So the NAO has just posted today a progress report on this £2.4bn (to 2023) project which the Department for Work & Pensions is undertaking with the help of Accenture, IBM, HP and BT.
The NAO's conclusions don't make pretty reading - in summary they are:
- value for money in the early implementation has not been achieved;
- the programmes timetable and scope were over ambitious, risks were taken to try and achieve them, there wasn't an explanation of how the plans had been decided upon on and their feasibility wasn't evaluated;
- a new "agile" project management approach (typically used in small collaborative teams) was used on a project which grew to over 1000 people - and such an approach had never before been used on a programme of this size and complexity (the Agile approach uses iterative and collaborative project management);
- there was an absence of a detailed view of how Universal Credit was meant to work (you can see how this might lead to some problems in implementing a large IT system);
- there was poor control and decision-making which undermined confidence in the programme and contributed to a lack of progress;
- there was a lack of IT expertise and senior leadership and frequent changes in senior management (5 senior responsible owners since mid 2012)
The NAO's recommendations focus unsurprisingly on: Producing a realistic plan with clear programme objectives, linked to policy design & service requirements; Using a management approach that allows policy experts, operational teams & systems developers to work together; establish effective governance processes & structures; Tightening ﬁnancial management & control over spending; Re-assessing existing programmes & capabilities in light of the experience on Universal Credit.
But what was the root cause of the failures found and underlying issues identified?
In my opinion much of what has gone on seems to have been caused by a familiar story as concerns large projects and early announcements about them being:
- initiated by high flying strategy wonks who are largely driven by political or news management imperatives and who won't be in the firing line if it all goes horribly wrong;
- based on little consultation with implementation grunts and a surprisingly low level of specific detail in the feasibility studies upon which the announced costs and timescales are based.
In other words - if you want to give a project a good chance of succeeding then have a detailed scope with corresponding cost and programme estimates
(ht to wikipedia for the image)