WITN09550100 Paul Anthony Gardner - Witness Statement

Evidence on official site

WITNO09550100
WITN09550100

Witness Name: Paul Anthony Gardner
Statement No.: WITN0O9550100
Dated: 9 May 2023

POST OFFICE HORIZON IT INQUIRY

FIRST WITNESS STATEMENT OF PAUL ANTHONY GARDNER

I, MR PAUL ANTHONY GARDNER, will say as follows:

INTRODUCTION

1. I am currently employed by Fujitsu Services Limited (“Fujitsu”). I am a Service
Manager for the Triole for Services (“TfS”) Service Management Platform, where I am
involved in finance and licensing. I am currently on phased retirement and am due to
fully retire in December 2023. I have no involvement in the Horizon project in my
current role.

2. This witness statement is made to assist the Post Office Horizon IT Inquiry (the
“Inquiry”) with the questions put to me in a Rule 9 Request dated 13 April 2023 (the
“Request’). It is based on my direct knowledge of relevant matters. I was assisted in
preparing this statement by Morrison Foerster, who represent Fujitsu in the Inquiry.

3. The topics in the Request relate to events that took place at least 15 years ago. I
have tried to remember matters as best I can but have not always been able to do so.

4. Where I have seen documents relevant to the Inquiry’s Request for the purpose of
preparing this statement or where I have referred to documents, these documents are

referred to using the Inquiry’s URNs and are listed in the index accompanying this

Page 1 of 28
WITNO09550100
WITN09550100

statement. I understand that some of the documents referenced in my statement were
not previously provided to the Inquiry. These documents are from my own records or
are ones that Fujitsu and its external legal advisers showed me in the course of
preparing this statement so as to assist my recollections. They have been provided to
the Inquiry alongside my statement.

BACKGROUND

5. I graduated from Coventry University with a BSc (Honours) in Applied Biology in 1984.

6. Following graduation, I worked in various management and IT roles at McDonald’s
Restaurants Limited (“McDonalds”). By 1996, and while still employed by McDonalds,
I was working as a consultant supporting Fujitsu’s provision of the McDonalds’ IT
service desk."

7. I joined Fujitsu in December 1998. My initial role was as a Team Manger on the same
McDonalds IT helpdesk I had been working on previously. In 1999, I was promoted to
Helpdesk Operations Manager. I spent several years in that role, working on various
helpdesks Fujitsu provided to large retail clients.

8. In November 2005, I joined the Horizon Service Desk (the “Helpdesk”) as Operations
Manager. Prior to joining the Helpdesk, I had no involvement with the Horizon project
or the Post Office account. I continued in the role of Operations Manager on the
Helpdesk until May 2008.

9. Within Fujitsu at the time, the Core Services group provided various services that were
required on multiple customer accounts. The Helpdesk was one such service. I, and

the individuals I managed, fell within the Core Services organisational structure. My

1 At the time, Fujitsu was known as “ICL”.

Page 2 of 28
WITNO09550100
WITN09550100

manager at the time was Helen Robinson (Operations Manager, Service Desks —

Customer Services). She managed several operations managers on various different

accounts. On a day-to-day basis, however, I worked alongside and was accountable

to the Post Office account's Helpdesk Service Delivery Manager (the “SDM”). During

my time on the Helpdesk, the SDM I worked with most frequently was Liz Melrose.
10.My role as Operations Manager on the Helpdesk involved:

a. Overseeing and supporting a team of approximately 80 team managers and
agents, who delivered first and second line support to Horizon users that
called the Helpdesk. As part of this responsibility, I (i) directly managed
about five team managers, who in turn each managed about 10 to 15 first
and second line support staff, and (ii) had overall responsibility for the
training and recruitment of Helpdesk staff and improving existing Helpdesk
procedures.

b. Ensuring that the contractual Service Level Targets (“SLTs”) for customer
service and support to subpostmasters were met.

c. Managing the provision of technical support for rollouts of new hardware or
new software releases to branches.

d. Supporting the SDM on liaising with the Post Office regarding Helpdesk
performance and processes, including feeding into the monthly Service
Review Books (“SRBs’”) received by Post Office. While interfacing with the
Post Office was primarily the SDM'’s role, I would sometimes liaise directly

with Post Office staff and attend meetings with them alongside the SDM.

Page 3 of 28
WITNO09550100
WITN09550100

11.After leaving the Helpdesk in May 2008, I remained at Fujitsu as a Toolset
Implementation Consultant. My role shifted from the management of helpdesks to an
internal technical services and financial role. I have, since leaving the Helpdesk (i)
supported the set-up of new accounts on the TfS call logging system, (ii) managed the
decommissioning of legacy call logging systems on various Fujitsu accounts and their
transition to TFS, (iii) led reviews on how Fujitsu could better utilise toolsets such as
telephony Individual Voice Response (“IVR”) to manage helpdesk agent performance
across the business, (iv) managed aspects of the decommissioning of TfS and
Fujitsu’s transition to the newer ServiceNow platform, and (v) managed the revenue
recovery for Fujitsu’s provision of helpdesk toolset services and licenses to various
global customers.

12. Due to my familiarity with the call logging systems used at various points on the Post
Office account, I would occasionally be asked how to properly extract call log data
from PowerHelp or TfS (as applicable) by individuals working on the Post Office
account, even after I had left the Helpdesk. I would not conduct the extractions myself
but, rather, would give guidance on how they were to be done. I had understood that
the extractions may have been to support the Post Office in investigations of
subpostmasters. Save for this, and giving my successor on the Helpdesk some ad
hoc assistance during our handover period, I do not recall having any involvement
with the Post Office account or the Horizon system after I left the Helpdesk in May
2008.

