period. Consequently no benefits were included in the business case. PO Consulting
researched the projects impact on outlets quality of service at various point in the
projects life, the conclusions are summarised in Annex D.
Quality of performance shows an improving trend but because of other initiatives it is
often difficult to link this directly to Horizon but in certain instances, for example
week numbers and manual transcription errors, Horizon is thought to be the main
reason for the improvement. The quality of performance trend during implementation
is detailed in Annex D.
Supplier performance: ICL Pathway performance can be broken down into three
phases:
How did they perform during the delivery of NRO?: Achievement of the peak beat
rate of 300 offices a week was against all expectations, however, considerable POCL
resources were dedicated to managing ICLP and it subcontractors to ensure this
successful outcome.
(ii) I How did they perform during CSR¢+ roll out? This was delivered to plan.
(iii) Post Implementation: Operationally the service has proved to be very reliable
and robust .
The contract with ICL is for a service not a system.. The performance of this service was
laid down in the contact specification which defines 76 measures which form the basis
of a Service Level Agreement with ICLP. This SLA is actively managed.
If a Minimum Acceptable Threshold for service is not met for three quarterly Service Level
Measurement Periods in any 24 month period, then Post Office Limited may terminate
the agreement. A number of areas exist where Post Office Ltd. has cause to terminate
the contact. However, legal advice warns of the possibility that the court would
consider termination for cause a disproportionate remedy relative to the business impact
of the contractual failure.
The detailed data on ICL Pathway performance against the service level agreement is
detailed in Annex E.
Related benefits: Horizon has been a key enabler for other projects such as Cash
Handling and Distribution SAPADS and Universal Banking. The business case
included an NPV benefit of £2.8m for SAPADS based on an inter business
contribution to the system costs. The project authority for SAPADS, has NPV savings
of £1.7m which can be clearly identified to Horizon and other savings of £14.8m.
Without the Horizon platform to provide key data the SAPADS project would not
have been able to proceed.
The successful introduction of Local Collect would not have been possible without the
Horizon platform. There are other projects currently being developed, such as 3D
parcel labels which would not have been possible without an automated platform. Had
Horizon not gone ahead the ECCO+, APT, and ALPS systems would have required
5
In Confidence
POL00334118
POL00334118
In Confidence
replacement, based on the original costs of these projects this was an avoided capital
cost of £15m.
26. Balance of Benefits: The positives
© Operationally Horizon has proven to be reliable and robust.
e A significant number of new services and products have been developed and
deployed at a cost below that which would have been incurred for stand alone
solutions.
e The system has been very positively received by agents and their staff and Post
Office Ltd staff.
The negatives
e The cost of the service is more than the business can afford.
¢ The system architecture was driven by the core BA requirements. This architecture
is arguably more complex than that which would have been required for network
automation alone . Because of the extensive testing required, new developments.
which require additions to the infrastructure do not meet business expectations in
terms of speed to market or cost.
¢ The business is locked into a monopoly supplier on terms largely imposed by that
supplier which means the business is forced to pay more for new developments.
Risk
A measure of the success of the risk management policy was that the roll out went to plan and
the impacts identified in the business case were contained.
27. This PIR only covers the risk identified in the business case and does not cover the
risk prior to May 1999. The National Audit Office dealt extensively with the
management of risk prior to May 1999, of which it was critical. The business case
identified two high probability risks as follows:
e Delay to completion of National Roll Out
e Adverse operational impact
These and other risks were managed by the monthly Automation Programme Board.
Annex F summarises the risks and how they were mitigated.
Lessons Learnt
With such a large and complex project it is only to be expected that a considerable number of
learning points have been captured.
28. The progress made during the period covered by this PIR is in stark contrast to that
made in the earlier stages of the programme. A number of significant factors explain
this, each of which ought to be considered as key learning points. They are:-
e ICL was jointly managed by Post Office Counters Ltd and the Benefits Agency.
The decision to transfer responsibility for managing ICL to Post Office Counters
Ltd, with it, in turn, reporting on progress to the Benefits Agency, was a key factor
in getting the programme under control and delivery on track.
6
In Confidence
POL00334118
POL00334118
In Confidence
e At the point of transfer of responsibility for managing ICL to Post Office Counters
Ltd a single point of accountability for the programme within Post Office
Counters Ltd was established with Executive Committee status. The clear lines of
command and sharp delivery focus that this created were critical to successful
delivery. This single point of accountability with EC status was maintained
through the creation of Post Office Network.
e The contract was let before requirements were completely tied down. This led to
a considerable gap in ICL’s understanding of what had to be delivered which, in
turn, led to delays. Once requirements had been tied down and suitable
governance put in place, progress proceeded to plan.
e There were gaps in the approach to risk management. The extent to which parties
attempt to transfer risks between them significantly shaped subsequent
behaviours. Although Horizon was rolled out to plan and has proven robust in
operation, ICL have taken a consistently risk averse approach to developments on
the platform.
e During this phase of the project considerable Post Office resources were employed
managing the performance of ICL and its sub-contractors. This ensured that ICL
met its contractual requirements and that any gaps were aggressively followed up.
Without such investment in supplier management, it is unlikely that the progress
of roll-out and the delivery of CSR+ would have been as successful as they were.
Other learning points are detailed in Annex G and the numerous lower level learning
points will all be captured on the Lessons Learnt Intranet site.
Other Issues
[ Many of the contractual relationships were fraught with conflicting interests. ]
29.
30.
Quality and use of consultants: Consultants were used extensively throughout the
project although the services acquired were not strictly consultancy. Horizon was a
systems project on a scale often referred to as the largest systems project in Europe.
There were significant gaps in in-house capabilities to manage a project of such
complexity and the use of consultants, therefore, contributed to the success of the
project. Consultants do not bring business knowledge to the table and there were
examples, particularly around testing, where problems occurred because of this.
As the project approached contractual acceptance, the level of consultancy resource
was reviewed. The balance of effort in the project was due at this time to shift
significantly from systems build and test to implementation . This shift placed much
greater emphasis on business knowledge and the number of consultants was reduced to
below 20% of the former level. The bulk of what was delivered post May 1999 was
after this “cull” of consultants.
Relationship with ICL: The stresses and strains of repeated delays and reviews
shaped relationships as did the PFI based contractual relationship. The codified
agreement was seen by many within Post Office Ltd. as having been imposed and on
terms very much in ICL’s favour and was the source of much resentment. Senior ICL
Pathway personnel have reflected that there was a strong sense of resentment within
their organisation also. Relationships could be described as adversarial.
