POL00027278 - Memorandum for POL Board from Lesley Sewell, Kevin Gilliland and Alisdair Cameron - Front Office Contract Award to IBM UK Ltd and Fujitsu Horizon

Evidence on official site

POL00027278

POL00027278

RESTRICTED

MEMORANDUM FOR: The Board of Post Office Limited

FROM:

Lesley Sewell, ClO & Operations Director
Kevin Gilliland, Network & Sales Director
Alisdair Cameron, CFO

SUBJECT: Front Office Contract Award
DATE: 21% May 2015
Recommendation

1. The Board is asked to:

.

Approve the award of the Front Office contract to IBM UK Ltd; and

Approve a one-year option to extend the existing Fujitsu Horizon Front Office contract
as contingency against delays, protecting continuity of operational service.

Executive Summary

2. Front Office contract award

.

The IT strategy is to substantially upgrade Post Office Limited systems and
infrastructure, replacing hardware and software across the business. This is an absolute
requirement to ensure operational continuity, as well as a substantial opportunity to
create a simpler, agile and more cost effective business.

The strategy is enabled by four substantial pieces of public procurement: End User
Computing (EUC); Network; Front Office; and Back Office. The EUC procurement has
completed, with Computacenter now in place. The Network procurement has completed
and the Board is also today being asked to approve the contract award to Verizon (the
subject of a second paper). The Back Office procurement is underway and is due for
completion in Q2.

The Front Office procurement scope includes replacing the Horizon application currently
supported by Fujitsu and moving the Common Digital Platform (CDP) from Accenture.
The revised Front Office system will be fully operational, with new software and support
services across 11,600 branches, by the end March 2017.

The Front Office procurement has followed a public sector competitive dialogue
procurement procedure, with the final tender responded to by Accenture, IBM and CSC.
Fujitsu declined to take part. The combined costs of the competing parties are
estimated at c. £15m, giving an indication of how much work and planning has already
been undertaken.

IBM scored the highest on all three key evaluation criteria areas: cost, quality and
delivery / implementation.

External legal counsel has been provided by Cameron McKenna Smith (CMS). This
included a comprehensive legal risk review supporting the recommendation to award
the contract to IBM. This was approved through the Front Office Steering Committee
and has subsequently been ratified by Group Executive. It should be noted that there is
an ongoing risk of procurement challenge from unsuccessful bidders which prior to
contract award could lead to a suspension of the process. Post contract award this
would normally be a claim for damages. The risk is low as the thresholds under which
challenges can be brought are expiring and the view from CMS is that we have followed
due process.

PAGE 1
POL00027278

POL00027278

RESTRICTED

The preferred bidder was announced on the 10" April 2015, and the Alcatel period
expired on the 20" April 2015 with no procurement challenge received.

The capital investment required to establish the new service and to exit from our
existing suppliers, Fujitsu and Accenture, is estimated at £123m (inclusive of a £19m
contingency)

The total OpEx requirement over the initial seven-year term of the contract is £73m
which represents a £20.2m per annum saving on our current like for like spend.

Three clients have contractual rights to approve the change, being the Home Office, the
DVLA and the Environment Agency. They were approached as soon as the identity of
the preferred bidder could be made public, and all have now indicated their acceptance
of IBM as the new Front Office service provider.

Following the award, the business will work intensively with IBM and Fujitsu to agree a
detailed plan, validating the approach and timetable and agreeing the change
constraints that will operate over the next two years.

Replacing the hardware (through EUC) and software within the branch represents a
significant change for the business. Mitigating IT risks requires software to be retained
in the most vanilla, out-of-the-box state as possible. This requires the business to
substantially change its ways of working. The branch roll-out will have to be very tightly
controlled to ensure that no Sparrow-style claims can be made.

Reflecting the scale of the task, a new programme director has been appointed. A
Steering Committee has been in place for several weeks, including four members of the
Group Executive. A full day workshop was conducted on 8" May 2015 with forty leaders
attending from across the business to ensure shared understanding and clear
accountabilities. A risk register is being developed; using lessons learnt from previous
problematic high profile IT implementations. A plan for expert, independent assurance is
also being formed.