HELPDESK OPERATIONS

Page 4 of 28
WITNO09550100
WITN09550100

13.As I note above, I worked on the Helpdesk between November 2005 and May 2008.
The Inquiry has asked me to consider a number of documents that describe Helpdesk
processes, and I have done so. These are: ICL Pathway/POCL Interface Agreement
for the NBSC and HSH Interface, Version 2.0, dated 13 July 2000 (FUJ00080406),
POA Customer Service Incident Management Process Details, Version 3.0, dated 23
March 2005 (FUJ00079939), End to End Support Process, Operational Level
Agreement, Version 2.0, dated 17 June 2003 (FUJ00079897), Service Level Targets
for Horizon Services, Version 3.0, dated 19 August 2005 (FUJ00001966), Horizon
Service Desk Joint Working Document, Version 1.0, dated 27 July 2007
(FUJ00080459), Call Enquiry Matrix and Incident Prioritisation - Reference Data,
Version 1.0, dated 23 April 2007 (FUJ00080498), Call Enquiry Matrix and Incident
Prioritisation - Software, Version 1.0, dated 23 April 2007 (FUJ00080499), Call
Enquiry Matrix and Incident Prioritisation - HSD, Version 1.0, dated 17 April 2007
(FUJ00080501), POA Customer Service Major Incident Process, Version 3.0, dated 7
July 2009 (FUJ00080071), RMGA End-to-End Customer Complaints Procedure —
Joint Working Document, Version 2.0, dated 24 December 2007 (FUJ00080034),
RMGA Customer Service Problem Management Process, Version 2.0, dated 22 April
2008 (FUJ00080043), and Processes - ID 408501 - Engineer Refused Access
Process (undated) (FUJ00138733). Some of these documents appear to have been
prepared outside the period I was involved with the Helpdesk. Those documents are:
FUJ00080406, FUJ00079939, FUJ00079897, FUJ00001966, and FUJ00080071.
While I am generally familiar with the processes described in these documents, and

some of them have assisted me in preparing this statement, I am not in a position to

Page 5 of 28
WITNO09550100
WITN09550100

comment in detail on Helpdesk procedures or operations during periods when I was
not working on the Helpdesk. Unless I expressly refer to a different time period, my
observations in this statement are based on the period I was working on the Helpdesk.
Purpose of the Helpdesk

14. The purpose of the Helpdesk was to provide initial technical support by phone to Post
Office branches, including end to end management of incidents reported by callers to
the Helpdesk. The support provided by the Helpdesk covered any part of the counter
equipment installed at a Post Office branch and the associated software. First line and
second line support agents worked within the Helpdesk. Service Delivery Units
(‘SDUs’), including, for example, the SSC, received escalations from the Helpdesk
when first and second line agents were not able to resolve a reported incident. These
units interfaced with the Helpdesk but did not form part of it.

Roles on and management of the Helpdesk

15.As noted above, I worked alongside the Helpdesk Service Manager in relation to the
management of the Helpdesk.
16. The main Helpdesk roles were as follows:

i. First Line Agent. First Line Agents were the initial point of contact for anyone
calling the Helpdesk. In summary, their primary function was to take incoming
telephone calls, record relevant information about the incident the caller was
reporting, and either resolve the incident or pass it on to another team to
progress/resolve.

ii. Second Line Agent. Second Line Agents were more senior and experienced

Helpdesk staff, with greater technical knowledge than First Line Agents.

Page 6 of 28
WITNO09550100
WITN09550100

Second Line Agents provided support to teams of First Line Agents allocated
to them. They would directly give advice to First Line Agents during calls if First
Line Agents were unsure how to resolve an incident. Alternatively, such calls
could also be passed on from a First Line Agent to a Second Line Agent.
Second Line Agents were also responsible for training First Line Agents and
monitoring calls taken by First Line Agents so as to ensure procedures were
being followed and correct advice was being given.

Team Manager. Team Managers each managed a team of about 10 to 15 First
and Second Line Agents. Team Managers had overall responsibility for the
telephony performance of First and Second Line Agents, and the quality of call
logging, incident resolution, and training by those in their teams. While I worked
on the Helpdesk, I recall there were usually about five Team Managers, who
each reported directly to me.

Incident Management Team (“IMT”). The IMT was a group of experienced

agents, led by a Team Manager, who mainly dealt with major incidents when
they arose, though they were deployed to assist with standard frontline calls in
periods of high call demand. They had a number of roles. These included:
i. Identifying and reporting on trends in incoming calls.
ii. Organising the Helpdesk’s response to call storms, which is a
situation where many calls are made into the Helpdesk reporting the
same incident. This could occur, for instance, after the rollout of a

new software release. In these instances, the team would set up a

Page 7 of 28
WITNO09550100
WITN09550100

master ticket and make sure first line support all assign their calls to
this master ticket to avoid duplication.

iii. Monitoring the desk’s SLT compliance for open incidents, and
providing an escalation route for Helpdesk Major incidents into the
Service Management Team.

iv. Managing complaints about Fujitsu or Horizon from end users (as set
out in more detail in FUJ00080034).

