POL00028149
POL00028149
7 PEV2-1 RESTRICTED - COMMERCIAL
Bringing Technology to Post Offices and Benefit Payments
REPORT ON EVALUATION OF RESPONSES TO THE SSR Pa
Author: D Selwood, M Reade, M Purchase Version: Issue 1.0
Authority: I D Selwood 12 July 1995
Reference: PEV2-1
Contents Page
GLOSSARY
1, PURPOSE.......
2. SUMMARY OF EVALUATION TEAM’S CONCLUSION:
3. THE EVALUATION PROCESS...
4. SUMMARY OF PROPOSALS...
5. EVALUATION SCORES
5.1. Overall Picture.
5.2. Sensitivity to Programme Objectives.
5.3. Pilot Programme vs Roll-out.
5.4. Facilities, Roll-out and Delivery
6. STRENGTHS AND WEAKNESSES
6.1. Introduction
62.BT
6.3. Cardlink
6.4. IBM...
6.5. Pathway
7. DISCUSSION OF RESULTS
8. COMMERCIAL EVALUATION
8.1. Risk Appraisal
8.2. Impact on Business Case
Annexes
A Evaluation Team Members
B Sensitivity Analysis Weight Changes
Cc Summary of Proposals
D Detailed Scores
E Scores for Sensitivity Analyses
12 July 1995 Page 1 of 22 Issue 1.0
POL00028149
POL00028149
PEV2-1 RESTRICTED - COMMERCIAL
GLOSSARY
ACT Automated Credit Transfer
ARTS Association of Retail Technology Standards (used in context of
standard data models for EPOS)
CAPS Customer Accounting and Payments System
Card authentication The processes and systems that support the accurate
identification of cards.
Cardholder The processes and systems that support the accurate
verification identification of individuals using a card.
CIS Counter Interface Service
CMS Card Management Service
ECCO Electronic Cash Registers at Counters (existing POCL system)
EPOS Electronic Point of Sale
ESNS Electronic Stop Notice System
IPR Intellectual Property Rights
OSS Operational Support Services
PAS Payment Authorisation Service
PIN Personal Identification Number (used in conjunction with a
card to verify the cardholder)
SSR Statement of Service Requirements
TMS Transaction Management Service
Watermark Trademark of Thorn Secure Science (used to support card
authentication)
12 July 1995 Page 2 of 22 Issue 1.0
POL00028149
POL00028149
PEV2-1 RESTRICTED - COMMERCIAL
1. PURPOSE
Li. The purpose of this report is to present to the Evaluation Board:
(a) _ the Evaluation Team’s findings on the proposals received from the five Service
Providers in response to the Statement of Service Requirement (‘SSR’) issued
on 13 April 1995; and
(b) _ their conclusions on the implications for shortlisting each Service Provider for
detailed discussions, demonstration of products and services and draft contract
negotiations with the aim of inviting best tenders.
12 The Evaluation Board are to meet on the 13 and 19 July to consider the team’s
findings and make the shortlist decision. Issue 1.0 of this report has been prepared
for input to the meeting on 13 July.
2. SUMMARY OF EVALUATION TEAM’S CONCLUSIONS
2.1. The team have concluded that the Evaluation Board should select a shortlist of two or
three from the following:
. BT
. Cardlink
. IBM
. Pathway.
22. Section 6 sets out the main strengths and weaknesses of each of these Service
Providers and Section 7 discusses the results for consideration.
3. THE EVALUATION PROCESS
3.1, All the Service Providers who received the SSR submitted proposals on 8 June, i.e.:
. BT
. Cardlink
. EDS
. IBM
. Pathway.
Bay The evaluation was conducted according to the various papers lodged with the
programme’s lawyers before responses were received:
(a) Proposal Evaluation Process (PWKP2-2, Issue 1.0, 7 June 1995);
12 July 1995 Page 3 of 22 Issue 1.0
POL00028149
POL00028149
PEV2-1 RESTRICTED - COMMERCIAL
(b) Proposal Evaluation Model (PWKP2-3, Issue 1.0, 7 June 1995);
(c) _ Proposal Evaluation Weights (PSC 12/95 (PWKP2-6), 22 May 1995);
(d) Evaluation of SSR Chapter 8 - Commercial Polices and Relationships,
(PWKP2-5, Draft D, 5 June 1995);
(e) Service Provider Risk Register (PWKP2-4, Final, 17 May 1995).
a2 The members of the evaluation and facilitation teams are listed at Annex A.