We recommend that a further update is provided to the Board in October 2015.

Fujitsu extension option

.

The Fujitsu contract expires at the end of March 2017. Large IT and business change
programmes are notorious for slipping, creating trade-offs between quality and
timeliness. Even if the plan with IBM creates some contingency before March 2017 —
the initial indications are c. three-months, it would be high risk to assume that all
branches will be operating on the new system by the time Fujitsu are scheduled to exit.

Fujitsu has signalled for some time that any ongoing support beyond March 2017 would
require capital upgrades to its core and resilience data centres as well as running costs.
The resilience data centre is due to be decommissioned by September 2018, with
planning permission granted for flats.

To mitigate this risk, Fujitsu have been engaged in negotiations to provide an option to
extend the current term by a further year. This will commit Post Office to a minimum
spend of £7m (subject to final negotiations). Should we need to make use of the full
extension, the worst case view of spend is £62m, however we are working to mitigate
the risk of requiring this investment while demonstrating prudence by incorporating the
worst case within the draft 3 year plan.

We are working with Fujitsu and IBM in design workshops to secure an agreed plan
which mitigates our exposure to further capital and operational spend with Fujitsu, while
not trading off on securing operational continuity. We expect to have this in place by
September 2015. We are also considering alternative solutions such as increasing
investment in IBM in order to ease dependency and spend with Fujitsu.

PAGE 2
POL00027278
POL00027278

RESTRICTED

«Given the need to secure continuity of operational service, and the current confidence
level in our plan, the Board is requested to approve the option for a one-year extension
— see Section 6 for further detail.

Financial Metrics

4. Overview

« The Board is requested to approve awarding the contract to IBM UK Ltd and to approve
the option to extend Fujitsu for one year. The combined maximum value of this approval
is £258m (including capital and operational spend), which is provisioned within the draft

3 year plan.

IBM £61m £73m £134m
Fujitsu extension £7m nia £7m
Post Office costs £43m nia £43m
Sub -Total £111m £73m £184m
Post Office Delivery £19m nia £19m
Programme Contingency

Fujitsu worst case £19m £36m £57m
Total £149m £109m £258m

N.B. all numbers inclusive of VAT

« The worst case capital investment is £149m, of which £104m (+ £19m contingency) is
Tequired to establish the Front Office service and to exit safely from Fujitsu and
Accenture. £7m is the minimum spend committed on agreeing to the option to extend
Fujitsu, with an additional £19m provisioned within the three-year plan in the event we
need to pursue the full one-year extension and associated infrastructure refresh.

e Of the £104m required for Front Office, £61m is provisioned for payments to IBM, while
£43m is provisioned for exit services from Fujitsu and Accenture, as well as for Post
Office to deliver the business change programme required to benefit from the new IBM
solution. A further £19m is set aside for contingency.

e Previous plans had estimated a requirement for £74m, which was understated. This
was calculated on indicative pricing submissions from an earlier phase of the
procurement process and has been updated by final tender pricing. The revised
provision also recognises the scale of the change required to back office processes and
is largely driven by cost to deliver the programme end to end throughout the business
and the transition costs to support the exit from Accenture and Fujitsu. These changes
underpin the current request for £104m.

¢ The operational spend is £73m over the seven-year term of the proposed IBM contract.
This represents an additional annual benefit of £13.2m over and above the forecast
£7m per annum. Were we to run the contract for its potential duration (including 3 one-
year extensions), this would realise an additional £105m of OPEX benefit.

¢ More detail on the financial arrangements and contract terms are included in section 5
and 6 below.

PAGE 3
POL00027278

POL00027278

RESTRICTED

IBM pricing

IBM provided the lowest bid price as part of their final tender.

As set out above, provisions of £134m are required for payments to IBM. The £73m of
operational spend is committed on the premise IBM are successful in deploying the
solution as set out in the plan. The £61m of capital investment breaks down as follows;
£52.3m of milestone based payments for successful delivery; a capped liability of £5.7m
for secondary TUPE between Fujitsu or Accenture and IBM; a £3m cap to fund known
potential changes to requirements.