v. Specifically supporting problems at Post Office “Hot Sites” or “VIP
Sites”. From my recollection, “Hot Sites” were defined by Post Office
and usually included newly opened sites or single counter branches
(where any technical issues were more likely to have a significant
impact on the branch's ability to trade). VIP Sites were also defined
by the Post Office, and included “flagship” sites or those that might
be regularly visited by Post Office senior management.

v. Communications Management Team (“CMT”). The CMT was a group of

experienced agents, led by a Team Manager, who specifically monitored and
supported on network issues faced by branches and any issue that prevented
a branch from communicating with the central servers. Like the IMT, the CMT
was also deployed to assist with standard frontline calls in periods of high call

demand.
17.Having reviewed the Post Office Ltd / Fujitsu Services Retention of the Incident and
Communications Management Teams document dated 4 April 2006

(WITN09550101), I recall that the IMT and CMT teams were provided at additional

Page 8 of 28
WITNO09550100
WITN09550100

cost to the Post Office, and they were retained during my time on the Helpdesk as
they were considered to benefit the Post Office branch estate.

Recruitment for Helpdesk Roles

18.1 reviewed various Helpdesk skill matrixes — specifically the Team Manager HSH Skills
Matrix (undated) (WITN095500102), HSD Customer Support Technical Skills Matrix
dated 3 November 2005 (WITN095500103), and the D1 — D4 Core Skills Matrix for
PSE (undated) (WITN095500104) — to assist me in recollecting some of the
observations I make in this section. I refer to these skills matrixes below as they set
out the performance framework for the key Helpdesk roles. I do not recall if each of
these skills matrixes were in force throughout the entire period that I worked on the
Helpdesk. There may have been changes in format or structure over time. However,
I believe they do reflect the substance of the performance framework that applied
during my tenure on the Helpdesk.

19.First Line Agents were generally recruited using external employment agencies. Most
individuals were usually initially employed on a temporary basis, but could later be
offered permanent positions. The performance framework for First Line Agents is set
out in WITN095500103. The minimum requirements when recruiting for the First Line
Agent role were good previous customer service experience and a basic knowledge
of IT. Bespoke training on the Horizon system would be provided to any new agent,
and so prior knowledge of the system was not required. After recruitment, First Line
Agents would continue to be assessed during a probationary period, and I recall
individuals were let go if they were not performing adequately. My recollection is that,

at the time I worked on the Helpdesk, the standard of IT knowledge of the staff within

Page 9 of 28
WITNO09550100
WITN09550100

a high proportion of Post Office branches was generally considered very low. As a
result, many calls into the Helpdesk could be resolved by someone with only basic IT
knowledge. As such, while First Line Agents were vetted for IT skills, good customer
service skills were generally considered to be more important than advanced IT skills.

20.Second Line/IMT Agents were generally promoted within the Helpdesk from
experienced First Line Agents who had gained strong knowledge of the Horizon
system and had begun to show the relevant skills required for the role as defined in
WITNO095500104.

21. Team Managers were generally promoted within the Helpdesk, transferred from other
Helpdesks in Fujitsu or, occasionally, joined from the Fujitsu graduate recruitment
programme. Potential candidates for Team Manager roles were usually identified by
an “Ongoing Management Review” process. This involved operations managers
across various Fujitsu helpdesks identifying potential candidates from their teams for
management roles. An example of how the process applied can be seen in the OMR
for Commercial & Public Sector Call Centres, Customer Services slide deck dated
September 2005 (WITN09550105). In order to be selected for the Team Manager
role on the Helpdesk, a candidate would need to be assessed by existing Team
Managers as performing at Level D4 of the Second Line (PSE) Skills Matrix
(WITN095500104) and had to successfully pass a Team Manager Assessment
Centre. At the Assessment Centre, individuals would undertake various exercises and
would be assessed by me and other Operations Managers on customer service,
technical knowledge, and people management. Team Managers were expected to

perform in accordance with WITN095500102.

Page 10 of 28
WITNO09550100
WITN09550100

Training for Helpdesk Roles

22.Training for new First Line Agents involved about two weeks of “classroom style”
training in the dedicated Helpdesk training room. Training would be conducted by
Second Line Agents and, to the best of my knowledge, would cover the use of call
logging software, how to categorise and escalate calls, how to use available resources
to resolve calls, and how to resolve common issues that came up on the Helpdesk,
amongst other things. Trainees had access to fully working Horizon equipment that
would replicate the Post Office branch environment. Where possible, the training
would also include a visit to a Post Office branch site see how the branch operated in
practice. During the course of this training, trainees were required to complete tests
on the course material and would have needed to pass in order to be deployed as a
First Line Agent.

23. This would then be followed by about a further two weeks “on the job” training where
the First Line Agent would take calls under the supervision of a Second Line Agent.
Second Line Agents would listen in and record a sample of calls taken by the First
Line Agent during this period. Calls would be scored and reviewed with the First Line
Agent. We referred to this process as “call coaching”.

24.Following the initial “on the job” training period, First Line Agents were routinely
assessed by means of call coaching and regular performance reviews relating to
defined objectives and call statistics. A call coaching database was maintained to
monitor agent performance across the desk. Overall scores were monitored on an
ongoing basis. First Line Agents were required to meet the standards in the First Line

Agent matrix.