3.4. The team met to produce its initial findings from 14 to 22 June. Each proposal was
read and the principal features summarised by a ‘pre-reader’, and the five pre-readers
presented their findings to the evaluators. Each proposal was then considered in turn
by the groups responsible for evaluation under the Characteristics, Viability and
Commercial criteria. In the larger Characteristics and Viability groups a minimum of
two ‘prime readers’ were appointed for each criterion or group of criteria. They were
tasked with reading in detail Service Providers’ responses on particular criterion /
criteria, scoring the responses individually, progressively agreeing the scores and
supporting rationale among themselves and then presenting the findings to their
group for quality assurance and final agreement. Requests for clarification were
issued to Service Providers where appropriate.
335, After all proposals had been scored the markings were revisited and rationalised to
ensure consistency of marking standards across all five Service Providers.
Sensitivity Analyses
3.6. The resultant scores were then processed according to the weights agreed by the
Evaluation Team and Evaluation Board before the proposals were received. The
scores were also subjected to a number of sensitivity analyses which are described in
Annex B and summarised as follows:
(a) giving more weight to the first objective in SSR 1.2.1 (‘a fraud-free method of
...”) by effecting a 20% increase in weight to relevant criteria;
(b) giving more wei; second objective (“extending automation to POCL 's
of .”) by effecting a 20% increase in weight to relevant
criteria. This also took on board a commitment to POCL to test the effect of a
change in emphasis in the lower-level weights under 1.1.1 Service
Architecture;
(c) taking account of Mr Peaple’s request to test the impact of giving more
importance to roll-out than to pilot;
(d) _ picking out component scores for BA and POCL perspectives on facilities and
services.
12 July 1995 Page 4 of 22 Issue 1.0
POL00028149
POL00028149
PEV2-1 RESTRICTED - COMMERCIAL
Commercial Evaluation
3.7. All Commercial group members read the relevant parts of all five proposals and
scored the Service Providers. No weightings were attached to these criteria.
Service Provider Risk Register
3.8. In the course of agreeing their scores the groups also agreed their associated rationale
on the score forms and where there were significant concerns or issues recorded these
on Service Provider Risk Register submission forms. The rationales, concerns and
issues were then collated to create the Service Provider Risk Register in respect of
each proposal.
Key Strengths and Weaknesses
59. The Evaluation Team met on the 5 and 6 July to agree their overall assessments of
each proposal. The outcome from these sessions was that the team cleared the
outstanding points / clarifications on Characteristics and Viability and agreed the
conclusions and main strengths and weaknesses presented in this report.
4. SUMMARY OF PROPOSALS
4.1. A summary of the main features of the proposals is given at Annex C. It is worth
bringing out some detail of two of the most significant characteristics of the proposed
solutions, so that the Evaluation Board can put into perspective the comments made
on individual proposals in subsequent paragraphs:
(a) eet vexed
(b) the type of card authentication and cardholder verification methods used.
4.2. These are discussed below.
Centralised vs Distributed
43. There is considerable semantic debate in the IT industry as to what constitutes a
centralised or a distributed solution. However, in this particular context, the
distinction is straightforward. In a centralised solution virtually every transaction
involves data being sent from a post office to a central processing point along a
telecommunications line and then data being returned from the central point to the
post office. A distributed solution does not always necessitate such traffic since the
data to enable the transaction to be performed is often held locally at the post office.
44. There are a number of trade-offs to be made when making a choice between the two
options:
@
lependently of any central
processor (and so keep working in the event of central system malfunction);
12 July 1995 Page 5 of 22 Issue 1.0
PEV2-1
POL00028149
POL00028149
‘irate:
RESTRICTED - COMMERCIAL
4.5.
4.6.
(b) _ the work done within the Programme indicates lower data network costs for a
distributed solution;
(c) the system ge a distributed solution is more complex and,
because distributed solutions have not been around for so long, is less mature
than the equivalent for a centralised solution;
(d) dealing with ‘foreign’ encashments (i.e. encashments at a post office other than
the claimant’s normal one) is more complex for a distributed solution because
the task of finding the data is more difficult.