Known but capped risks were established to set out certain areas of requirements we
knew lacked stability, but for which we were able to secure competitively sourced
quotes for alternative solutions. For example, the current plan for refreshing the branch
counters through EUC is a dependency on being able to roll out the new Front Office
solution. There is a risk that the roll out plan might be delayed due to dependencies on
Fujitsu, therefore we asked bidders to price cap the effect of delays.

To allow comparison, the Accenture proposal price was c. £171m and the CSC
proposal price was c. £151m. Overall, each bid compared favourably with our original
expectations on reduced run costs.

The IT Transformation business case for Front Office services projected c. £7m per
annum of OPEX benefit. The IBM proposal will unlock £20.2m per annum of OPEX
benefits, representing an OPEX benefit of over £160m over the duration of the contract.
Run benefits have also been included within the three-year plan.

We have confidence in the forecast benefits as we have fixed the cost of delivery,
carried out benchmarking of our services, and received comparable offers from
Accenture and CSC.

Fujitsu extension price

We are seeking approval for a one-year contract extension for the provision of Horizon
Front Office services through to 31st March 2018 to provide contingency in the event of
delays.

The maximum price payable would be £62m: £36m of operating costs and £26m of
capital improvements to bring elements of the main and back up data centres back into
support.

The option includes a mechanism to secure a further extension. However, the disaster
recovery data centre is currently scheduled to be decommissioned in September 2018 —
see risks below.

Given the need to secure continuity of operational service, and the current confidence
level in our plan, the Board is requested to approve a one-year extension to the current
Fujitsu contract. This commits Post Office Limited to a minimum financial exposure of
£7m, representing early termination charges should we be able to negate the need for
service beyond March 2017.

The programme team will continue to explore options to reduce any further capital
investment; opportunities will become clearer as the plan matures during the design
phase. A further update will be provided to the Board in October, which will outline the
options available, the respective costs and associated risks.

The IBM Front Office contract includes financial remedies that assist with recovering our
financial exposure should IBM fail to deliver on time. For example we hold back 20% of
their implementation charges (valued at c. £9m) until we have fully accepted their
solution.

In addition, should IBM fail to deliver on time they have a capped liability to Post Office
of £36m (up to a period of 12 months of delay) in respect of Post Office Limited having

PAGE 4
POL00027278

POL00027278

RESTRICTED

to extend the Fujitsu service. IBM will suffer the delta between Fujitsu run costs and
their own proposed run costs each month (valued at approximately £1m - £2m per
month).

Previous Approvals

7. Modernisation of our Front Office IT is a key tenet of the Post Office Strategy:

.

The Board approved the IT Strategy in March 2012; this identified the need to
modernise our Horizon point of sale solution as part of procuring a new IT supply chain.

The work on Business Transformation, presented at Board in November 2014 identified
implementation of a new Front Office solution as a key requirement and enabler of
growth and cost efficiency; cost and benefits were included as part of the investment
strategy for Business Transformation.

The IT Strategy was re-presented to Board in January 2015, aligned to Business
Transformation and updated on progress and plans to secure the modernisation of our
Front Office.

The investment plan presented at Board in March 2015 included provisions to effect IT
Transformation and Front Office modernisation.

Proposal

8. Overall, the IBM proposal was the most compelling on solution, implementation and
commercial terms. Key features include;

.

.

access to the latest technology and a proven market supplier;

an enhanced replacement to the Horizon platform provided by Fujitsu, and will be
a key enabler in modernising our branch experience for our customers;

in modernising the branch experience IBM propose a collaborative and partnering
approach to working with other suppliers that replace key aspects of the current
Fujitsu contract, and also provide integration with clients;

A stable and robust core stock control and accounting platform that will mitigate
against Sparrow scenarios in the future.

A focus on mitigating operational service risk presented by the expiry of Fujitsu
and Accenture contracts in March 2017 and January 2016 respectively.

9. The optional extension with Fujitsu provides us with a prudent risk management strategy
for managing the continuity of service risk.

Risks

10. The Board should note that each risk retains ‘Controlled’ status and is within the Technology
and Operations risk appetite of Adverse.