Page 11 of 28
WITNO09550100
WITN09550100

Instructions to Helpdesk agents and resources available to resolve calls

25.All Helpdesk roles had documented processes to follow in terms of receiving calls,
categorising them, attempting to resolve them and escalating them. These instructions
were covered during initial training. That information may have been available to
agents after completing training, though I do not recall this in detail or how agents
would have accessed those materials. If instructions needed to be changed or
supplemented, these would usually be communicated by me or the team managers
using email or at team meetings.

26.Those working on the Helpdesk had access to an online knowledge base which
served as a repository of information that agents could search to assist in the
resolution of calls. I do not clearly recall the name of the database but, having reviewed
FUJ00079939 and the Horizon Service Desk Operational Development Plan Current
Activities document (WITN09550106), I believe it might, at various points, have been
called HSH One or HSD One. For convenience, in this statement, I will refer to the
database as “HSH One”. The HSH One database held call scripts, articles containing
instructions on how to resolve common problems, and also contained the KELs which
had been produced by the SSC.

27.My recollection is that call scripts would at least cover the types of points set out at
Section 4 of FUJ00080459 - e.g. how to answer a call, what to initially ask a caller,
and the mandatory information an agent needed to record on the call log. Having
reviewed FUJ00079939, I recall there may have been scripts on how to deal with

specific problems or scenarios. However, I do not have a clear recollection of these.

Page 12 of 28
WITNO09550100
WITN09550100

28.The information on HSH One was generally prepared by Second Line Agents and
Team Managers, but some information will have been provided by SDUs such as the
SSC or the Engineering team.

29.The call history for all branches would also be available on the call logging software
(PowerHelp or TfS as applicable) if required.

30.My understanding was that most Helpdesk agents would download or copy onto their
own desktops the instructions for resolving most common issues, so they would have
them easily to hand. As such, most agents would not be searching HSH One on every
call they took.

31.Beyond what I have noted above, I only have a vague recollection about the scripts
or other resources available on HSH One. Much of the material had already been
prepared before I joined the Helpdesk, and I do not recall having any involvement in
drafting scripts or other resources to assist Helpdesk staff with resolving problems on
calls.

Procedure for resolving calls

32.As an initial point, I would observe that the procedures set out at pages 11 to 20 of
FUJ00079939 and pages 13 to 18 of FUJ00080459 largely reflect my recollection of
the procedure Helpdesk staff would follow in order to resolve a call.

33.The Helpdesk received calls on a variety of issues covering a very wide range of
scenarios. It would not be possible for me to set out in this statement the procedure
that would be followed for every scenario a Helpdesk agent might have encountered.
However, I have set out below the typical process that would have occurred for new

incidents received by the Helpdesk:

Page 13 of 28
WITNO09550100
WITN09550100

. When a call was received by a First Line Agent, the agent would first greet
the caller and take the caller's branch FAD code and other security details
to validate the call.

. Once validation had been successfully completed, the agent would seek to
verify if the call related to a new or existing issue, based on the caller's
description of the problem and the branch’s records on the applicable call
logging system.

. If the call was about a new issue, the agent would open a new ticket for the
call on the call logging system.

. The agent would take more information about the problem, and seek to
classify the call into one of four categories:

i. Incident. An incident was any hardware or software issue with
Horizon.

ii. Advice and guidance. Advice and guidance calls were calls seeking
information on how to use the Horizon system (i.e. they did not report
any problem in the functioning of Horizon).

iii, Out of scope calls. Out of scope calls were calls relating to issues
that did not fall within the scope of the IT services Fujitsu was
providing to Post Office. Most often, these related to Post Office
business issues (e.g. the correct procedure for selling a stamp).

iv. Quality. Quality calls related to the quality of service the caller had
experienced in a prior interaction with the Helpdesk or other Fujitsu

staff.

Page 14 of 28
WITNO09550100
WITN09550100

e. If a call was classified as an “advice and guidance” or “out of scope” call,
the Helpdesk would not proceed with the call and would advise the caller to
contact the NBSC, as described at paragraph 4.1.2 of FUJ00080459.
During my tenure on the Helpdesk, we also referred to calls of this nature
as “inappropriate calls”, as they were calls that were for the NBSC and often
related to issues Post Office was best placed to resolve. The Inquiry has
specifically asked me how calls about accounting discrepancies were
handled. As noted in FUJ00080459, which is a “Horizon Service Desk Joint
Working Document” approved by both Post Office and Fujitsu, these were
treated as “inappropriate calls” and were directed to the NBSC. I was an
approval authority for this document, which was also reviewed by others in
Fujitsu and the Post Office. I provide further details below on how these
types of calls were dealt with while I worked on the Helpdesk.

f. If a call was classified as an “incident”, the First Line Agent would then
proceed to classify the incident’s severity as Low, Medium, Major, or
Critical, according to the criteria set out on page 15 of FUJ00079939. If an
incident was considered “Major” or “Critical” it would usually be escalated
to the IMT as well as the SDM.

g. If the incident was classified as “Low” or “Medium”, the First Line Agent
would then attempt to resolve the issue based on their own knowledge or
the resources they had available on HSH One.

h. If they were unable to do so, they might seek guidance from a Second Line

Agent or pass the call on to them.