Card Authentication & Cardholder Verification
To ensure a fraud-free method of paying benefits it is necessary to ensure benefit is
paid to the right person. For cards, this may be thought of in two parts:
(@) Card Authentication
The mechanisms for ensuring the card is the original issued and not a
counterfeit card. Two strands support card authentication. Regieairte. .
ep card (e.g. on a magnetic stripe) which may be checked by the
system, and ‘d in terms of the visual and
tactile features (e.g. holograms, ultra-violet codes).
() Cardholder Verification
The mechanisms for confirming that the card is being used by its authorised
holder. For example a mn the candy use of a
personal identification num!
Various mechanisms are technically possible for both. These have varying degrees of
effectiveness, acceptability to the cardholders and cost. The Service Providers all
based their proposals on the use of a magnetic stripe card.
id as such
probably represents ost ption to cardholders. However experience
shows that copying the cards and forging signatures.
The significant options presented in proposals were:
(a)
This technology patented by Thorn Secure Science embeds a hidden code in
the magnetic stripe. This presents a good mechanism for card authentication as
standard card readers and writers currently on the market cannot copy these
cards.
—
The use of PINs is presented as a means of supporting cardholder verification.
However there are policy and public acceptability issues to be considered.
(©) Integrated Circuit (“smart”) cards
12 July 1995 Page 6 of 22 Issue 1.0
POL00028149
POL00028149
PEV2-1 RESTRICTED - COMMERCIAL
This is a relatively new technology that allows significant amounts of
information to be stored and updated on each card. The proposals recognised
but should be
considered for the future.
(d) Signature panel / Verification codes / Encryption
All the proposals discussed various options for making the card method of
payment more secure. These will need to be taken forward during the
Demonstrator.
4.7. Further discussion is provided against specific Service Providers in Section 6.
fi EVALUATION SCORES
§.1. Overall Picture
5.1.1. The overall scores for each Service Provider as agreed by the groups are shown in the
following table.
e third column shows
the number of criteria (out of a total of 170) scoring less than four, which is
indicative of the level of risks.
=a
Chi isticsI _ Viability I( 4
BT a 6.0 7) eg
Cardlink 5.1 16
a 4 wea Ge
IBM iy 13
Pathway Sa 5.5 13
5.1.2. Details of the scores for the main groups of criteria are given at Annex D. The above
scores for Characteristics and Viability are used to show the positioning of the
proposals on the evaluation grid in Figure 1.
Overall Positioning
8 x (°8T
2 le a Cardiink
are jas
H eM
6 x Pathway
3
3 5 7
Viability
Figure 1 Overall Positioning
12 July 1995 Page 7 of 22 Issue 1.0
POL00028149
POL00028149
PEV2-1 RESTRICTED - COMMERCIAL
5.1.3. Figure 1 shows an overall picture combining all aspects of Service Providers’
proposals. The position of the crossed lines at a score of five is not precise. The
guidance on scores set four as the minimum acceptable, so an average score of over
four would be required for a good response. Hence further inspection is needed of
the underlying scores and risks before drawing any firm conclusions.
5.2. Commercial Scores
5.2.1. The commercial group scored the proposals against seven commercial evaluation
criteria as shown in Figure 2.
Figure 2 Commercial Evaluation
5.2.2. _ The guidance for scoring the commercial evaluation criteria differs slightly from that
for Characteristics and Viability and is given in Annex D. Commercial issues for
specific Service Providers are discussed in Section 6, and Section 8 provides a
summary of the commercial findings including the assessment of the business case.
12 July 1995 Page 8 of 22 Issue 1.0
PEV2-1
POL00028149
POL00028149
ie gs
RESTRICTED - COMMERCIAL
53. Sensitivity to Programme Objectives
Tel.
ps Be
5.4.
5.4.1.
55,
Suk,
The first two sensitivity analyses identified in section 3 are concemed with
emphasising the two primary programme objectives. The results of applying these
weight changes are shown in Figure 3 (details of the scores are given in Annex E).
Emphasise Fraud-Free Payment Emphasise Other POCL Clients
7 7
% oT .
