POL00028344
POL00028344
2.0 AUG 1999-430
Electronic memo
To Bruce McNiven/POCL/POSTOFFICE, John Meagher/POCL/POSTOFFICE, David
W Miller/POCL/POSTOFFICE, David W Miller/POCL/POSTOFFICE
cc ""Andrew Simpkins’ <andrew.simpkin:
Baines/POCL/POSTOFFICE
Hard Copy To
Hard Copy cc
From Chris French <chris.frenc
Date 19/08/99 18:04
Subject Acceptance - Additional Offices
Bruce,
here is the brief paper in bullet point style for the meeting tomorrow with
Dave Miller.
I have shown this to Keith Baines who was content. (he did not press for
the gap at AB, RAB as long as it is clear we would stop if acceptance still
had not been achieved, but I left it in as we agreed)
By the way, Pathway are bringing Christou, Bennett, Openheim, Coombs, to
the meeting with Dave at 10 am.
Kind regards
Chris
- Scenario-Accept-2.doc
POL00028344
POL00028344
r- HN
Acceptance - Second Stage
This note reviews the value of implementing additional offices in the Live Trial , and
recommends when this could occur, the number of offices and the benefits.
It is part of the strategy to maintain the initiative with POCL
Objectives:
To de-risk the next stages of the project:
- acceptance decision
- RAB decision
- roll-out, beat rate
We must have a proper acceptance process for this mid-term acceptance opportunity
within the 3 month period for re-test.
Why add any offices?
1) There is a set of changes we wish to introduce that can only be demonstrated by
new offices, e.g.:
- new training (courses, material)
- new scripts for the HSH
2) The existing offices cannot test some improvements because:
- they already have old habits e.g.:
- not ringing the HSH,
- rebooting when they hit known problems
oo work arounds to avoid known problems, and
- subconscious avoidance
- we cannot test any new business rules or compliance rules in offices used to
operating Horizon in other ways, this area is seen as one of the largest risks to
the roll-out.
3) The Instability Acceptance Incident causes Pathway and ourselves the greatest
concern. This has over twenty causes, some of which are rare events within current
sample size. A significant increase in the number of the trial offices greatly increases
the opportunity to identify and deal with these events.
Additional offices would enable us to perform a proper evaluation once the service
was stable.
4) We have only tested our, and Pathways processes, up to about 60 offices per week
This is a long way from the beat rate of 300, and the difficulties are not all linear with
the number of offices. Only by adding new offices can we lear any lessons on how
to do this better or give confidence in the process. We need to demonstrate the
ability to handle much higher volumes in:
- training (Peritas)
- logistics of implementation
- information flow
- changes/fixes
Finally it is highly desirable in large roll-outs to have two separate test processes:
POL00028344
POL00028344
or
- a trial to prove the service and the surrounding business processes in the outlet
- a pilot (using the proven software/processes) to prove the roll-out processes.
This has been a concern from the start, but the contract had, until now, prevented a
proper test of the roll-out processes. We now have the opportunity to put right this
omission.
At the least we should test the roll-out procedures at half the planned beat rate. Le.
run one week with 150 offices implemented.
5) We need more offices to stress their service to identify weak points. We need an
informed view on these risks before roll-out.
6)There are other reasons, of lower priority, for adding more offices:
- decreases the cost overall to POCL
- maintains some momentum and morale
- reduces ICL’s costs
The risk of not doing these tests is that we fail to push Pathway to an acceptable
service, and that we gloss over POCL weaknesses that could have been rectified.
Timing
There are conflicting pressures:
- to give ICL the best chance to get proper improvements: we need a later date
- to reduce wasted cost: we need to have a positive AB and RAB as early as
possible.
Given these, and the wider POCL needs, an end September date is likely to be the
target.
We believe that even one extra week before re-test would increase the performance of
the service. The first week of 40 offices should not be implemented and we would
like to confirm this immediately.
The schedule we recommend is:
Week commencing Original schedule Revised Schedule
31 August 40 1 (a special office)
6 September 60 60
13 September 90 90
20 September 158 158
27 September 178 0 (gap for AB/RAB)
4 October 198 198
11 October 198 198
18 October 198 198
25 October 198 198