Page 15 of 28
WITNO09550100
WITN09550100

i. Having reviewed FUJ00080498, FUJ00080499, and FUJ00080501, I recall
that while taking details about a call and attempting to fix the relevant
problem, agents would also be required to categorise the problem based on
the type of problem, the cause of the problem and (if a fix was successful)
the repair that was applied. The available categories were pre-defined by
senior Helpdesk staff or the SDUs, and available for selection on the call
logging system. These would be modified from time to time as new issues
were identified. If a Helpdesk agent was unable to diagnose the problem or
fix it, he/she may not be able to complete the categorisation and that would
be done when the call was escalated and resolved by an SDU.

j. If the Helpdesk agents were unable to resolve the incident, they would pass
the ticket on to the relevant SDU (e.g. Engineering for hardware issues, the
SSC for software issues). Different processes were used for transferring
calls to an SDU depending on which SDU was relevant. For SDUs where
there was a high volume of tickets transferred (such as the SSC or
Engineering), there was a link (known as an “OTI link”) set up between the
Helpdesk’s call logging system and the ticket management system the SDU
used (Peak in the case of the SSC, and CRiSP in the case of Engineering).
The OTI link allowed information from the call logging system to be
automatically copied over into the relevant SDU’s ticket management
system. For SDUs where there were a much lower volume of calls

transferred, a manual email process was used.

Page 16 of 28
WITNO09550100
WITN09550100

k. Once an SDU had resolved the incident, the call was required to be passed
back to the Helpdesk and a First or Second Line Agent was meant to call
the relevant branch to confirm the incident had been resolved. In practice,
SDUs often obtained the caller’s agreement to close the call when their
issue was resolved, and the Helpdesk would not call the branch back. This
was because staff at branches often found it inconvenient to receive calls
on issues that had already been resolved. I also recall that, in some cases,
IMT may have called branches back regarding call closures, though I cannot
clearly recollect the situations in which this happened.

I. Once an incident was resolved, the call would be closed by the Helpdesk.
As noted above this was usually done with the caller's agreement.
Occasionally calls may not have been closed by the Helpdesk, but directly
by the SDU via the OT] link. However, I do not recall how common this was.

Calls about accounting discrepancies

34. The Inquiry has asked me how calls from branches about unexplained accounting
discrepancies would be classified in the table set out on page 11 of the Fujitsu
Services Post Office Account - Service Review Book (February 2007) dated 14 March
2007 (FUJ00083429). As I note above, these would be classified as “inappropriate
calls” for the purpose of FUJ00083429 and were to be directed to the NBSC.

35.During my time on the Helpdesk, it was not generally the Helpdesk’s role to resolve
unexplained discrepancies in branch accounts. Where a call related to an accounting

discrepancy and if the relevant issue was covered by a KEL, I would expect Helpdesk

Page 17 of 28
WITNO09550100
WITN09550100

agents to apply the instructions in that KEL. Aside from that, discrepancies in branch
accounts would be referred to the NBSC in accordance with FUJ00080459.

36.While accounting discrepancy calls would initially have been referred to the NBSC, it
was open to the NBSC to refer the call back to the Helpdesk if the discrepancy
remained following the conclusion of their support process or if it appeared to be
attributable to a technical issue. This is noted at footnote 1 of FUJ00080459. Where
calls were referred back to the Helpdesk, Helpdesk agents would follow usual call
handling procedures, including referring the incident to the relevant SDU if they were
unable to resolve the problem.

37.In practice, some Helpdesk agents may have provided initial technical support for a
cash discrepancy, such as rebooting the Horizon counter at the branch, as this
sometimes cleared discrepancies. However, if the discrepancy still remained, then the
call would be referred to the NBSC helpdesk, who I understood would attempt to
resolve the discrepancy.

38. The Inquiry has referred me to the First Witness Statement of Amandeep Singh dated
13 January 2023 (WITNO6660100). I am aware from reading Mr Singh’s statement
and watching his oral evidence online that, prior to my joining the Helpdesk, support
for accounting discrepancies was provided by members of the Helpdesk to callers. By
the time I joined the Helpdesk, that position had changed and calls regarding
accounting discrepancies were primarily dealt with by the NBSC. As I was not involved
in the discussions around the decision to change the procedure, I cannot provide
further detail on it.

Monitoring quality of assistance offered by the Helpdesk

Page 18 of 28
WITNO09550100
WITN09550100

39. The quality of assistance provided by the Helpdesk was monitored by various means:
a. First Line Agents were monitored via call coaching. I believe this occurred
at least quarterly, as call coaching results fed into First Line Agents’
quarterly reviews. In addition, IVR statistics were tracked for each agent and
used to review individual agent performance for metrics such as speed of
answering calls and availability to answer calls.

b. The performance of the desk as a whole was monitored via daily reviews of
the desk’s compliance with the SLTs. My recollection is that the SLTs that
applied during my time on Helpdesk were of the nature set out in in the
“Service Levels for Support Desk Services” section on page 8 and 9 of
FUJ00001966. I believe the Post Office received a weekly update on SLT
compliance from the Helpdesk Service Manager, and SLT figures also
formed part of monthly Service Review Books which reported various
metrics relevant to Helpdesk performance.

c. In addition to the SLTs, I believe a number of other metrics were tracked by
the IMT, including the number of complaints received regarding Helpdesk
performance, and specific updates on VIP\Hot sites.

BUGS, ERRORS AND DEFECTS

40.1 have been asked by the Inquiry whether, during my time working on the Helpdesk, I
was aware of any bugs, errors or defects with the Horizon system.
41.1 was aware that the Horizon system did have software errors. These errors resulted

in calls to the Helpdesk, with relevant workarounds or fixes being recorded in KELs.