L x
= I Cardiink le x
5 ens 5
ra
eM =
xPatw ay
s 3
3 5 7 2 8 7
Figure 3 Sensitivity to Programme Objectives
positioning between proposals is largely unaffected. Similarly for component
services under Characteristics and Viability, changing the emphasis for the primary
programme objectives does not change the relative positioning of the proposals.
Pilot Programme vs Roll-out
Similar to the programme objectives, the sensitivity test putting more weight onto
roll-out and less to the pilot programme does not alter the relative positioning of the
proposals. The scores are given in Annex E.
Facilities, Roll-out and Delivery
Returning to the standard weights, differences in positioning of the proposals can be
seen considering the first level down in the model: Facilities, Roll-out and
ese are shown in Figure 4.
Facilities Delivery
Figure 4 - Facilities, Roll-out & Delivery
12 July 1995
Page 9.of 22 Issue 1.0
POL00028149
POL00028149
PEV2-1 RESTRICTED - COMMERCIAL
5.5.2. In addition, the Viability scores shown in Figure 1 include a factor for the Pilot
Programme for which the scores were s follows:
EDS
Score 22.
(b)
The strong position for
the top right quadrant;
(c) meena sacneergaa have question marks against their Roll-out
an livery, these having the worst Characteristics and the worst Viability
respectively;
(@) IBM hold a consistent position forall three areas, being high on Characteristics
and borderline on Viability. Thus their Viability needs further consideration;
proposa and all have rae as apparent from the
number oPctiiota scoring less than 4. The next sections discuss the strengths and
———E—
12 July 1995 Page 10 of 22 Issue 1.0
POL00028149
POL00028149
PEV2-1 RESTRICTED - COMMERCIAL
6. STRENGTHS AND WEAKNESSES
6.1. Introduction
6.1.1. _ After considering the scores and sensitivity analyses, it was once
‘orks done ee of ED. They ha fared worst in both Characteristics
an ility assessm substantial gaps in the information provided meant
that no score could be given in a number of areas. Parts of their proposal were felt by
the Characteristics group to be fundamentally unacceptable in business terms and
significant doubts had been raised as to the viability of what had been put forward.
None of the sensitivity analyses had improved their position significantly relative to
the other proposals. They also fared worst on the Commercial scoring giving no
reason to suppose that their approach here was so innovative that the proposal should “
be revisited to see if scope for clarification existed.
6.1.2. At its workshop on 5/6 July, the team considered the scores prota at pet si in
association with h sa
conclusions are set out below for the remaining four Service Providers.
62. gEE
62,1.
This Service Provider has
i ce > viability of the solution or of a critical analysis of th
mn. The audit facilities proposed are comprehensive. The
package is widely used in the retail environment Cs, 000 peations) but there is a
problems in areas such as:
. the transaction times at the counter would be lengthy because of the customer
having to enter the PIN on each visit to claim benefit (and also when exercising
the option to alter it at the counter);
+ the transaction authorisation process atthe counter is weak, relying totally on
an unprotected magnetic stripe and the PIN, with the system (not the clerk)
making the decision to accept the card. However this Service Provider is
prepared to accept the risk of fraudulent use of the card, and is also the only
one to accept some risk in the area of public acceptability;
12 July 1995 Page 11 of 22 Issue 1.0
POL00028149
POL00028149
PEV2-1 RESTRICTED - COMMERCIAL
. there is a high fraud risk when the system is down as the PIN cannot be used
for verification and repudiation would be a risk. Research from the finance
sector suggests that there would be a large number of phantom withdrawals at
such times;
. ili bes th because of the physical use of the
PIN and because it is suggested that the customer would be responsible for the
misuse of the card and PIN, as in the USA. The team expect that this would
require a change of legislation in the UK;
. ‘ cerned that the use of the PIN would be a move away from the
i at the counter;
+ no viable altematives are proposed to the PIN; secondary features of dynamic
signatures or photographs are mentioned but not proved;
. the proposals for proxies, in particular casual agents, are considered
unsatisfactory in a PIN-based system.
6.2.3. I Technically the team had no reason to doubt that the TMS would work but very little
evidence or rationale for the design was given. The plans for resilience and recovery
are excellent.
6.2.4. I While the proposal for OSS and audit trails has good potential, there are no plans for
the migration of the existing services and no demonstration of the requirements for
office accounting.