7
In Confidence
POL00334118
POL00334118
In Confidence
The success of National Roll out, of CSR+ implementation, other successful delivery
and joint awaydays between senior management teams all contributed to a “thawing”
of relationships. However, this could not be described as a successful co-operative
relationship. ICL expects and achieves levels of profit significantly above normal
industry expectations whilst transferring one hundred per cent of risk back to Post
Office Ltd.
Conclusion
31.
This phase of the Horizon project, particularly given the scale and complexity of both
the roll out and CSR+, was successfully delivered and this represents a considerable
achievement. However, the Horizon project was overall much delayed, resulted in a
write off of £571m, left the business burdened with operating costs it cannot sustain,
has delivered a solution which does not meet business expectations in terms of time to
market and with a supplier intent on leveraging its monopoly fully.
Recommendation
32.
That the PIRBoard note the contents of the report including the learning opportunities
and agree to this project’s closure.
Sponsor: Dave Miller Project Director: Dave Smith
Date:
March 2002
8
In Confidence
POL00334118
POL00334118
In Confidence
SUMMARY OF PROJECT INFORMATION ANNEX A
GENERAL
Project Title: Horizon
Project Sponsor(s): Dave Miller
Project Director(s)David Smith, Programme Integration Director
FINANCIAL DATA £m Original Authorised Actual
Authority Outturn
Capital 564.000 561.693
Non-Recurring Revenue 41.427 41.600
TOTAL AUTHORISED SUM 605.427 603.293
Specific DFE 8.800 sense
MAXIMUM AUTHORISED EXPENDITURE 614.227 603.293
Sunk Costs 47.841 47.841
SUM TO DETERMINE AUTHORITY LEVEL 662.068 651.134
FORECAST RESULTS (no NPVs calculated PIR.
Project is run on an Annual Equivalents basis)
Pre Tax NPV at 12% £m * -180.965 -96.188
Post Tax NPV at 12% £m * -97.802 +5,839
Pre Tax Internal Rate of Return % * -ve -ve
Post Tax Internal Rate of Return % * -ve -ve
Average ROCE % * -ve -ve
EFFECTS ON TARGETS Cumulative Cumulative Cumulative
Cash Flow £m Business 121.459 +6.248
Cash Flow £m Group n/a n/a n/a
P&L Steady State per Annum n/a n/a n/a
TOTAL P&L -608.437 -483.057
KEY TARGET DATES Original Target Outcome
Project Authority to Proceed Sep 1999 Oct 1999
1800 Office live 1 Nov 1999 8 Nov 1999
5590 Offices live 1 May 2000 3 May 2000
10680 Office live 1 Nov 2000 17 Aug 2000
15000 Offices live 31 Dec 2000 1 Dec 2000
Ready for Service Apr 2001 Jun 2001
PIR Oct 2001 Mar 2002
OTHER TARGETS Original Target Outcome
Manpower Savings/Increases Nil Nil
Quality of Service Nil Nil
Operating Performance Nil Nil
Customer Service Nil Nil
9
In Confidence
POL00334118
POL00334118
In Confidence
HORIZON NATIONAL ROLL OUT ANNEX B
PERFORMANCE AGAINST ORIGINAL AIMS AND OBJECTIVES
The main objectives as defined in the Project Initiation Document (PID) and performance
against each are:-
1. Assure through effective planning and management of the components of the
plan with ICL Pathway (ICLP) that implementation and training delivered to
outlets meets contractual and business requirements on time, cost and quality
specifications.
1.1 All milestones have been met, including the PID deliverables, and the project came in
under budget
1.2 The structure of the programme was systematic and enabled all key activities to be
undertaken at the relevant time
1.3. Roles were clear and interfaces with other functions, particularly territories, generally
good. Some confusion existed in the early days about the decommissioning and
recovery of legacy system equipment though this was soon resolved
1.4 PRINCE methodology was used throughout the project. The application of proper
project disciplines throughout the project has provided a benchmark for others to
follow.
1.5 There has been an evident reluctance by ICLP to treat Post Office Network as a valued
customer from the early days of the Horizon system. This may, in part at least, have
emanated from the commercial implications of the switch from a PFI to direct
contract. Protracted deliberations on the PFI and project slippage created a potent
situation which undoubtedly influenced behaviours.
1.6. I Whatever the cause, without active management by the project team, including the
provision of extra resource, ICL Pathway would not have performed as well, and time,
cost and quality specifications would have been in jeopardy. ICL Pathway planning
and supplier performance was variable and required detailed control mechanisms,
systems (eg. issues management, supporting satellite and mobile solution surveys) and
people (eg. Contract Compliance Managers, Product Knowledge Managers) to be put
in place by us to manage progress and performance of activity at all outlets. Training,
initially on course scheduling then latterly on course occupancy levels, was an aspect
that required extra management attention throughout. PON also had to extend its
implementation team from 10 people to 450 as it inherited added responsibilities, for
example, overseeing surveys, undertaking data migration and supporting outlets post-
migration.
1.7 Training administration attracted criticism throughout around convenience of course
locations, inflexibility of dates and course occupancy. Subpostmasters generally found
the second day of the Counter Managers course was not focused enough on balancing,
suspense accounts, lottery and bureau de change. During live trial, the decision was
taken not to extend the length of the training which would have increased PO costs.
The view at the time was that to make a real as opposed to perceived difference in
training would require more than a trivial increase.
10
In Confidence
18
1.10
2.1
2.2
2.3
2.4
2.5
POL00334118
POL00334118
In Confidence
Balancing training was perhaps the weakest link until outlets became familiar with the
system. Distribution of the Balancing With Horizon guide did much to bolster
knowledge and confidence levels and was well received.
The scheduling of installation activity in outlets was geared more towards supplier
convenience than the demands of customer service. This is more a reflection of
contractual specification than supplier performance and emanated from the
requirement to contain costs by not imposing undue constraint on installation activity.
Having said this, the impacts were generally well managed though some tensions were
encountered during the later stages of the project.
The management of suppliers and provision of contingency arrangements by ICL
Pathway was variable; third party and sub-contracted arrangements were seemingly the
weakest link (e.g. provision of communication, particularly satellite, links;
modification to counters). Greater visibility of performance management information
and contact with suppliers without transferring risks to ourselves would have been
helpful.
One of the strengths of the project was the comprehensive and structured
documentation of processes and change management.
Satisfactory roll out of Horizon to all outlets taking remedial action to secure business
performance by initiating or developing policies/frameworks as appropriate.
Activity in the outlets generally worked well though problems with ISDN line
provision and satellite permission delays eroded achievement of weekly targets on a
number of occasions. Some problems were encountered with appointments and the
standard of some survey and modification work at outlets left something to be desired
The absence of a coherent and robust outlet and product reference data system was
evident at various times during the project. Additional data verification processes had
to be introduced and databases constructed.
