Europe’s biggest city council faces £100M bill in Oracle ERP project disaster Birmingham City Council is set to pay up to £100 million ($123 million) for its Oracle ERP system — potentially a four-fold increase on initial estimated expenses — in a project suffering from delays, cost over-runs and a lack of controls. Newly appointed council leader John Cotton told regional news outlet the Birmingham Mail …
"The solution will not be running in 10 years' time, I bet you. So an overall loss."
In ERP vendor thinking, once the client is on one or another product, they will be there for 10 years minimum. Oracle sales folk would fight like rats in a sack to win one. DBs you might be able to turn over more quickly, but not an ERP system.
That said, if this goes belly up they might not get in there for year 1.
"For £100m you could not just hire some techies, you could start a software company to build the solution. At least then you'd have a solution, and a software company."
How do you think Oracle came up with their products in the first place...
Another go on the fcuk it up and fcuk off somewhere else to fcuk up again roundabout.
Nobody __ever__ saves money moving to Oracle.
Choosing Oracle for anything should be classified as negligence and tax payers money recovered from the people who authorise the spending.
If the system isn't broke then surely the solution is to leave it alone.
And anyone who thinks / says that a major IT project is going to come in on time and in budget is either demented or else on a backhander.
Even worse that the Oracle system implemented isn't intended for this sort of application.
Regardless of anything else, someone is laughing all the way to the bank (clue - it's not the tax payer).
> If the system isn't broke then surely the solution is to leave it alone.
Absolutely. But the problem then is that you get morons and consultants (sorry, I repeated myself there) saying things like “still running a mainframe system and COBOL in this day and age? Sheesh, grandad, get with the times!”. Phrases like “technical debt” start getting bandied around. Managers and people who should know better get seduced by shiny-shiny cloud vendor press releases. And… well, this sort of situation arises.
I've read so many of these stories. And every time, every fucking time, it's the same story. A bloody great big system that is going to cost squillions, even in the unlikely event that it doesn't cost a multiple of the original tender, is bought in from one of the big corporates, who's main area of expertise is extracting money from clients that don't have the skills or expertise to define what they need, or manage its supply.
And they never fucking learn!
...is bought in from one of the big corporates, who's main area of expertise is extracting money from clients that don't have the skills or expertise to define what they need
You'll find the big corporates are quite happy to pay expensive lawyers to tie the public sector in legal knots. The public sector just can't compete with the legal budget. The corporates know this and are happy to continue to pay for the expensive lawyers as it's a legal way to win big contracts - and to continue squeezing money out of the public sector.
I think there needs to be some distinction between the systems integration team and the Oracle Costs.
How much of this 100m is going to Oracle, they'll have had the license costs. The cost to implement, manage and transform the business will potentially be going into someone else's pockets. Larry's not going to be taking all of that to the bank.
They chose Insight Direct and Evosys as integrators, which is like hiring a cryptobro as your retirement planner. They were absolutely boned from the start, at every stage, and frankly the idiots who made the decision to go this way need to be held to answer with more than just being voted out.
The people who work with Oracle are all in on the scam!
And then there is the Oracle "evangelists" who get hired by these companies/organizations who push Oracle solutions like it's some kind of religion!
We've had them in our company, luckily we saw them for what they are and got rid of them (unfortunately not before they cost us millions!).
Every time this sort of debacle happens I'm reminded of the cartoon where the owners of a company are looking at, on the one side, huge computer stacks tended by white-coated acolytes, and on the other, an old guy hunched over a desk.
Caption: "You mean we need all that just to replace Fred?"
How about some ideas about what an alternative solution would look like? Something plausible, not the usual "I could write this in my shed with a couple of mates" type of answer.
There are a few challenges to consider,
a) This is a big enterprise, a multi billion £ "business" ;
b) Whilst on the surface it appears there are many local authorities doing similar stuff, the reality is that there are major differences between how each and every one operates. Each need a bespoke solution. "Make them all work the same way" is about a realistic a plan as "just stop crime" is as an answer to prison overcrowding.
c) there is a reason the same suppliers get asked every time - they are running these systems already. whilst there are well publicised issues, a huge amount of local and national government stuff totally depend on them to function.
d) The depth and breath of functionality required is huge. It is hard to visualise how much these systems do if you've no experience with them.
e) the customers/end users lack the skills or understanding to support these projects. Keeping the lights on day to day is challenge enough. This is both in business and IT skills.
f) there are existing legacy systems that have to be kept running whilst the new project is done. Migration of these to new processes/systems is high risk, high complexity and time critical.
g) doing nothing is not an option. Existing Hardware is approaching end of life with no like for like replacement. Software and even operating systems are no longer supported. Many existing processes can't manage the changes that will be required to support new business requirements.
Each need a bespoke solution. "Make them all work the same way" is about a realistic a plan as "just stop crime" is as an answer to prison overcrowding.