6.2.5. I The PAS proposed is supported by good evidence of modelling, which gave the team
confidence in the solution, and the design for financial reconciliation of benefit
payments is thorough. However, the solutions for proxies and urgent payments are
very weak, and the team felt that they would not be workable in practice. The
proposal is subject to the constraints of all the solutions proposing centralised
authorisation; and the design of PAS and CMS suggests that the two might not be
readily separable if that were required in future.
6.2.6. for the physical
implementation in post offices, b (no mention of
TMS, PAS) and to the migration of applications and dependencies of the systems.
BT’s experience and awareness of the problems of rolling out large-scale
programmes is apparent in the realistic timetable and good proposals for publicity
and programme communications. The response on service delivery is also strong and
shows their experience and a good understanding of the issues, but the plans for the
pilot are missing any indication of units of time.
BT Commercial Issues
6.2.7. I The Service Provider proposes a
0
acommercial th
12 July 1995 Page 12 of 22 Issue 1.0
POL00028149
POL00028149
e PEV2-1 RESTRICTED - COMMERCIAL
Summary of BT Proposal
6.2.8. Apart from the major weakness of the use of the PIN and the other weaknesses
above, the characteristics and viability of this proposal were considered to be
generally good (with characteristics being the stronger) although the proposal is more
thorough on the POCL aspects than on those of BA.
existing ECCO and automated payments systems. It uses an rexetny EPOS vackae
which, although from a relatively small supplier, is widely used. The Service
Provider is planning to work with the software supplier to customise it into the five
generic functions, but the extent of this development is not known at this stage.
6.3.2. __ A robust solution is proposed for TMS using. a product that is from the UK financial
arena and to be implemented in a relatively short timescale.
6.3.3. The PAS proposal shows a good understanding of the business issues. The approach
to CMS in oomerchensive, og assed iti Thr Seem Selenoo Weenie»
‘ecology leader in card security) to support card authentication, and clearly
ought out procedures for proxies and temporary token management.
6.3.4. i i d the plans do not give
confidence of having been thought through (e.g. the logistics of card production, the
fact that the design and build of the bespoke keyboards has not been a
aggressive timescale for the roll-o
arly savings if achieved, it is nce considered vi:
duce ee
recognise Programme constraints such as els) of staff.
6.3.5. The team were concerned at potential Ady Cardlink and
BA/POCL.
partner, and and POCL will be unable to react at the pace proposed. This
may limit the time for BA and POCL consideration and acceptance of the Service
Provider’s deliverables; and any delay may incur contract penalties. The approach to
contract transfer is generally over confident, seeing no reason why the contract would
need to be transferred.
6.3.6. lowever it is subject to the
a solutions proposing ap gar Also, the decision
to 5 hold all the PAS, TMS and CMS databases on the same physical platform could
make it difficult to secure separate provision of one or more of the elements later.
12 July 1995 Page 13 of 22 Issue 1.0
POL00028149
POL00028149
i PEV2-1 RESTRICTED - COMMERCIAL
Cardlink Commercial Issues
6.3.7. I It was announced at the time of submitting the proposal
\. This apy ed to have h
bid; for examp) e, he lack of availability of Barclays’ premises could necessitate
the creation oa earn a that
the consortium n formed for this procurement, there is not an established
culture or procedures for the companies working together nor any indication of how
many staff are currently employed or are expected to be employed in the future. This
is also a potential issue for the other Service Providers. As with the Pathway
proposal, there are concerns that the creation of a consortium company provides
potential for the major shareholders to limit their liability later in the procurement
process unless this is clarified before shortlisting.
6.3.8. The Service Provider proposes a y
with POCL staff as co directors I There seis bed some
advantages’ to p, stich an approach but there may also be issues around conflict of
interest and ability to develop business outside the new company.
Summary of Cardlink Proposal
6.3.9. This is a well presented proposal covering the main aspects required. If shortlisted
the main concems relate to:
. inadequacies in the current roll-out proposals;
. the apparent attitude of the Service Provider and the associated implications for
a long-term relationship;
. reducing the risk transfer by asking for sign-off of the system design before
award of contract and potentially through the creation of a consortium
company;
. lock-in through the single joint venture company proposed for all new
business, the difficulty in separating PAS, CMS and TMS, delaying the
agreement of rights to systems until transfer and proposing transfer only as a
going-concern.
