UK govt Matrix has unenviable task of consolidating several different ERP systems The UK government has kicked off procurement of an ERP system for eight Whitehall departments which consolidates nine different software systems – a project potentially more complex than a snake's wedding. According to civil service chief operating officer Alex Chisholm, the group, dubbed Matrix, is the "trickiest one" among …
I have no knowledge of these things, beyond what I've read about earlier disasters- so just thoughts and questions. Like;
What problem are they trying to solve
Do the needs of all the different departments match enough to combine the software
Is there enough knowledge and understanding of what these system do and how they need to do it
Are the problems they're trying to solve ( above) sufficient to require the cost and disruption this will cause
Is it too complex for anyone to get their heads round
Has the cost included the retraining of all the staff (This last bit I have experience of, albeit on a small scale, and the answer is usually no.)
I wonder whether they will specify the correct security requirements when they do the analysis, or just try to 'bolt it on afterwards' when they realise that they need some segregation of data, applications and access? Strikes me that the requirements analysis for amalgamating the IT of 8 government departments including the treasury will be fraught with difficulty. Each department will want (certainly should have) a 'senior user representative' to ensure they can still perform their functions. Groups of eight trying to agree on something contentious have a history of long term deadlock.*
Oh well, I'll just have to keep my fingers crossed.
*https://www.telegraph.co.uk/news/science/4221839/Eight-people-on-committee-leads-to-decision-deadlock-scientists-say.html
"A computer modelling exercise, which used mathematical theories to calculate how quickly committees of different sizes were likely to reach agreement, singled out the number as uniquely bad for decision making...."
Perhaps they need a meeting or two to review major projects over the last 20 years to work out why some worked - and some did not work, and actually apply the lessons learned!
Simple lessons like
Agree the major requirements (and users) before you start, rather than half way through.
Do not add on "if we add this... function - we can do this clever nice to have at 10* the original cost."
How many millions of records in the database, and what end user response time is expected.
Prototype it at scale before you start coding.
Migrate an ERP system - Nope, been involved with that before, it's a nightmare.
Migrate from seven different ERP systems - Double nope.
Migrate from seven different government ERP systems, from different vendors and with different requirements - Hell no!
Sounds like a hell project for all concerned.
My suspicion is that serving 7 sets of highly entitled-feeling senior bureaucratic types, with an even more entitled-feeling ( and possibly titled too) top bureaucrat leading each one is going to be horrific. Each and every one is going to want their pet foible built into the system. Each and every one is going to want to take credit for some "improvement". Every rivalry and resentment is going to played out.
There isn't enough popcorn in the world.
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2024