Policies and approaches to support roll out were developed and deployed as
appropriate. Documentation, change control and issue management were universally
acknowledged as major strengths
The management of problem outlets (eg. too expensive, no solution, refusals etc.) were
all handled sensitively with the co-operation of the NFSP. The agreement of the NFSP
to use their Discretionary Fund to finance the extra cost of automating those outlets
above business limits indicates both the advantage of partnership and seeking
inventive ways of meeting excessive costs. There was however a reluctance by the
retail line to manage these situations in a prompt and proactive manner.
Stages of the programme (e.g. live trial and acceptance) were used to good effect to
review progress and improve a number of roll out dependencies. In particular, a
deliberate decision was taken to extend the Christmas 1999 installation break. This
was a crucial factor in preparing for the rigors of relentless implementation and
enabled various preparation projects to be undertaken before national roll out
commenced. In particular the Balancing Guide To Horizon filled a real gap.
11
In Confidence
3.1
3.2
3.3
3.4
3.5
3.6
3.7
41
42
43
44
45
POL00334118
POL00334118
In Confidence
Plan and manage Post Office Network resource requirements according to the
needs of the project.
Numbers on the project team were managed according to need via a project manpower
plan which was adjusted to reflect significant changes in the ICL Pathway
implementation plan. The nature of outlet migration activity resulted in some
variations in weekly scheduled hours worked amongst field staff but these were
managed out within the life of the project.
The involvement of senior field management in the sponsorship and development of
implementation activity from an early date proved to be an effective arrangement as
too did the use of experienced outlet staff.
The organisational structure of the team was generally sound though the revised
gradings during SCS implementation, coming as they did just before national roll out
began in earnest, caused a degree of consternation and unrest. One very significant
problem that had to be overcome in the early stages of national roll out was
inconsistent business approach on temporary promotions. Many staff had been used in
such a way for over 2 years some as long as 6 or 7 years.
A notable success was the way field teams bonded and supported each other as a result
of sound leadership across the project.
Training and preparation arrangements for the CSR+ software upgrade, which were
handled by the project, generally worked well.
Equipment resources (e.g. cars, mobile phones, laptops) and hotel bookings were
logistical challenges at times but were managed efficiently and economically. Central
London proved to be a particular problem from a parking, transporting and security
viewpoint and required special arrangements to be made.
A comprehensive staff release plan and database were developed to manage the release
and redeployment of over 400 people from the project.
Confirm readiness of other business units to engage in Horizon implementation
through effective liaison and communication.
The involvement of PONEC members on the project board facilitated both a high
degree of ownership across the business and close and active co-operation in the
delivery of the project without becoming intrusive.
The creation of the Horizon Evaluation Review Forum brought together key people
from affected units across the business to co-ordinate activity and provided an
effective vehicle for brokering solutions to emerging problems. Salient pieces of work
included scenario planning, impact analysis of service failures and driving down
amounts in suspense accounts.
An influencing plan to secure the active participation of people across the business was
developed and deployed and used to inform the communications plan.
Communications was actively driven by the project manager throughout with regular
briefings via all the main internal channels, reactive external publicity and attendance
at subpostmaster forums at national and local level.
A regular dialogue was maintained with key stakeholders, including NFSP General
Secretary, CWU and CMA, territorial staff and support functions to secure their active
co-operation.
12
In Confidence
HORIZON NATIONAL ROLL OUT
In Confidence
PERFORMANCE AGAINST KEY MILESTONES
POL00334118
POL00334118
ANNEX C
Key Milestones
Plan Date
Actual Date
Commencement of National Roll Out
authorised
27 September 1999
24 September 1999
£68m contractual payment to ICL
Pathway
27 September 1999
24 September 1999
Business Case Authorisation
13 October 1999
13 October 1999
1800 offices live and £90m contractual
payment to ICL Pathway
1 November 1999
8 November 1999
(invoice date)
Review of high level Acceptance
Incidents
24 November 1999
24 November 1999
Network Business Support Centre in
place
13 December 1999
13 December 2000
Resumption of National Roll Out 24 January 2000 24 January 2000
4000 offices live 31 March 2000 20 March 2000
5590 offices live and £90m contractual 1 May 2000 3 May 2000
payment to ICL Pathway (invoice date)
Upgrade to CSR+ pilot authorised 6 September 2000 11 September 2000
Commencement of CSR+ roll out
23 October 2000
23 October 2000
10680 offices live and £90m contractual
payment to ICL Pathway
1 November 2000
17 August 2000
offices live
1 September 2000
(invoice date)
Completion of CSR+ migration
3 December 2000
15 January 2001
15000 offices live
31 December 2000
1 December 2000
Final outlet live in core roll out and
£90m contractual payment to ICL
Pathway
9 March 2001
6 March 2001
Secondary roll out of non core outlets
complete
15 June 2001
30 June 2001
13
In Confidence
POL00334118
POL00334118
In Confidence
ANNEX D
ASSESSMENT OF IMPACT ON QUALITY OF SERVICE AT OUTLETS
PO Consulting assessment stated:
Extensive monitoring and analysis of the impact on queuing performance was
undertaken during the roll out of Horizon. The analysis suggest that Horizon had a
significant adverse impact on queuing times in all outlet times in the first 3-6 months
after installation. The on-going impact is more difficult to assess, but the analysis
suggests that main offices/ (BO, MSPO & FPO) have not been adversely affected by
Horizon. It indicates that BOs may even have benefited from Horizon in terms of
queuing performance. There is some evidence to suggest on-going SPSO queuing
performance may have been adversely affected by Horizon, extent of the impact is
very difficult to quantify. There is also some evidence to suggest that the average time
to serve a customer has increased significantly in the last 3-4 years, some of this
increase is likely to be linked to Horizon.
ASSESSMENT OF IMPACT ON QUALITY OF PERFORMANCE AT OUTLETS
The level of transaction errors made at the counter is measured by the Transaction Accuracy
Measurement. The scores for period covering national roll out are scheduled in the table
below.
99.8
99.7
99.6
99.5
99.4
99.3
99.2
99.1
99
Comparable figure prior to April 2000 are not available.
Counter Transaction Accuracy Measures Since April 2000
—?- Series1
—-— Series2
=e >eespbses ke >Ees 68
BSPPSZLRPPEP ELLER EEG
§38 5 § @s 3s 25 8
<= 3 ose se 3 le)
14
In Confidence
POL00334118
POL00334118
In Confidence
ANNEX E
ICL PATHWAY PERFORMANCE TO CONTRACT SPECIFICATION
Business Impact
The business impact model is a tool which will enable the ICL Pathway Service Level
Agreements (SLA’s) to be assessed from a business wide perspective by PON. The business
impact scores will show how any SLA failures are affecting the Network and where the most
pain is felt.