6.3.10. ee ti is taken forward, strong management and negotiation will be
64. IBM
6.4.1.
6.4.2. — A particular strength is the use of Thorn Secure Science’ to
support card authentication (also used by Cardlink) which, combined with the use of
12 July 1995 Page 14 of 22 Issue 1.0
POL00028149
POL00028149
PEV2-1 RESTRICTED - COMMERCIAL
a customer signature for cardholder verification, is likely to minimise potential
adverse customer reaction.
6.4.3. While the proposal offers the facilities required, it does have weaknesses. In
particular:
(b) the TMS solution, while theoretically sensible and robust, is not proven in the
proposed configuration with the volumes expected and gives doubts on
scalability;
(c) _ there is a good approach to the desig
constraints of all the solutions proposing ceatralived authorisation.
6.4.4. The proposal for sequencing the roll-out of benefits ord, an appreciation of the
reduction of ans Sone ae
6.4.5.
Ms solution provides for a peri lem dey
tional trial and start of roll-out by some six mont In the team’s assessment
ion may prove to be more realistic than some of the SE Service Providers. The
long period for system development implies that a significant development / tailoring
exercise is required; this does not appear to be consistent with the description of the
apparently existing facilities of the packages to build the service.
6.4.6. _ The PAS, CMS and TMS central systems are to be operated from existing sites by
FDR, an organisation experienced in these areas. The proposals for fallback and
recovery to maintain the availability of the central system are well presented.
However these could be inadequate for disaster recovery due to the close geographic
location of FDR’s main and backup sites (11 miles).
IBM Commercial Issues
6.4.7. The Service Provi that a future option may be to offer the service via a
‘is brings into question the extent to which risks might
be underwritten by the parent company or shareholders.
Summary of IBM Proposal
6.4.8. ~ Outwardly the proposal is satisfactory but it lacks sufficient explanation of how the
services would be delivered. While not in itself a significant determinant for
12 July 1995 Page 15 of 22 Issue 1.0
POL00028149
POL00028149
* PEV2-1 RESTRICTED - COMMERCIAL,
shortlisting, when combined with the issues on POCL roll-out this indicates that
significant clarification should be expected during the Demonstrator during which
there is a risk that unacceptable issues may arise.
osed by this Service Provider contains a
development and contract transfer. The distributed solution ee a great deal of
resilience (e.g. a post office can work independently of the network for 2 days) and
potentially fewer problems on response times, although there would be potential
problems on foreign encashments. The modelling work done within the Programme
team indicates an expectation of lower network costs from a solution of this kind.
6.5.2. considerable part of Pathway’s proposal - including the system architecture - is
. Whilst this gave a LAeRtee of
confidence through the existence of a working system, t c ‘400
offices compared with 20,000 in the UK) led to concé V he geability of
such a large distributed system.
6.5.2 The CIS proposal shows a feareerareiier igs ae as
mentioned before, this is enhanced by the existence of a w is kind
elsewhere. However, the Service Provider ee ee
rd d client groupings, rather than a functional approach (an example is
Gunter Sem otros with POCL’s preferred generic In-pay). This
raised concerns over potential obstacles to future POCL developments.
6.5.3. I The PAS proposal is generally acceptable although more complex than a centralised
solution. One implication of this is that the PAS and TMS use the same distributed
systems, which could make it difficult to secure separate provision of these services
later. There are good proposals on payment monitoring, although proxy arrangements
are not adequately addressed.
The sals for the Pilot Programme a out
= iieeescing the constraints that will exist.
acceptance of products (because of the need for eeeicnen oO eae to mica
would result in PAS and CMS not being proven by the end of the Operational Trial.
12 July 1995 Page 16 of 22 Issue 1.0
POL00028149
POL00028149
PEV2-1 RESTRICTED - COMMERCIAL
As a consequence roll-out is proposed to start in November 1996, six months later
than requested i in the SSR. To compensate for this, the Service Provider proposes a
ure to try and maximise the
luction of order book fraud.
Pathway Commercial Issues
6.5.6.
6.5.7.
