Europe's largest council could face £12M manual audit bill after Oracle project disaster Europe’s largest local authority faces a $15.58 million (£12 million) bill for manually auditing accounts which should have been supported by an Oracle ERP systems installed in April 2022. Birmingham City council Oracle megaproject got a code red – and they went live anyway READ MORE The £3.2 billion ($4.1 billion) budget …
They all need multi-tens of millions of pounds custom solutions
Surely UK councils have enough commonality that the bulk can be done by one solution provided for and paid by central government and custom plugins bolted on at that council's expense for wanky vanity projects
I get the feeling the likes of oracle have cottoned on to this anyway but there's money to be made in reinventing the wheel every time
90% of council functions are wanky vanity projects. Our council has loads of harbours. Birmingham doesn't. Birmingham owns an airport, our council doesn't. Our council has - I think - 300+ parish councils. Birmingham has two. Our council interacts with, I think, four NHS CCGs. Birmingham: one. Our council is thousands of square miles of countryside, Birmingham is about 100 square miles of urban city. Two councils may provide identical services, but purchase them from two different suppliers with two different sets of billing and accounting procedures.
you do raise an interesting perspective i never realised the scope of actually.
for hundreds of millions though i'd hope 1 solution could load individual modules to handle a council on the moon and one in london with equal ease, i see so many high numbers in government IT spending they're beginning to lose all sense of size. (I think government accountants must have been hit by that a long time ago if they're still approving these things)
There is the flip side to trying to create a central authority for software. In that you're likely to have one that really bad solution over many lesser bad solutions. Government work in general is very litigious and trying to normalize that with software is off more difficult than expected.
This isn't to say that Oracle didn't screw up. They most likely did
Two councils may provide identical services, but purchase them from two different suppliers with two different sets of billing and accounting procedures.
Only two? Our council seems to use several suppliers just for the services from that one council. Hence the 7 recycling bins, collected by 4 separate lorries, that we have to use.
Yes, our seven districts councils have been abolished and merged into one unitary county council, so yes, we have seven different waste collection departments to merge, seven different licensing regimes to merge, seven different local planning policies...
We were sold the merger on things like: the road is county council highways, but the grass verge is district council open spaces, it's never coordinated. But now we've got: the road is county highways, but the grass verge is county open spaces, it's never coordinated.... sigh.
It seems to me a lot of this is a sort of governmental technical debt: Various authorities are created to fill various needs, but there's no overall design, of course, and over time requirements change but the authorities either don't change, or the changes themselves aren't well thought-out.
That said, from here it's always seemed that the UK has this problem particularly badly. Where I live, there are municipal governments, county government, Pueblo government (and the interactions between state and reserved lands are ... interesting, to say the least), state government, and Federal government; and then there are all sorts of other organizations with some political power, like the land-grant associations1 and the acequia associations2. On the Federal side, you have plenty of interested and often conflicting players, like the Bureau of Land Management and the Forest Service (the county has a lot of BLM land and national forest) and Reclamation and Agriculture and Indian Affairs. Coast Guard is about the only Federal agency that isn't up to something in these parts.
Yet somehow we don't seem to end up with quite the same sorts of messes I've read about in these articles and comments. Certainly the political situation is fraught in various ways and there are inefficiencies and screw-ups, but nothing of this scale (even scaling down for the much smaller population).
1Inheritors of certain land rights dating back to the reconquest of Nuevo México by Spain in 1692. They're still around.
2Which maintain the traditional irrigation-ditch system and control access to surface waters, as apportioned by treaties and Federal laws and the Office of the State Engineer and several of the longest-running lawsuits in US history. Water rights are hugely important here.
"From here it's always seemed that the UK has this problem particularly badly."
As someone who grew up in a country with regional and national departments doing what's delegated to councils in Britain(*), I wholeheartedly concur
(*) Education, police, fire in particular if funded/administered at council levels run into the issue of rich areas getting better services than poor ones
"our seven districts councils have been abolished and merged into one unitary county council"
The problem is that decades of manual systems have been locked into computerised ones and now the staff _WON'T_ adapt to change even if the systems chosen are objectively better
Of course to achieve THAT involves talking to the users, scoping out what's needed, trying to standardise as much as possible on what's best common practice AND taking a baseball bat to any politicians or managers who try to "help" (aka, line their mates' trousers)
-- Two councils may provide identical services, but purchase them from two different suppliers with two different sets of billing and accounting procedures. --
From the sentence above I'm unsure whether its the councils that have different sets of billing and accounting procedures or the suppliers. If its the former I think I've spotted the problem, if the latter then with the amount the councils spend I think they could be persuaded to change.