Page 19 of 28
WITNO09550100
WITN09550100

However, in my role, I did not have detailed knowledge of specific software errors. I
was not directly involved in diagnosing or resolving software errors on the system.
42.While the Helpdesk staff could provide solutions to known software issues based on
the KELs, they were not responsible for diagnosing and identifying solutions to new
software issues. If a software incident could not be resolved by consulting the KELs,
the appropriate process was for the ticket to be passed to the SSC for analysis and

resolution.

EVIDENCE OF SOME SUBPOSTMASTERS REGARDING THEIR EXPERIENCES

WHEN CALLING THE HELPDESK

43.The Inquiry has informed me that it has heard evidence from a number of
subpostmasters who allege that, when they called the Helpdesk with a technical issue,
they were told they were the only one experiencing the problem they had called about.

44.To the best of my knowledge, during my time on the Helpdesk, Helpdesk agents were
never instructed to inform callers that they were the only person experiencing a
particular issue. I do not recall anyone giving such advice to a caller. If this was
overheard by a Second Line Agent or Team Manager, or picked up by the call
coaching process, I would have expected it to be corrected.

EMAIL OF 7 NOVEMBER 2007 (FUJ00138214) AND SPREADSHEET ATTACHED TO

FUJ00138214 (FUJ00138215)

45. The Inquiry has referred me to an email from me to John Casey, Paul Hailey, Sarah
Hill and others, subject heading “FW: Process issues at HSD”, dated 7 November
2007 (FUJ00138214), which contains a chain of emails. The first email dated 2

November 2007 is from Mik Peach to Liz Melrose, and contains a number of criticisms

Page 20 of 28
WITNO09550100
WITN09550100

of Helpdesk staff's interactions with the SSC. Liz Melrose subsequently forwards the
email to me on 6 November 2007. On 7 November 2007, I forwarded it to the team
managers who reported to me at the time.

46. Due to the time that has elapsed, I do not recall the emails on FUJ00138214. However,
my recollection is that the issues described in Mik Peach’s 2 November email were
not common. In any case, some errors in correctly or punctually routing calls were to
be expected given the Helpdesk was handling thousands of calls a month.

47.While I do not recall the email from Mik Peach, the Helpdesk’s general practice at the
time would have been to take such procedural issues seriously and to resolve the
alleged issues. While I do not recall the specific action taken in this case, Team
Managers would usually identify the agents involved and provide some one-to-one
coaching on areas where they had been lacking.

48.1 would agree with Liz Melrose’s assessment in her email of 6 November that the
rollout of the new TfS call management system did lead to pressure on the Helpdesk.
Having reviewed an email chain involving FSL Amber Alerts, me and others, subject
heading “RE: Amber Alert Ref 998 — Various T4S Customers - UPDATE REPORT”
dated 22 January 2008 (FUJ00152279) and an email chain involving me and others,
subject heading “FW: HSD — Update on Telephony Failure” dated 31 January 2008
(FUJ00152280), I recall there were initial performance issues with the TfS system,
specifically intermittent problems with the OTI interface that passed tickets from the
Helpdesk to the SDUs, which resulted in those tickets having to be transferred
manually. A summary of the issues that occurred is contained in the documents I

mention above and in the TFS Issues Log (undated) (WITN09550107). These issues

Page 21 of 28
WITNO09550100
WITN09550100

resulted in an “Amber Alert” where I believe all issues were captured and resolved.
My recollection is that the difficulties with the TfS system continued sporadically for
several months before they were fully resolved.

FIRST WITNESS STATEMENT OF AMANDEEP SINGH

49.The Inquiry has referred me to paragraph 6 of the First Witness Statement of
Amandeep Singh (WITN06660100). The behaviour described there is unacceptable
and disgraceful. I do not recall ever witnessing such behaviour on the Helpdesk or
becoming aware of such behaviour. I believe that, while I was working on the
Helpdesk, if such behaviour had been observed, it would have resulted in disciplinary
action.

DOCUMENT FUJ00002254

50.1 can confirm that I was not involved in the production of the Audit Trail Functional
Specification, Version 12.0, dated 8 October 2010 (FUJ00002254).

ADEQUACY OF SUPPORT PROVIDED BY THE HELPDESK TO SUBPOSTMASTERS

51.1 believe that the Helpdesk generally did provide adequate support to subpostmasters.
I have this view for the following reasons:

a. Our SLTs were generally met, and our performance on those could have a

direct impact on the overall quality of a subpostmaster’s experience when

calling the Helpdesk.?

2 To assist my recollections about the Helpdesk’s SLT performance, I reviewed the SRBs from the period
I worked on the Helpdesk. These are: Service Review Book (November 2005) dated 15 December 2005
(FUJ00152251); Service Review Book (December 2005) dated 16 January 2006 (FUJ00152252); Service
Review Book (January 2006) dated 14 February 2006 (FUJ00152255); Service Review Book (February
2006) dated 7 March 2006 (FUJ00152256); Service Review Book (March 2006) dated 18 April 2006
(FUJ00152258); Service Review Book (April 2006) dated 15 May 2006 (FUJ00152259); Service Review
Book (May 2006) dated 27 June 2006 (FUJ00152260); Service Review Book (June 2006) dated 14 July
2006 (FUJ00152261); Service Review Book (July 2006) dated 14 August 2006 (FUJ00152262); Service