The model is derived from matrices which contain scores based on the percentage failure of
the SLA against the number of offices affected. The model will allow comparison between
the SLA results on an even scale rather than identifying which SLA’s have failed by the
biggest margin.
The matrices are divided into four key areas of the business which are impacted by ICL
Pathway’s performance (customer, support, client and outlet) and from these an overall score
is produced for each SLA.
The scores are as follows:
100% shows that the SLA was achieved to the agreed minimum acceptance threshold.
* Scores above 100% shows that the SLA target has been exceeded by ICL Pathway
therefore having a positive impact on the business
* Scores below 100% show the extent to which the SLA has failed based on the impact on
the business. The historical data will enable decisions to be taken based on sustained
failure rather than reacting to ‘blips’.
e Ascore of 0% show that the failure of the SLA is totally unacceptable to PON and should
be the focus for immediate action to improve the SLA results.
e¢ N/A means that ICL Pathway are not currently reporting on these SLA’s.
Although the matrices scores have been agreed across the business it should be noted that
this is currently a learning model. As such the scores may be subject to change if the
impact is considered to have a greater or lesser impact than the results suggest.
The tables below shows the business impact for the past 20 months.
15
In Confidence
POL00334118
POL00334118
In Confidence
Business Impact Scores Table
Service Level Agreement Mar-00 I Apr-00 I May-00 I Jun-00 Jul-00 Aug-00
IRDMS reference data file delivery - day D 0 0 0 0 0 0
IAPS reference data file delivery - day D 0 0 i) 0 i) 0
IAPS data delivery - day D 0 0 i) 0 0 90
[TPS data delivery - day D 0 0 0 0 0 0
[OBCS stop list delivery - day D 0 0 0 0 0 0
IOBCS data file delivery - day D 0 0 0 0 0 0
Business incidents 0 0 0 0 0 0
Priority A - 4 hours (local) 25 12.5) 6.25 43.75 81.25 68.75
Calls answered within 40 seconds 31.25 0 12.5 68.75 68.75 81.25
Priority A - 6 hours (local) 43.75 6.25) 0 43.75 43.75 43.75
Calls abandoned 62.5) 6.25) 18.75 75 100) 100)
Priority B - 10 hours (local) 62.5) 50 56.25 81.25 87.5) 87.5
Priority B - 12 hours (remote) 62.5) 87.5) 43.75 100) 150) 200]
Priority B - 8 hours (local) 75 43.75 75 87.5 150) 150)
Level 1 calls - 5 minutes 75 62.5) 62.5 50! 150) 150)
Calls engaged 75) 75 37.5 150 150) 150)
Priority B - 24 hours (remote) 87.5) 62.5) 6.25 87.5 81.25 87.5
Level 2 calls - 45 minutes 87.5) 75 75 75 75 87.5
Site installation - quality compliance 87.5I___ 87.5 87.5[N/A #N/A #N/A
Calls answered within 20 seconds 100 50 50 50} 31.25 31.25
Level 2 calls - 30 minutes 100 87.5) 87.5 100 150) 150)
IAPS reference data file delivery - day C 100 87.5) 100) 100 100) 100)
ITraining course timeliness live delivery 100 100 87.5) 87.5 87.5] 100)
Level 1 calls - 10 minutes 100 100 100) 100 100) 100)
Cash account - 2nd line call back 100 100 100) 100 100) 100)
Training venue quality 100 100 100 100 100 100
ITraining course quality 100 100 100) 100 100) 100)
[Training courses NOT canceled 100 100 100) 100 100) 100)
ITraining course competency levels 100 100 100) 100) 100 100)
ITPS data delivery - day B 100 100 100) 100 100) 84.38
IOBCS stop list delivery - day C 100 100 100) 100 100) 100)
IOBCS data file delivery - day C 100 100 100) 100) 100 100)
Site modification - quality compliance 100 100 150IN/A #N/A #N/A
Priority A - 12 hours (remote) 100 100] #N/A 100 100) 100)
IRDMS reference data file delivery - day B 150 100 100) 150) 100) 100)
APS data delivery - day B 150 100 100) 100) 100) 84.38
IAPS data delivery - day C 150 100 100) 150 100) 150)
IRDMS reference data file delivery - day C 150 100 150 150] 150 150
APS reference data file delivery - day B 150 100 150) 150) 100) 150)
IOBCS data file delivery - day B 150 100 150) 150 100) 150)
[TPS data delivery - day C 150 150 150) 150 150) 150)
IOBCS stop list delivery - day B 150 150 150) 150) 150) 100)
(Cash account - 2nd line availability 200) 200) 200) 200 200 200)
Site installation - repair timeliness 200) 200) 200) 200] 200 200)
Priority A - 8 hours (remote) 200) 200] _#N/A 200 200 200
(Cash account scripts #N/A #N/A #N/A_ [N/A #N/A #N/A
Priority A - 6 hours (intermediate) #N/A #N/A #N/A 200I _#N/A #N/A
Priority A - 9 hours (intermediate) #N/A #N/A #N/A 87.5 87.5] 87.5
Priority B - 10 hours (intermediate) #N/A #N/A #N/A 200 200 200)
Priority B - 15 hours (intermediate) #N/A #N/A #N/A. 87.5, 87.5) 87.5
ISite modification - repair timeliness #N/A #N/A #N/A 84.4] #N/A #N/A
16
In Confidence
In Confidence
Sept 00 I Oct 00 I Nov 00 I Dec 00
IRDMS reference data file delivery - day D 0 0 0 0
APS data delivery - day D 0 0 0 0
TPS data delivery - day D 0 0 0 0
[OBCS stop list delivery - day D 0 0 0 0
[OBCS data file delivery - day D 0 0 0 0
Business incidents 0 0 0 0
Priority A - 12 hours (remote) 200 Too [31.25 25
Priority A - 8 hours (remote) 200 200_I 43.75 I 62.5
OBCS stop list delivery - day C 150 100 100__I 68.75
(Calls answered within 40 seconds 68.75 I 68.75 I 625 I 81.25
Priority A - 6 hours (local) 68.75 I 100 100 I 81.25
Priority B - 10 hours (local) 81.25 I 87.5 I 81.25 I 87.5
Priority B - 24 hours (remote) 87.5 I 875 I 81.25 I 87.5
{Training venue quality 100 87.5 I 81.25 I 87.5
Level 2 calls - 45 minutes 375 I 875 I 875 I 875
Priority A - 9 hours (intermediate) 875 I 875 I 875 I 875
Priority B - 15 hours (intermediate) 87.5 I 875 I 875 I 87.5
Priority A - 4 hours (local) 36.25 I 200 100 87.5
[APS reference data file delivery - day D 0 0 0 100
ITPS data delivery - day B 100 Too I 84.375 I 100
[Training course timeliness live delivery 87.5 100 I 87.5 100
Level 1 calls - 10 minutes 100 100 100 100
[Cash account - 2nd line call back 100 100 100 100
[Training course quality 100 100 100 100
[Training courses NOT canceled 100 100 100 100
[Training course competency levels 100 100 100 100
[OBCS data file delivery - day C 150 100 100 100
IRDMS reference data file delivery - day B 150 150 100 100