You suggest this against a background where our brand-new, sparkly government is spouting lots of nonsense about further devolution of powers to English regions. Given how well that is working in Scotland and Wales, I can really see that being a good thing. - NOT.
I can't speak to UK in particular. But governments are often very difficult to build software for. Usually weird and conflicting requirements and unknowns.
I worked on an election software and the variance in legal rules across countries and states was staggering and painful to build for. Each county or state added to twice as long as the last one to implement and doubled the test surface.
I still generally avoid IBM and Oracle though. They take cost overruns to a whole new level.
In the simplest model, there three layers of council.
1. County Council
2. District Council
3. Parish Council
County and District Councils have core responsibilities which remain standardised with the exception of councils with a coastal area, but even those are standardised responsibilities.
Parish Councils have such a limited number of responsibilities that an excel spreadsheet is pretty much all they need.
There is absolutely no need for councils to go shopping for complex software that they then proceed to try and "customise" because they feel somehow unable to adjust their internal processes to match the COTS standards prevalent with those packages, or because they see themselves as a "special case".
The reality is that no council is so "unique" in what they do that they could not use a Government Digital Service (GDS) developed and supported platform, and indeed some common apps used by councils are already provided by GDS.
A bit of forethought would allow for things like unitary realignments, vanity projects and changes in responsibility from central government.
Yes, it would take money to develop, but you only have to look at the ridiculous cost spent by Birmingham City Council to achieve... absolutely nothing, to justify the dev costs.
So they want to spend £12m of public money on a box-ticking exercise covering historical data.
They are bankrupt. Sack everyone, draw a line in the sand and just begin afresh. Every other option will just waste more money.
And it doesn't cost that much to do your accounts, old school, unless your accountants are working from home in Hawaii.
We really need to develop better, simpler systems. Not these universal packages that just don't work.
-- We really need to develop better, simpler systems. Not these universal packages that just don't work. --
Its a bit like NPfIT - over specify initially then add everyone's pet desire in, accommodate every possible way and method of working, extend to cover every possible edge case and off you go.
Oracle didn't implement it, I expect they are just the software vendors not the implementation partners.
There are others using the same software fine but of course failures it's much more news worthy. It seems though from the article a lot of it is also external systems to Oracle but it's hard to tell.
I guess if it's not auditable the. It's likley going to be external system integration as tracing transactions through finance is intrinsic to the system.
They'll probably find that the problems were raised during the design phase by experienced staff on the ground. Then ignored by senior management. Who are tunnel-visioned on delivering the project by a fixed time. So they can collect their bonuses and move on to another organisation before everything hits the fan.
Oracle. SAP. All have their strengths and weaknesses.
Not surprising, the smoke and mirrors sales people have usually bullshitted their way in and convinced those with no understanding that their solution will be magical, brilliant, transformational .... Oracle have a lot to answer for but also the morons at the councils who believe the bullshit, don't do any due diligence and sign on the line. They all deserve each other
Sadly, you have hit the nail on the head.
Procuring bespoke software is not the same as sourcing as paperclips. It is an extremely challenging task and even when undertaken by someone with decades of relevant skills, ability and experience has at best about 50% chance of delivering on time and budget. Assigning this task to someone who has little or no relevant skills, ability and experience is a recipe for disaster as we see time and again with Government/Public Sector IT projects. The IT vendor makes huge profits delivering something which is usually late and over budget and then enjoys years of additional work and obscene profit fixing the problems. And because it all gets funded by the taxpayer no-one on the procurement side faces any level of accountability.
This was basically the cause of the Post Office Scandal.
So £12 million for a manual audit
The cost to implement the new system has climbed to £131 million so far
You could do the audits manually for a decade for that! And that's not counting the ongoing maintenance costs, which you could conservatively estimate at 10% of initial system cost.
Alternatively, you could try to avoid spending the required amount - but after 10 years, the new system would have rotted to the extent of needing an upgrade or replacement anyway.
So my modest proposal: go back to doing things manually for good and forget about ERPs. This is clearly the cost efficient solution compared to the horror of these implementations.
The problem is that replacing these large legacy projects is only successful if you take the smallest possible bit, migrate it, test it, fix it and watch it, then take the next smallest possible bit and so on. This approach simply doesn't fit into the 'give us a fixed, massive upfront cost and fixed scope quote' bucket that manglement wants.
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2024