Page 22 of 28
WITNO09550100
WITN09550100

b. The Helpdesk strived to improve the service provided to branches by
applying a “Sense and Respond” methodology. In 2007, the Helpdesk
proactively reviewed high volume call categories and produced action plans
to reduce call volumes and improve the end user experience of
subpostmasters. Having reviewed the report from that review, the Post
Office Ltd and Fujitsu Services Sense & Respond Action Plan dated 24
September 2007 (FUJ00141282), the activities that resulted from that
review included (i) an Epson Printer refresh rollout (ii) providing branches
with cleaning kits for their Pinpads, (iii) an improved reboot process, (iv)
revision of relevant scripts to improve diagnosis of hardware faults, and (v)
improvement of the quality of spares given to engineers to reduce “fail on
fit” issues.

c. The Helpdesk would often receive compliments from subpostmasters for
the service given. I recall that the Helpdesk team was proud of the work that

it did in support of the Post Office and there was even a noticeboard in the

Review Book (August 2006) dated 14 September 2006 (FUJ00152263); Service Review Book
(September 2006) dated 13 October 2006 (FUJ00152264); Service Review Book (October 2006) dated
14 November 2006 (FUJ00152265); Service Review Book (November 2006) dated 14 December 2006
(FUJ00152266); Service Review Book (December 2006) dated 15 January 2007 (FUJ00152267); Service
Review Book (January 2007) dated 14 February 2006 (FUJ00152268); Service Review Book (February
2007) dated 14 March 2007 (FUJ00152269); Service Review Book (March 2007) dated 17 April 2007
(FUJ00152270); Service Review Book (April 2007) dated 15 May 2007 (FUJ00152271); Service Review
Book (May 2007) dated 14 June 2007 (FUJ00152272); Service Review Book (June 2007) dated 13 July
2007 (FUJ00152273); Service Review Book (July 2007) dated 14 August 2007 (FUJ00152274); Service
Review Book (August 2007) dated 14 August 2007 (FUJ00152275); Service Review Book (September
2007) dated 12 October 2007 (FUJ00152276); Service Review Book (October 2007) dated 14 November
2007 (FUJ00152277); Service Review Book (December 2007) dated 15 January 2008 (FUJ00152278);
Service Review Book (January 2008) dated 15 January 2008 (FUJ00152281); Service Review Book
(February 2008) dated 14 March 2008 (FUJ00152282); Service Review Book (March 2008) dated 14
April 2008 (FUJ00152283); Service Review Book (April 2008) dated 15 May 2008 (FUJ00152284);
Service Review Book (May 2008) dated 15 June 2008 (FUJ00152285); and Service Review Book (June
2008) dated 14 July 2008 (FUJ00152286).

Page 23 of 28
WITNO09550100
WITN09550100

office containing excerpts of messages from subpostmasters (or other
branch staff) praising the service the Helpdesk had given.

d. The Helpdesk was also nominated as one of three finalists for an award
given by the Helpdesk Institute (“HDI”) (now known as the Service Desk
Institute), an industry wide body promoting best practice on IT service
desks. While we didn’t eventually win, I believe such external recognition

shows the passion the Helpdesk team had to go the extra mile.
52.For completeness, I note that subpostmasters’ experience of the support process
would usually be affected by factors aside from the Helpdesk’s performance. By way
of illustration, if there were delays within Engineering around the delivery of
appropriate parts, that could affect the length of time it took before a problem was

resolved, thought this would not be attributable to the Helpdesk’s service.

Statement of Truth

I believe the content of this statement to be true.

Signed: _GRO

Dated: 9 May 2023

Page 24 of 28
WITNO09550100
WITN09550100

INDEX TO FIRST WITNESS STATEMENT OF PAUL ANTHONY GARDNER

Exhibit No.

Description

Control Number

URN

ICL Pathway/POCL Interface
Agreement for the NBSC and HSH
Interface, Version 2.0, dated 13
July 2000

POINQ0086577F

FUJ00080406

POA Customer Service Incident
Management Process Details,
Version 3.0, dated 23 March 2005

POINQ0086110F

FUJ00079939

End to End Support Process,
Operational Level Agreement,
Version 2.0, dated 17 June 2003

POINQO086068F

FUJ00079897

Service Level Targets for Horizon
Services, Version 3.0, dated 19
August 2005

POINQ0008137F

FUJ00001966

Horizon Service Desk Joint
Working Document, Version 1.0,
dated 27 July 2007

POINQ0086630F

FUJ00080459

Call Enquiry Matrix and Incident
Prioritisation - Reference Data,
Version 1.0, dated 23 April 2007

POINQO086669F

FUJ00080498

Call Enquiry Matrix and Incident
Prioritisation - Software, Version
1.0, dated 23 April 2007

POINQ0086670F

FUJ00080499

Call Enquiry Matrix and Incident
Prioritisation — HSD, Version 1.0,
dated 17 April 2007

POINQ0086672F

FUJ00080501

POA Customer Service Major
Incident Process, Version 3.0,
dated 7 July 2009

POINQ0086242F

FUJ00080071

10.

RMGA End-to-End Customer
Complaints Procedure — Joint
Working Document, Version 2.0,
dated 24 December 2007

POINQ0086205F

FUJ00080034

11.

RMGA Customer Service Problem
Management Process, Version
2.0, dated 22 April 2008

POINQ0086214F

FUJ00080043

12.

Processes - ID 408501 - Engineer
Refused Access Process
(undated)

POINQ0144928F

FUJ00138733

13.