[OBCS stop list delivery - day B 150 150 100 100
IRDMS reference data file delivery - day C 150 150 150 100
APS data delivery - day B 150 150 150 100
TPS data delivery - day C 150 150 150 100
Cash account - 2nd line availability 200 200 200 100
Priority B - 8 hours (local) 87.5 100 100 150
Priority B - 12 hours (remote) 200 100 100 150
APS reference data file delivery - day C 150 100 100 150
(Calls answered within 20 seconds 18.75_I 150 150 150
Level I calls - 5 minutes 100 150 150 150
Level 2 calls - 30 minutes 150 150 150 150
[APS data delivery - day C 150 150 150 150
[OBCS data file delivery - day B 150 150 150 150
APS reference data file delivery - day B 150 150 100 200
Priority A - 6 hours (intermediate) #N/A_[_200 200 200
Priority B - 10 hours (intermediate) 200 200 200 200
Site installation - repair timeliness #N/IA_I 200 200 200
(Calls abandoned 100_I #N/A_I #N/A_[ #N/A
(Calls engaged 150__I #N/A_I #N/A_[ #N/A
Cash account scripts EN/IA_I #N/A_I_#N/A [I ANA
Site modification - quality compliance #N/A_I #N/A_I #N/A_[ ANA
Site modification - repair timeliness #N/A_[ #N/A_I #N/A_[ #N/A
Site installation - quality compliance #N/A_[ #N/A_I #N/A_[ #N/A
17
In Confidence
POL00334118
POL00334118
POL00334118
POL00334118
In Confidence
Jan 01 I Feb 01 I Mar 01 I Apr 01 I May 01 I June 01
APS data delivery - day D 0 0 0 0 0 0
APS reference data file delivery - day D 0 0 0 0 0 0
Business incidents 0 0 0 0 0 0
OBCS data file delivery - day D 0 0 0 0 0 0
‘OBCS stop list delivery - day D 0 0 0 0 0 0
RDMC reference data file delivery - day D 0 0 0 0 0 0
TPS data delivery - day D 0 0 0 0 0 0
APS reference data file delivery - day B 150 150 81.25 100 81.25 31.25
Calls answered within 40 seconds 100 81.25 87.5 150 81.25 81.25
TPS data delivery - day B 100 65.62 100 100 100 81.25
Level 2 calls - 45 minutes 87.5 87.5 87.5 87.5 87.5 87.5
Priority A - 9 hours (intermediate) 87.5 87.5 87.5 87.5 87.5 87.5
Priority B - 10 hours (local) 87.5 87.5 87.5 87.5 87.5 87.5
Priority B - 15 hours (intermediate) 87.5 87.5 87.5 87.5 87.5 87.5
Priority B - 24 hours (remote) 87.5 87.5 87.5 87.5 87.5 87.5
Priority A - 6 hours (local) 100 81.25 100 100 81.25 100
Training venue quality 87.5 87.5 87.5 100 81.25 100
APS reference data file delivery - day C 100 100 100 100 100 100
Cash account - 2nd line call back 100 100 100 100 100 100
Level 1 calls - 10 minutes 100 100 100 100 100 100
OBCS data file delivery - day C 100 100 100 100 100 100
OBCS stop list delivery - day C 100 100 100 100 100 100
Priority A - 12 hours (remote) 100 100 100 100 100 100
Training course competency levels 100 100 87.5 100 100 100
Training course quality 100 100 100 100 100 100
Training course timeliness live delivery 100 100 100 100 100 100
Training courses NOT canceled 100 100 100 100 100 100
Priority A - 4 hours (local) 150 150 150 150 150 100
RDMC reference data file delivery - day B 100 100 0 0 12.5 150
OBCS stop list delivery - day B 150 150 100 100 75 150
APS data delivery - day B 100 68.75 100 150 150 150
APS data delivery - day C 150 150 150 150 150 150
Calls answered within 20 seconds 150 150 150 150 150 150
Level I calls - 5 minutes 150 150 150 150 150 150
Level 2 calls - 30 minutes 150 150 150 150 150 150
OBCS data file delivery - day B 150 150 150 150 150 150
Priority B - 8 hours (local) 150 150 150 150 150 150
RDMC reference data file delivery - day C 150 150 100 150 150 150
TPS data delivery - day C 150 150 150 150 150 150
Cash account - 2nd line availability 200 200 200 200 200 200
Priority A - 6 hours (intermediate) 200 200 200 200 200 200
Priority A - 8 hours (remote) 200 200 200 200 200 200
Priority B - 10 hours (intermediate) 200 200 200 200 200 200
Priority B - 12 hours (remote) 150 200 200 200 200 200
Site installation - repair timeliness 200 200 200 200 200 200
18
In Confidence
In Confidence
July 01 I Aug 01 I Sept 01 Oct 01
APS data delivery - day B 150 150 150 150
APS data delivery - day C 150 150 150 150
APS data delivery - day D 0 0 0 0
APS reference data file delivery - day B 200 200 #N/A_I N/A
‘APS reference data file delivery - day C 150 150 #N/IA__[ N/A
APS reference data file delivery - day D 100 100 #N/IA__[_N/A
Business incidents 0 0 0 0
Calls answered within 20 seconds 150 150 150 150
Calls answered within 40 seconds 87.5__[ 87.5 30 87.5
Cash account - 2nd line availability 200 200 200 200
Cash account - 2nd line call back 100 100 100 100
Level 1 calls - 10 minutes 100 100 100 100
Level 1 calls - 5 minutes 150 200 200 150
Level 2 calls - 30 minutes 150 150 150 150
Level 2 calls - 45 minutes 100 87.5 75 87.5
OBCS data file delivery - day B 150 150 150 150
‘OBCS data file delivery - day C 100 100 100 100
OBCS data file delivery - day D 0 0 0 0
OBCS stop list delivery - day B 150 150 150 150
OBCS stop list delivery - day C 100 100 100 100
OBCS stop list delivery - day D 0 0 0 0
Priority A - 12 hours (remote) 0 100 100 100
Priority A - 4 hours (local) 100 150 87.5__I 81.25
Priority A - 6 hours (intermediate) 150 200 200 200
Priority A - 6 hours (local) 87.5_I 87.5 [I 81.25 30
Priority A - 8 hours (remote) 12.5 200 200 87.5
Priority A - 9 hours (intermediate) 87.5 I 87.5 87.5 87.5
Priority B - 10 hours (intermediate) 200 200 200 200
Priority B - 10 hours (local) 87.5 [I 87.5 87.5 87.5
Priority B - 12 hours (remote) 200 200 200 200
Priority B - 15 hours (intermediate) 87.5__[ 87.5 87.5 87.5
Priority B - 24 hours (remote) 875 [I 87.5 87.5 87.5
Priority B - 8 hours (local) 150 150 150 150
RDMC reference data file delivery - day B 100 100 100 100
RDMC reference data file delivery-dayC I 150 150 150 150
RDMC reference data file delivery - day D 0 0 0 0
Site installation - repair timeliness 200 I a#N/A I ANIA I NA
TPS data delivery - day B 100 100 100 150
TPS data delivery - day C 150 150 150 150
TPS data delivery - day D 0 0 0 0
Training course competency levels 100 100 100 100
Training course quality 100 100 100 100
Training course timeliness live delivery 100 100 100 100
Training courses NOT canceled 100 100 100 100
Training venue qualit 100 100 100 100
19
In Confidence