Summary of Pathway Proposal
6.5.8. This proposal has two major areas of risk, under technical and commercial. The
technical risks associated with the distributed architecture cannot be conclusively
addressed ii in the Demonstrator, though some of the aspects in scaling from current to
associated wath, cardholder erific: tion a
a rT ioneset acd land the degree ‘of tisk to be
accepted by the Service Provider would need to be explored further during the next
stage of the procurement. The delay in starting roll-out may change during
Demonstrator / Negotiation, and the ‘staged delivery’ issue might be safeguarded
contractually by linking the payment profile to the availability of user application
services.
12 July 1995 Page 17 of 22 Issue 1.0
RESTRICTED - COMMERCIAL
POL00028149
POL00028149
1. DISCUSSION OF RESULTS
YB Whilst reducing the field from five to four was very easy, the process of reducing
below that is not clear-cut even taking account of the sensitivity analyses and the
analyses of the main strengths and weaknesses. The team felt that none of the
remaining proposals was unworthy of further consideration, and that the Board
should consider the four survivors and select a shortlist of two or three. In summary
the main strengths and weaknesses identified in Section 6 are summarised in the
following table.
BT Cardlink IBM Pathway
Strengths
CIS generic & EPOS widely used Proxy & fallback Distributed
flexible TMS procedures architecture
EPOS widely used Silanes eal Understanding of five I Based on working
generics system. mie
<Ms) Design of facilities ae
PAS modelling Watermark card Detail of Pilot and
roll-out plans
PAS financial BA roll-out
reconciliation PAS, CMS, TMS to
Roll-out of CIS bern by EDR
to
Roll-out viability Lack of information I Card authentication
Customer Pilot Programme about Tillmaster Risks associated with
acceptability TMS distributed
Card authentication POCL roll-out architecture
Viability of solution Delays in start of pean oe
Proxy & urgent — Easel and generic functions
poms Foreign encashments
Roll-out of TMS & Be eae Pe
PAS Lack of detail met
Late roll-out
Transcash approach
Wwe
ds
y Uh ?
a ide a
ee
12 July 1995 Page 18 of 22 Issue 1.0
POL00028149
POL00028149
PEV2-1 RESTRICTED - COMMERCIAL
BT Cardlink IBM Pathway
0 =
Grow POCL’s BT Options on use of New business with Extend ESNS roll-
business existing assets & Service Provider’s out
Marketing of IPRs I ff (¢.g.Romec) __ I partners ~~ ermance in peak
and software jointly I “Smart store” type Extend ESNS roll- tiods linked to
approach to new out yment
business (pee pia Options for low-cost
counter equipment in
smaller post offices
ee I
Optional data Limit on liability dueI I Option for limited Limit on liability due
warehousing facility I to consortium recourse company, to consortium
in TMS Attitude & culture II PAS & CMS may hot I BA/POCL to take
rs & — may not Sign-off system be scparshie ‘d fraud risks
separable’ design before Girob
contact award wy PAS & TMS may not
Joint venture for all. be separable
new business
ae
I Delay agreement on
IPRs until contract
ination
Transfer only as a
going-concern
PAS, CMS & TM:
( may not be separable
72.
@
the incentive to the Service Providers to put maximum effort into Stage 3
would be reduced if they had a one-in-four chance of winning rather than one-
in-three or one-in-two; this is not in the best interests of the Programme;
(b)
handling negotiations with and demonstrations by four Service Providers
instead of three or two would have a major impact on the Programme Team in
Stage 3 for which the timetable is already extremely challenging, and there
would be a serious danger of the timetable being substantially elongated and/or
risk reduction being curtailed in the demonstrations and contract negotiations;
©
it is considered to be a poor return on the investment of time and money in
Stage 2 to have achieved nothing more than the elimination of one Service
Provider.
12 July 1995
Page 19 of 22
Issue 1.0
POL00028149
POL00028149
PEV2-1 RESTRICTED - COMMERCIAL
gre
derive from the proposals themselves; it ag = cay to use merely grounds of
expediency (e.g. the reduction of the burden on the Programme Team in handling
only two Service Providers and thus the increased confidence in the firmness of the
timetable).
TA. If contemplating a shortlist of two, careful consideration needs to be given to the
chances of the remaining two Service Providers reaching the stage of being invited to
submit Best Tenders so as to ensure continuing erecta and thus maximum ave
should only be Canisniplated for the negative reason that there are felt to be very
strong grounds why one of the three should not be taken further.