The Horizon change constraint required during Front Office implementation may
impact other key business initiatives. (Operations Risk — Controlled)

There is less flexibility in the event of a change of scope or business strategy. This is
because on signature of the contract, Post Office is committed to a minimum level of
financial expenditure. (Financial Risk — Controlled)

PAGES
POL00027278
POL00027278

RESTRICTED

¢ Failure to fully transition to the new Front Office solution by March 2017 may result in
a delay in benefits realisation and significant additional Fujitsu cost being incurred to
maintain continuity of service. (Operational Risk / Financial risk — Controlled)

e Failure to deliver within the three-year transformation window may have significant
ramifications to the wider business transformation, and require major funding outside
the existing cost envelope. (Operational / Financial Risk — Controlled)

11. Further detail on each risk and mitigating actions is contained in Appendix 1.

12. In addition to the specific risks outlined above, the Board had previously requested that we
explore lessons learnt from other failed external major projects. The results of this analysis
are also detailed in Appendix 2.

Conclusion

13. Further to the completion of the competitive dialogue process for Front Office, and as a key
mitigation to the continuity of service risk, the Board is requested to;

e Approve the award of contract to IBM UK Ltd.

e Approve the change of terms with Fujitsu to provide an option for a one-year
extension of Horizon services.

Lesley Sewell Alisdair Cameron Kevin Gilliland
CIO & Operations Director CFO Network & Sales Director

PAGE 6
POL00027278

POL00027278

RESTRICTED

Appendix1: Key Programme and Contract Risks

Further detail on how the key programme and contract risks are being controlled is set out in this
appendix. All the identified risks are within the Technology & Operations risk appetite of adverse.

Programme risks:

1.

There is a risk that the Horizon change constraints required during Front Office
implementation may impact other key business initiatives.

(Operational Risk — Controlled)

In order to deliver the IBM solution and to exit safely from Fujitsu by March 2017, the Front
Office programme will need to limit change to scope or change to products and services
during the transition period. Any potential benefits of such changes will need to be balanced
against the risk of additional investment through the optional Fujitsu extension.

Mitigating Actions:

e The Front Office Delivery Programme is being mobilised to engage the business
effectively in the design process.

e Early communications from the programme focused on educating the business lines on
key milestones which may impact or constrain their initiatives. These constraints are
being tested with IBM to ensure their validity.

e Product profitability has come into the scope of the Front Office Delivery Programme
allowing better integration of plans.

There is a risk of less flexibility in the event of a change of scope or business strategy. This is
because on signature of the contracts, Post Office Limited is committed to minimum levels of
spend as outlined in the contract.

(Financial Risk — Controlled)

The financial planning is predicated on the fixed price attained for IBM to develop and deploy
the new Front Office solution. Changes to scope or strategy put the financial plan at risk. The
solution does offer flexibility to change scope, and IBM as a partner to Post Office accept the
need to manage change and yet ensure continuity of service, however we need to manage
this carefully.

Mitigating Actions:

e The forecasts are included in the three-year plan, any variation to scope will be subject to
robust change control.

e — The Front Office plan includes some degree of commitment to partnership and
commercial flexibility which can be leveraged.

Failure to fully transition to the new Front Office solution by March 2017 may result in a delay

in benefits realisation and significant additional Fujitsu cost being incurred to maintain

continuity of service.

(Operational Risk / Financial risk — Controlled)

As highlighted in the main body of this paper, there is significant financial exposure should we

need to extend Fujitsu beyond March 2017 and thereby invest in an infrastructure refresh for

the Horizon estate. The IT OPEX benefits will only be realised once we achieve a steady

state for the new IBM solution, and that is predicated on Fujitsu exit.

Mitigating Actions:

e — The potential additional costs are included in the three-year plan.

PAGE 7
POL00027278
POL00027278

RESTRICTED

e The scale of the proposed benefits in Front Office are significant offering a much quicker
pay back profile so delays to realising the benefits expected can be recovered.

Failure to deliver within the three-year transformation window may have significant
ramifications to the wider business transformation, and require funding outside the existing
cost envelope.