POL00334118
POL00334118
POL00334118
POL00334118
In Confidence
SUMMARY OF THE MAJOR RISKS FOR THE HORIZON PROJECT ANNEX F
PROGRAMME DESCRIPTION OF RISK POTENTIAL IMPACT COUNTER MEASURES I OWNER I L I C RISK I COMMENTS
BOARD DATE IDENTIFIED RATIN
G
NRO Scheduling of occupancy levels Significant number ofextra. I Ongoing discussion with I Douglas [3 [3 27 © Impact of
26/02/01 of training events impacts on courses needed - has created a Pathway to agree the Craik recent
14/01/01 number of courses needed, outlet potential deficit (Elm ongoing cost basis for improvement
08/12/00 readiness and user perceptions exposure) training courses and to activities to be
16/11/00 Untrained staff in live maintain adequate monitored
18/10/00 environment training performance. © Analysis of
30/09/00 ICLP model
21/08/00 prior to
17/07/00 review of
30/06/00 remedies
23/05/00
27/04/00
NRO CSR+ release implementation HOHD & NBSC help desks I Joint and cross-project_ I DonGrey [4 [3 36 © Compliance
16/11/00 impairs the NRO beat rate unable to cope with call levels planning taking place to and
18/10/00 Release problems cause handle transition and contingeney
30/09/00 interruption to full NRO beat training/changeover arrangements
21/08/00 rate problems. being
17/07/00 Roll out programme slippage developed
30/06/00 - loss of impetus and potential
23/05/00 impact on contractual
milestones
CSR+ migration and training
move into pre-Christmas
period
CSR+/Everest I Data alignment of processes to Lead times to introduce dataI_I As part of Everest release I Ruth 4 [3 40
30/09/00 support integration of PIP, OBC changes to SAP ADS plus the management, plan to Holleran
and automated distribution and need to integrate reference ensure both development
transaction processing processes and business as usual
may not meet commercial alignment of selling and domains progress issues
requirements selling activity may cause in advance of Hemel go
discontinuity live with SAP ADS (July
2001)
20
In Confidence
POL00334118
POL00334118
In Confidence
PROGRAMME I _ DESCRIPTION OF RISK POTENTIAL IMPACT COUNTER MEASURES I OWNER [LL] ©] S ] ©]. RISK I COMMENTS
BOARD DATE IDENTIFIED RATING
‘CSR+/Everest_ I HAPS capacity could be impaired I * Business continuity around I « HAPS capacity planning I Ruth 4 14 [4 [3 [44
30/09/00 if client migration and / or ICL AP client transmission may be I study has proved Holleran
Pathway fix to support five day disrupted volumes can be
transmission agreements with supported through
clients are delayed existing system until
March 2001. ICL
Pathway should have
addressed the problem
before then
NRO 30709700 I Industrial actionemanating I * Inability to sustain NRO beat I. Highriskareas(eg, I Do"Gry [4 [3 [2 [2 ]28 Developments
21/08/00 from introduction of BO rate Roll out schedule London) done by being closely
efficiency improvements would need to be revised time new entrant pay monitored with
and overall roll out introduced other PON
progmmmansay be functions
telayed * Scenario planning
done
* Contractual penalties
‘avtked © Programme
implications being
¢ Potential knock on effects managed at micro
to CSR+ migration level
CSR+ Lack of client readiness in © Delayed roll out ofservices I Meetings arranged with I A 3 fi [3 [2 [is Still need to
17/07/00 delivery of SPM and Quantum British Gas and Siemens Simpkins obiain clear
* Additional testing and I B Cragg guidance from
NB Confirmed Quantum will not pilot activities being clients
be ready for pilot see issues considered
© SPM now agreed to be
ready for pilot
© Plan being chased for
Quantum
NRO 17/07/00 I Horizon training courses do I * Additional pressure on © Continual joint Steve) 4 I3 72 [2 I 28
se ine not meet requirements of outlets, NBSC and review of training Grayston
a outlets Transaction Processing performance in place
* Contingency
arrangements being
scoped
21
In Confidence
POL00334118
POL00334118
In Confidence
PROGRAMME DESCRIPTION OF RISK POTENTIAL IMPACT COUNTER MEASURES OWNER Lic RISK COMMENTS
BOARD DATE IDENTIFIED RATING
CSR+ Risk to programme plan of Delay to OPTIP and CSR+ I ¢ CCN impact assessed Ruth 4 [2 16
23/05/00 ‘Tandem outsourcing. Plans not testing and signed off Holleran
27/04/00 yet impact assessed. Delayed business benefits
Increased costs
NRO Deferred offices creates OBC process and workload I ¢ close dialogue with PWY I Don Grey No score
27/04/0 business capability issue at tail not sustainable to limit deferrals during entered
end of roll-out. NRO
Joint activity to plan and
manage tail end
NRO Resource to support 2" & Additional pressure on resource plans in place to I Don Grey No score
27/04/00 subsequent balances does not NBSC, outlets and TP recruit auditors, trainers entered
meet requirements. ete,
scheduling in place to
allocate people on wkly
basis according to
NSBC/RNM needs
analysis,
CSR Lackiofsustained client Client products excluded © Ongoing monitoring of _ I Andrew No score
27/04/00 commitment to AP Client from E2E testing and /or go plan delivery Simpkins entered
Migration plans live of CSR+ Client representative to
Delayed business benefits become a member of the
Additional costs and AWG at appropriate
resources required to test once session
PET and UCT complete
CSR+ CSR to CSR+ planning required CSR+ release delays © Planning underwé Andrew No score
27/04/00 to assess additional support contractual roll-out timescales identify resources Simpkins entered
request for CSR to CSR+ office
migration and may not be
available
NRO Lack of continuity scenario Not identified © Keypinch points being I Don Grey No score
7/04/00 planning within project identified, scoped and entered
processes planned
22
In Confidence
POL00334118
POL00334118
In Confidence
SUMMARY OF KEY LEARNING POINTS ANNEX G
1. Information & Databases
1.1 The robustness of the inherited databases for the Project was clearly unacceptable.
Projects involving the Network of post offices need to identify key information
requirements early on and either ensure existing databases are validated and up to
date or create a new database.