Ws: In reaching their decision the Board may find it helpful to address a number of
questions:
(a) what are the realistic chances of a particular Service Provider being able to
overcome the deficiencies in its proposal during Stage 3 to a degree that would
be acceptable to the Programme?
(b) are there strengths or unique elements that a Service Provider brings that are
particularly attractive to the Programme and mean that if all other things are
more or less equal that Service Provider should be given the edge over another
contender?
(c) what is the ‘risk profile’ for the Programme that would result from selecting
particular combinations of Service Providers, e.g. do all members of a potential
shortlist exhibit the same risk, so that the Programanie stands or falls on the
outcome of that particular element? ‘
76. Applying these questions to the results of the evaluation team’s findings can produce
anumber of portfolios, as follows.
BT PIN issue up-f
dds nt. If the answer is that PINs are
unacceptal Je and the PIN is such an integral part of BT” 's proposal (see 6.2.7) that the
Board do not believe an alternative can be adopted without imposing an unacceptable
delay on the Programme, then the means of reducing from four is solved
ediately. A shortlist of Cardlink, IBM and Pathway gives a portfolio of:
(a) _ one distributed solution (Pathway) and two centralised; and
(b) a measured (IBM) and an aggressive (Cardlink) pilot/roll-out approach.
12 July 1995 Page 20 of 22 Issue 1.0
POL00028149
POL00028149
PEV2-1 RESTRICTED - COMMERCIAL
78. Another way of looking at the card issue is to say that both the Watermark proposals
should be kept since this technology is regarded as the best approach to reducing
fraud. In that case the portfolio is:
(a) Cardlink (with Watermark);
(b) IBM (with Watermark);
BT or Pathway; both have weaker card authenti on ss Eee
W to cardholder verification is likely to be more. to
7.9. This portfolio would put at risk the retention of a distributed option because of
Pathway being one of the proposals at risk, and the Board would need to consider
whether, other things being more-or-less equal, it wished to give preference to
retaining the ;
Strength of Solution vs Delivery
7.10. It is also possible to consider options according to the strength of their solution
versus the delivery of it, on the basis that it is more easy to change the
implementation than the solution being implemented. The fact that the SSR gave far
more guidance on what was wanted than on how it was to be implemented gives
further support to this line of reasoning. In that case the portfolio would be:
(a) Cardlink;
(b) Pathway;
(c) BTorIBM.
7.11. The Evaluation Board are invited to consider the above portfolios as a rationale for
their decision.
8. COMMERCIAL EVALUATION
8.1. Risk Appraisal
8.1.1. All of the Service Providers indicated a willingness to take on standard risks of
system operations including implementation and ongoing performance.
12 July 1995 Page 21 of 22 Issue 1.0
POL00028149
POL00028149
* PEV2-1 RESTRICTED - COMMERCIAL
8.1.3. For other commercial risks such as existing volume (BA) and new business the
Service Providers are generally seeking additional information rior to 9 making any
firm commitments. ified S
ating om 0.9% fr BA voles, the other two simply did not provide data
but sugge: ley Wo! req .
8.1.5. Bipaates Feied from a refusal to transfer which was tempered by an interest
in joint marketing (given the PFI nature of the procurement) through to a more
negotiable stance, but these were typically caveated by issues around third party
licences.
on new business
ioetetense of new eee petites within the financial, retail and
communication areas were proposed for POCL although there were few really new
deas and they were generally at an early stage of development.
i’ 8.1.7. I Whilst most of the issues identified are thought to be capable of resolution during the
8.2. Impact on Business Case
8.2.1. Service Providers were asked to provide indicative costings but with the assurance
that these would not be used as part of the evaluation. However, it is necessary to
feed the indications into the business cases in order to test whether, on the best
evidence now available, they are still valid. If they are not, there is no case for going
on with the Programme.
8.2.2. After they had completed their work on the scoring of proposals, from which point
they took no further part in the evaluation process, two members of the Commercial
group were given access to the costings submitted by the Service Providers. Final
assurances that the business cases are still viable have not yet been given, but it is
hoped that these will be forthcoming by the time of the second Evaluation Board
meeting (19 July).
12 July 1995 Page 22 of 22 Issue 1.0