Ellison's exemplar SAP-to-Oracle region rules out ditching Oracle Birmingham City Council – Europe's largest local authority, serving over one million customers – has agreed to re-implement an Oracle Fusion system following a failed rollout that saw costs escalate by more than £100 million ($125.6 million) and the authority unable to fulfill its statutory duties. In 2021, Oracle founder and …
What would it have cost to implement this using open source technology, eg using postgreSQL on Linux using Java or Python ?
This could be made available to other councils -- which must all have similar requirements.
True: there would be migration & other costs; but it would have to be less than pissing cash away to a USA based enterprise that will just suck licensing fees for years to come.
I know that large government IT projects do not have the best reputation (think: NHS), but it seems that ones done by private companies are not much better.
Yes: cash up front for the 5 years to develop it, but reduced costs later -- and you get something that is designed to do the job that councils want.
ERP systems are beasts. They're not trivial web-apps. It'll take an army of developers to write an open source ERP system.
ERP systems also have to fullfill numerous rules (e.g. Accouting, payroll, audit, etc). You're gonna need to get your open-source software validated that it follows all these rules & regulations. That ain't going to be a cheap task. And every time you update your software....
Just browse LWN.Net and see all the articles over the years the editor was written about trying to use open source software to run a small business.
"It'll take an army of developers to write an open source ERP system."
A lazy, small, incompetent army.
Build costs for an entry level ERP system are in the £1-5m range. There's no magic sauce, no especially difficult functionality. And that sort of money is non-trivial but would be easily found by a group of councils.
I am sure I could put together a team to build them a bespoke payroll system for £5mil. The biggest hurdle to making the system reliable would be how much of that budget would be left over to spend on "keeping the the development team fueled"...
I am sure you could, I am sure many others could, I am also sure that there were many other alternatives to Oracle (there are, I looked them up, one is open source) that could have done the job, but that's not what this is about. It is a combination of the English always wanting the shiniest, biggest, and what "they" believe is the best thing* and Larry and his crew of criminals finding a way to suck millions out of the city!
*How do I know this? I work for a British multi-national and I watch Brits hired as IT executives take a company with over 900K pounds cash in the bank and in less than 3 years waste most of that money. Yes, Oracle was part of that mess. We still deal with this mess. Azure instead of AWS, why? Because "oooooh" Microsoft! It's a mindset I just do not understand!
Those responsible for this should have been fired, investigated and surcharged or imprisoned, as appropriate.
If you live in Brum you should consider moving, because the costs of the wage case and this farce means that you are not going to be getting much in the way of services for the next decade.
Once they have thrown another hundred million of taxpayers' money at this, it will be time for them to look for its successor. I expect Oracle will be preparing for the tender by adding AI to their current offerings.
Councils really need to consider using less/simpler tech. Use accountants, printed forms, Excel, A4 pads, whatever. It must be cheaper and more reliable than this shameful disaster.
If you live in Brum you should consider moving, because the costs of the wage case and this farce means that you are not going to be getting much in the way of services for the next decade.
As of now the council don't know what their budget or cashflow is so they decided to cut a whole bunch of services just in case.
For those of us who have been involved with the technical parts of the procurements the outcome is fairly inevitable.
Procurement exists to ensure rules are followed and they will enforce that (rightly). Unfortunately the government puts a stack of criteria around the size & financial status of the companies responding, often excluding responses that are better. This because the company does not meet turnover and other criteria so are deemed to be high risk. Then you have the frameworks that are in place that simplify the process as you only need to invite companies on the framework, not open to all
The problems start when the tendering process essentially becomes an arse covering exercise to ensure that the council cannot be sued by the companies submitting unsuccessful tenders. This is because the private sector sees the public sector as a cash cow to be stripped of as much money as possible at all points.
That is not the fault of the councils.
Forgive my ignorance, but could somebody briefly explain why these councils feel the need to spend an eye-watering sum of money to switch from a (presumably) working SAP system to a start-again-from-the-ground-up Oracle system?
I mean.... why embark on this fraught journey, when they could have just stuck with what (I presume) had been working well?
Shiny
An example a good number of years ago my employer decided to replace a system using text terminals (ibm 5250) with a new system.
This was to be windows with loads new functionality.
In reality after a year the new product barely functioned would have required every device in the company to be replaced and the network upgraded (the old system used about 2k to do a transaction the new needed 10mb of data) and was abandoned after going to court.
When I left a few years later they were still on the old terminals because they worked…
It looks like BCC used SAP while IT was outsourced to Crapita between 2006-2019.
Crapita was dumped because they were viewed as expensive. There is also mention of the SAP systems being heavily customised and I would suggest (without knowing) that a heavily customised SAP probably wasn't patched/updated because it would break the customisations.
So the move to Oracle was with a relatively inexperienced IT team, eager to show they could outdo Crapita, taking on a very challenging migration and while the intention was to avoid customisation, the business pressured the implementation team to attempt to replicate at least some of the custom SAP functionality. There also seems to be some BCC signoff issues where the Council failed to approve work without understanding that that would lead to increased costs at that there might be a cheaper alternative possible.
It's worth pointing out that the annual costs for running SAP were ~£5m/year versus £750k/year for Oracle. As SAP has to remain in service for a number of years post-migration for audit purposes, each year of delay has added another ~£6m of operational costs which is part of the reason for the over runs.
There is a ComputerWorld article that covers this saga in depth.
Thank you AC. I found the article to which you referred...
Birmingham City Council’s Oracle implementation explained: What went wrong?
Good grief. I didn't realise just how deep the problems went on this one.
Does any setup involving Oracle actually ever work? I have seen lots of articles over the years of huge failures with the system never doing what people wanted it to do. I have seen a few attempts myself (One company I worked at had a pile of people working on doing an Oracle setup for a couple of years with no end in sight when I left). Does it actually ever work as promised?
Lennart Sorensen,
Yes, it does work as promised ..... *but* never on the project you are working on !!!
I have met many people who know of successful projects *but* never managed to be there when one has been completed, on time & to costs.!!!
Scope Creep and over-ambitious timescales, leading to escalating costs seems to be the norm, particularly when govt (local or otherwise) is involved.!!!
:)
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2025