2. Supplier Acceptance & System Acceptance
2a,
2.2
2.3
2.4
2.5
3.1
The key learning point for any future programme either internally based or with an
external supplier is to ensure the key requirements are:
e clean and unambiguous, with detailed performance metrics, remedies and
penalties
e have a defined business owner and a baton transfer arrangement in the case of
changes
¢ are customer/user centric i.e., there is not a presumption of operability or value
based on an intermediate level of knowledge
e shared and understood across the whole customer/supplier domain
e have a clear associated change management process which has line of sight
through the business owner down to the necessary level of detail in the
implementation plan
e have a clearly defined link with other change programmes within the business
e.g., organisational structures and associated impact assessments
The decision to continue with the project by moving to acceptance testing in a live
office environment worked well, it meant we could put pressure on them to put in
place agreed remedial actions to get the system to work. It should be understood that
contract negotiation remain open until remedial action is in place to rectify faults
identified by acceptance testing.
The business must constantly talk to the system supplier even if it is contracted by a
third party, in this case Benefits Agency. This way you can understand what is the
real state of the programme and what problems it is encountering so as you can
prepare your own risk profile. We had the impression from the Benefits Agency that
we were ready to roll and yet it took another 18 months to get it to testing in the live
environment. By talking to ICLP in 1998 we found that 80% of the time and money
had been spent on the Benefit Card and very little on EF-POS and our products.
Projects need to establish a process to monitor subcontractor performance even when
the subcontractor is not directly contracted for example subcontractors of the main
contractor. The establishment of an effective central mechanism to monitor
subcontractor performance and manage ICL Pathway infra structure progress enabled
subcontractor slippage, such as poor quality or not turning up, to be quickly rectified.
It also dealt with ICLP failures such as training when they failed to book the right
people on courses.
Senior Project staff must ensure that the supplier’s staff have the same level of
empowerment to make key decisions as the various levels of the project team.
Risk & Issue Management
The business must have the best technical people and support available, and then use
them to interrogate the supplier and get to know as much about the system as is
possible. That way you can understand what is and is not happening, what is working
etc. thus enabling you to understand the real risks and whether the solution will work
and their potential impacts on other areas.
23
In Confidence
3.2
3.3
POL00334118
POL00334118
In Confidence
The business must adopt a pragmatic approach and take into account risks to
succeeding stages. PON and the Benefit Agency took different views of the faults
that were identified by the laboratory acceptance testing. Post Office Network
formed the view that the impact of the faults was containable in 300 outlets during
live trial and that by working round the faults in the live environment it was possible
to make more progress and, therefore agreed to release authorisation.
Whereas the Benefit Agency took the view that the faults had to be corrected and re
tested before the release could be used in the live environment. It is debatable
whether the Benefit Agency were just very conservative or were looking to kill the
project off anyway. In retrospect the Post Office Network view was right. It saved
months and accelerated the process of further learning in a real environment. The key
thing was that we had identified that as it was only going into 300 outlets; if it had
not worked we would have regressed to the earlier version.
Staffing & Project Staff Support
4.1
4.2
4.3
44
4.5
4.6
4.7
There needs to be dedicated HR expertise attached to the project to develop clear
personnel policies and processes which support the project, and which are consistent
with overall business policy. All managers must adhere to these policies. The ability
to pull in and retain the required resourcing numbers and capabilities is critical to
success. This has to be supported by a clear exit policy and clarity on how people
will be paid both during and at the end of their period of involvement.
There needs to be a strong logistics management process which takes the burden off
the core implementation team. This will cover a variety of areas including the
provision of cars, laptops etc., which can be extremely demanding of time and energy
if not well handled. The principle should be to cluster key support around the core
delivery.
Projects need to consider whether sub projects require different leadership skills. The
creation of two separate projects, implementation (NRO) and CSR+ enable two
individuals with different leadership styles to be adopted. With the implementation
the project was controlled by an individual skilled at delivering change in the retail
line and driving through progress. With CSR+ an external consultant with specific
technical expertise was required to identify and resolve the technical issues. Before
the split there was evidence that the focus could move between the two areas rather
than the momentum being maintained on both.
Major projects should be run by a dedicated team which is protected from
organisational change, and need continuity with little staff turnover at management
level to develop and retain experience. It is recommended that experienced business
managers be used to deliver implementation rather than consultants. Staff from the
users areas and specialist support staff should be used in the development and
delivery stages of the project.
Give people a clear focus and responsibility and project staff will exceed
expectations. We did not necessarily select what appeared to be the best people, the
selection process looked for “reasonable” postal officers. They came from a wide
range of backgrounds and they performed well above expectation.
Ina project with large numbers of staff small sub-teams need to be established. The
Horizon team contained a large number of individuals and it was found that it needed
to be split into sub teams to create pride, mutual support, and competition. These sub
teams created their own self managed team spirit.
Projects which have a major impact on the retail line should be resourced from the
line but the project must be managed independently of the line.
24
In Confidence
POL00334118
POL00334118
In Confidence
Communications
5.1 The project should issue frequent and consistent communication which are produced
by a dedicated communication resource within the project. These staff should be as
expert in the implication of the project as the operational managers. They need to
establish a process to ensure that the impact of the project is fully explained and
understood throughout the business, not just a perception. Widespread
communication of the progress of the project was a major factor in building the
morale of a large dispersed team. The previous experience of ‘bolt on' support where
there is a necessary need to manage and translate the communication content is both
time consuming and ultimately ineffective.
Leadership & Levels of Empowerment
6.1 There has to be clear leadership standards not only within the programme itself, but
by all associated senior managers in supporting and conveying the underpinning
values of the change.
6.2 The need for strong and consistent leadership behaviours with clear roles and
responsibilities and accountabilities was paramount. On projects of this nature and
size a single team concept should be adopted to ensure consistency and avoid
communication gaps.
Testing & System Acceptance
7A Testing needs to include a volume test which gets as close to the real situation as
possible. The CSR+ migration testing performed in the laboratory by ICL was
inadequate it did not get any where near the real situation encountered in the outlets.