Post Office Ltd / Fujitsu Services
Retention of the Incident and

WITNO9550101

Page 25 of 28
WITNO09550100
WITN09550100

Communications Management
Teams document dated 4 April
2006

14.

Team Manager HSH Skills Matrix
(undated)

WITN09550102

15.

HSD Customer Support Technical
Skills Matrix dated 3 November
2005

WITN09550103

16.

D1 — D4 Core Skills Matrix for PSE
(undated)

WITNO09550104

17.

OMR for Commercial & Public
Sector Call Centres, Customer
Services dated September 2005

WITN09550105

18.

Horizon Service Desk Operational
Development Plan Current
Activities, Version 0.3, dated 6
January 2006

WITN09550106

19.

Fujitsu Services Post Office
Account - Service Review Book
(February 2007) dated 14 March
2007

POINQO089600F

FUJ00083429

20.

First Witness Statement of
Amandeep Singh dated 13
January 2023

WITNO6660100

WITNO6660100

21.

Email from Paul Gardner to John
Casey, Paul Hailey, Sarah Hill and
others, subject heading “FW:
Process issues at HSD”, dated 7
November 2007

POINQ0144409F

FUJ00138214

22.

Spreadsheet attached to
FUJ00138214, dated 7 November
2007

POINQ0144410F

FUJ00138215

23.

Email chain involving FSL Amber
Alerts, Paul Gardner and others,
subject heading “RE: Amber Alert
Ref 998 — Various T4S Customers
— UPDATE REPORT” dated 22
January 2008

POINQ0158473F

FUJ00152279

24.

Email chain involving Paul Gardner
and others, subject heading “FW:
HSD — Update on Telephony
Failure” dated 31 January 2008

POINQ0158474F

FUJ00152280

25.

TFS Issues Log (undated)

WITN09550107

26.

Audit Trail Functional
Specification, Version 12.0, dated
8 October 2010

POINQ0008425F

FUJ00002254

Page 26 of 28
WITNO09550100

WITNO9550100

Service Review Book (November POINQ0158445F I FUJ00152251
27. 2005) dated 15 December 2005

Service Review Book (December POINQ0158446F I FUJ00152252
28. 2005) dated 16 January 2006

Service Review Book (January POINQ0158449F I FUJ00152255
29. 2006) dated 14 February 2006

Service Review Book (February POINQ0158450F I FUJ00152256
30. 2006) dated 7 March 2006

Service Review Book (March POINQ0158452F I FUJ00152258
31. 2006) dated 18 April 2006

Service Review Book (April 2006) I POINQ0158453F I FUJ00152259
32. dated 15 May 2006

Service Review Book (May 2006) POINQ0158454F I FUJ00152260
33. dated 27 June 2006

Service Review Book (June 2006) I POINQ0158455F I FUJ00152261
34. dated 14 July 2006

Service Review Book (July 2006) I POINQ0158456F I FUJ00152262
35. dated 14 August 2006

Service Review Book (August POINQ0158457F I FUJ00152263
36. 2006) dated 14 September 2006

Service Review Book (September I POINQ0158458F I FUJ00152264
37. 2006) dated 13 October 2006

Service Review Book (October POINQ0158459F I FUJ00152265
38. 2006) dated 14 November 2006

Service Review Book (November I POINQ0158460F I FUJ00152266
39. 2006) dated 14 December 2006

Service Review Book (December I POINQ0158461F I FUJ00152267
40. 2006) dated 15 January 2007

Service Review Book (January POINQ0158462F I FUJ00152268
41. 2007) dated 14 February 2006

Service Review Book (February POINQ0158463F I FUJ00152269
42. 2007) dated 14 March 2007

Service Review Book (March POINQ0158464F I FUJ00152270
43. 2007) dated 17 April 2007

Service Review Book (April 2007) I POINQ0158465F I FUJ00152271
44. dated 15 May 2007

Service Review Book (May 2007) POINQ0158466F I FUJ00152272
45. dated 14 June 2007

Service Review Book (June 2007) I POINQ0158467F I FUJ00152273
46. dated 13 July 2007

Service Review Book (July 2007) POINQ0158468F I FUJ00152274
47. dated 14 August 2007

Service Review Book (August POINQ0158469F I FUJ00152275
48. 2007) dated 14 August 2007

Service Review Book (September I POINQ0158470F I FUJ00152276
49. 2007) dated 12 October 2007

Page 27 of 28
WITNO09550100

WITNO9550100

Service Review Book (October POINQ0158471F I FUJ00152277
50. 2007) dated 14 November 2007

Service Review Book (December I POINQ0158472F I FUJ00152278
51. 2007) dated 15 January 2008

Service Review Book (January POINQ0158475F I FUJ00152281
52. 2008) dated 15 January 2008

Service Review Book (February POINQ0158476F I FUJ00152282
53. 2008) dated 14 March 2008

Service Review Book (March POINQ0158477F I FUJ00152283
54. 2008) dated 14 April 2008

Service Review Book (April 2008) I POINQ0158478F I FUJ00152284
55. dated 15 May 2008

Service Review Book (May 2008) POINQ0158479F I FUJ00152285
56. dated 15 June 2008

Service Review Book (June 2008) I POINQ0158480F I FUJ00152286
57. dated 14 July 2008

Post Office Ltd and Fujitsu POINQ0147477F I FUJ00141282

Services Sense & Respond Action
58. Plan dated 24 September 2007

Page 28 of 28