(Operational / Financial Risk — Controlled)

Delay to this programme will necessitate a longer window of constraint to change for products
and services, and also delay process efficiency opportunities within the business. If the delay
is caused by IBM our financial position is protected within the scope of that deal, though may
not account for loss of market opportunity. If the delay is caused by Post Office, it is likely we
will incur further costs in paying for exit services from incumbent suppliers over a longer
timeframe.

Mitigating Actions:
e — Senior transformational programme governance is in place to steer business and IT

programmes effectively.
e Weare also seeking to establish formal external assurance and review of the plan.

PAGE 8
POL00027278

POL00027278

RESTRICTED

Appendix 2: Learning from Industry Benchmarked Major Programme IT Failures

Industry benchmarking activity has been undertaken focussing on understanding the reasons why
major ‘IT’ programmes fail, with the aim that Post Office learn from these and ensure that any relevant
risks are flagged and controlled from the outset for subsequent major programmes.

A Gartner paper ‘AAA-Rated Project Failures — Abdication, Avoidance and Apathy’ has been used as
one source which quoted several recent examples, and this has been used to highlight the specific
lessons learned which are most relevant to the current Post Office Change Agenda. This review was
undertaken particularly in the light of the commencement of the pending Post Office Front Office
Delivery Programme, to follow Procurement award.

The identified lessons learned, in conjunction with further collateral from other external benchmarking
work, ‘best practice’ input from Post Office Risk Function and feedback from Post Office senior
management have been used to create a checklist which is currently being used as part of the initial
risk work for the Front Office Delivery Programme and will be used as part of subsequent programme
assurance activities within Post Office.

Lessons have been learned from recent experiences from some of the following organisations: HM
Revenue and Customs, NHS, Centrica, Universal Credit, with key recommendations listed below.

Key Recommendations:

+ Dramatically simplify the business processes, business rules and complexity of both the
business solution and the technical design;

+ Manage to outcomes;

+ Measure every change back to the business case;

+ — Appoint a Programme Manager who is empowered to make decisions.

The lessons learned below have already been initially assessed and are being actioned by the
programme with independent assurance being established.

Governance:

+  Unrealistic goals, schedules and budgets are set with little allowance being made for
practical, achievable outcomes

+  Underestimation and overconfidence with regard to risk

+ Governance and oversight lack accountability and responsibility

+ Poor project discipline and process controls that impede the ability to make informed
decisions

Requirements:

+ Failure to stabilise requirements, or failure to define, control and track changes to
requirements

+ Failure to understand or address non-functional and technical performance requirements
(insufficient focus)

+ Production of loose requirements definitions results in numerous, subsequent changes

Data/Management Information (Ml):

+ Production of ‘management information (MI) is left to last or not delivered at all, undermining
ability of the business to operate effectively

People/Capability:

+ Lack of business leadership, leads to a lack of focus on business change

+ Programme Manager/Director is not provided with adequate authority impacting the ability to
drive the programme ‘at pace’

+ The business and technical design decisions are delegated to junior/less experienced
colleagues who over-complicate it

PAGE 9
POL00027278
POL00027278

RESTRICTED

+ Systems Integrator (SI) is too powerful and drives its own agenda or is too weak and lacks
the capability to do the job

Technology/System Capability:

+ Technical solution is too bespoke, error prone, expensive to upgrade
+ Lack of processing power in the implemented technical solution
+  Underestimation of complexity of solution

Business Readiness, Design and Processes:

+ Failure to address or adequately consider the need to simplify complex business processes
and rules in advance

+ There is insufficient focus on the management and changes to end to end processes

+ Insufficient user engagement resulting in no ‘buy-in’ of the people who will have to use the
system & underestimation of training & support needs

+ Processes can be designed to expect too much from the people using them — not sufficiently
idiot proof, encouraging the wrong behaviours

+ There is a tendency to replicate old ways of working that aren't necessary or ‘fit for purpose’

+ The business functions change the agreed design in mid-build

PAGE 10
POL00027278
POL00027278

RESTRICTED

PAGE 1