There were initial problems down loading the software to offices not identified in the
laboratory. The project needed to convert hundreds of outlets a night in order to
progress 12000, this was not being achieved.
7.2 The test programme needs to include a scenario “what happens if various
components fail” and the impact on other systems, not just testing if everything
works perfectly. The end-to end-process testing was deficient in that it only tested
what happened if everything worked. It did not test what happened to the processes if
things went wrong.
73 There is a need to trial all stages of the training process from cradle to grave not just
the training material. We reviewed and trialed the content of the training packages,
but we did not trail the logistics of the training process; that is organising the
invitation/confirmation at the training sessions. We trusted the ICLP contractor to
successfully organise the training sessions, but the booking system did not work and
we were asking them to enforce our rules on attendance. The impact was that when.
offices were not trained it prevented the office from “going live” and this in turn
delayed roll out.
7.4 Changes to the large number of varied outlets nationwide is very demanding. ICL
Pathway seriously under estimated the difficulty of getting offices ready, the original
13 weeks was totally impractical.
High Level Vision
8.1 There is a need to look at the wider business environment and the soft issues.
Projects are not just about “hard” issues. There is a need to think “How will the
project fit into the larger business environment, and what are the people issues likely
to be”. You need a “soft” change person to manage this aspect. Understanding these
wider issues earlier on in the project ultimately speeds the project up.
25
In Confidence
8.2
8.3
POL00334118
POL00334118
In Confidence
Business senior managers did not think wide enough because we did not understand
Service Management. The business had just accepted the Benefit Agency statement
that they would deal with Service Management. We had not understood that the
interface between the system, the staff/ agent and the customer was different to what
we were used to; it was a more difficult automated environment. We had not
appreciated that the scale of the problem was far greater until PA consultants
provided a model and sound advice.
The business must not devolve control of any of the core processes to third parties
even within your own business, because this can lead to a conflict of priorities even
within your own business. At one point Customer Management were trying to poach
all the Service Management work to the detriment of the project focus.
Policy & High Level Project Control
91
9.2
9.3
9.4
9.5
The project needs to understand the consultants that the third party is using, and what
is their role. When the Benefits Agency pulled out they had already handed over the
Programme Release Management to Authur Anderson who were just harassing
ICLP, they were not looking for solutions. The result was that ICLP had become very
defensive and held back information etc.
Consideration needs to be given as to whether very large projects need to be sub
divided into sub projects. The separation of the infra structure preparation from the
implementation (National Roll Out NRO) was fundamental to the successful
outcome. By the autumn of 1999 we already had 12000 offices prepared for
installation. However, the downside was the split into infrastructure preparation and
then NRO lengthened the overall process for individual offices from 13 to 42 weeks
and, therefore, the total programme took longer.
There is a need to build in check steps to review the original specification to ensure it
now reflects current operational needs. Key requirements were specified early on in
the project but we did not build in a step(s) to perform a further a review to check the
adequacy of the requirements. This led to problems later on in the project, eg at the
start of the project the specification was drawn up by reference only to branch offices
counters no thought was given at the time to how the kit would fit on other outlets
with the resulting problems of kit being to big during acceptance trials.
Consideration needs to be given to re-engineering products /processes, or a section of
key products/processes, at the time of automation in order assist the project and
perhaps generate further benefits. We just replicated the manual counter processes
we made no attempt to engineer any of the products or processes in order to assist in
their incorporation onto Horizon.
Projects need to establish processes to ensure that the impact of the project on other
processes is fully assessed, accountability defined, and remedial action planned
before the event. The quality of training for outlets contributed to the volume of
errors experienced by Transaction Processing through roll out and beyond. Also, the
knock on effect to Transaction Processing due to inexperience and length of learning
curve for Call Centre staff had not been anticipated. Incident and Problem
management team in particular was badly affected by this for a number of months,
resolving up to 30 calls a day from call centre staff. Network Business Support
Centre appeared slow to adopt the process for capturing information and updating the
knowledge database evident by the number of repeat requests for information.
26
In Confidence
10.
Il.
9.6
POL00334118
POL00334118
In Confidence
The executive board members need to be represented on the project boards of major
projects to demonstrate leadership and ownership to drive the project forward. The
fact that the Project Board was made up of the executive ensured support at the
appropriate level for an initiative of this scale. This was a fine example of
demonstrable leadership and the fact that it was the executive again gave visibility
and a sense of ownership to driving the project through.
Market Facing Units & Clients & Customers
10.1
10.2
10.3
10.4
10.5
Senior managers must assess any impact on clients systems and then develop a
strategy to manage this impact; do not assume they will accept the changes. We
failed to develop a strategy to manage client migration from the tandem machine.
The switch over from the tandem system to Horizon delivered no benefits to clients
and we expected them to have to fall in line. In some cases this did not happen and
caused delays. Their priorities were out of step with ours. We could have had more
sway over clients if project personnel had been involved in client contract
negotiations.
Projects need to develop processes to ensure that Market Facing Unit staff are told of
any systems changes and that they understand any impact on the client and contract
negotiations. Equally a process to ensure that contract changes negotiated by Market
Facing Units are compatible with the system specification. A problem arose with the
Horizon client interface being a 7 day a week system but clients often only expected
5 days of interfaces. A good example being British Gas Quantum and Siemens
which took months to resolve.
Projects need to be aware that switching to complex and complicated systems can
lead to increase new product development times and increases development cost
which can be a competitive disadvantage. The implementation of transaction changes
has become slower and costlier, time to market is longer. This is because changes
have to be tested far more to assess the risk to the rest of the system, and the process
is very complex and lengthy. With the Automated Payment Terminal (APT) it took
four weeks to complete a change ready for roll out, it now takes twelve weeks a
factor of three equally costs have increased by a factor of three or four. This has
given PAYPOINT who have stayed with APT technology an advantage.
Projects should not assume the client will want the system enhancement. Before we
had a network off approximately 8000 APT in 6000 outlets we now have it in all
office. British Gas did not want the Quantum transaction to be in more than 4-5000
outlets; presumably this was to ensure we did not obtain too dominant position and
undermine the viability of the other suppliers.
Projects should not assume the client will want the system enhancement. Because of
using APT technology Bill Payment was our most standardised and automated
transaction and so unlike some of our other transactions the switch to Horizon did
not produce any benefits. We automated what we already had, it was just like
expanding ECCO+ to all outlets.
Training
1.1
Training and live support was not planned sufficiently early in the life cycle of the
project and although remedial action was taken to address this I do not believe that
the outlets were as supported as they needed to be at the time of a critical business
event i.e. the completion of the outlet cash account.
27
In Confidence