If a bespoke system is really needed then it needs to be written from scratch. Although it may not be on the few mates in a shed scale that amoutns to the same thing only bigger. It doesn't seem to be what's happening here. According to the article they're starting with a specific ERP product. If a bespoke system is really needed then it would appear to be a case of "If that's where you want to go I wouldn't start from here."
HR should be entirely off the shelf. Every employer in a given jurisdiction needs the same stuff - salaried, hourly, casual etc. employees exist in every organisation of reasonable size. None of that is new or even any different to any other UK organisation.
The numbers vary, but not the legal requirements.
Financials - again, should be off the shelf. Every organisation has essentially the same financial tracking needs.
Yes, some things have different names - "council tax" instead of a "subscription" - and some are due on slightly odd cycles like only 10 months of the year, but there's nothing actually new or different about any of it. There's bills to pay, loans, debts and bonds to track, and income to monitor. Doesn't matter whether the "customers" get to choose their subscription rate or not.
The stuff that's different about local government was all entirely out of scope! This isn't planning, or bylaws etc, it's just HR and finances.
Scale is irrelevant really.
Every actual database system can scale this large. (Obviously not Access, don't use a toy)
I suspect the real problem is that nobody within the buying team has any actual coal-face experience of any other organisation, and so buys into a salesman saying "You're different and special so need expensive customization".
You're not special. You're just big.
I would also say that where things are different whoever speced up the RFQ didn't have and domain knowledge. I remember one company who'd had consultants in where there was one line "flexible invoice payment terms required" which during implementation had to be upgraded to a 15 page document.
I agree with the above comments. Having been involved in some quite big corporate system developments, all too often companies state their requirements based on what they do. Which leads to the inability to use any off-the-shelf software as-is because it doesn't do things exactly how the company has always done things.
The answer should be to challenge how things have been done and consider whether it would be prudent to change how things are done to how the software wants things to be done. Some people will claim heresy at this point but the reality is that the bespoke development (including customisation/ehnahcement of off the shelf packages) is typically many times more complex than initially thought (often because the people who are in charge of using the system are not skilled/experienced in explaining and detailing the requirements and further more, you're left with a highly customised package, more or less beholden to continuing to use whoever the developers were who did the modifications because only they understand it, and all leading to a cost hugely more over the lifetime of the software than simply using the software off the shelf and changing one's practices accordingly.
Bespoke software is costly - from requirements capture to ongoing support. Off-the-shelf software is cheaper in the short-term and the long-term with lower maintenance & support costs and the possibility of upgrades over time.
For what should be relatively standard processes such as HR & Finance, organisations should really consider whether they have to have custom software or whether they should change their practices to conform to the software.
"Financials - again, should be off the shelf. Every organisation has essentially the same financial tracking needs."
This is absolutely incorrect!
Government is "not for profit" The accounting rules/regulations are absolutely different! Business is geared towards profit, government/non-profit is geared to spend it all! (yes I am being sarcastic but you get the gist!)
Not every for profit business can use an general financial system. A system geared to manufacturing/sales cannot be used for Construction. We in construction need We in construction need job costing, the ability to convert estimate items to cost code items for a job. Subcontractors, joint venture partners. Payroll is an entirely different animal from other businesses!
"... a salesman saying "You're different and special so need expensive customization"."
First, an admission: I spent the final 21 years of my working life implementing Oracle ERP & HR for three different integrators. It was quite common for a client to insist that they were unique and that they couldn't possibly change their processes to fit the software. Seemed a very prevalent belief in public sector organisations. Sometimes we persuaded them to do so (yes really! Process re-engineering also generates revenue), other times we simply earned more money developing bespoke code (as amended by the many change requests from the client when they realised that what they asked for isn't what they needed - alas, the customer is always right).
Ah, the good old days ...
Even in retirement it appears to be beer o'clock.
b) Whilst on the surface it appears there are many local authorities doing similar stuff, the reality is that there are major differences between how each and every one operates. Each need a bespoke solution.
f) there are existing legacy systems that have to be kept running whilst the new project is done.
On the surface they _are_ doing similar stuff. Instead of a complete rewrite from scratch, refactor with a facade design pattern. So your surface/facade is the eventual new system, you start with an ETL between the legacy systems and the new, then the legacies are replaced one at a time under the covers. And here is the key step, you document the transformation from legacy to facade so the next council can do the same transformations on their own majorly different versions of the legacy systems. One reason this can work is because the councils are not in competition with each other, they should be willing to share best practices and even working code. You couldn't write this in your shed with a couple of mates, but you _could_ do a proof of concept starting with a small council and then move to progressively harder cases. Open source the darned thing and ten years from now it won't be obsolete.
As the previous owner of a literal 'three blokes in a shed' company writing software for local authorities, we did indeed start with a single department at a single authority as proof of concept. We handled HR, payroll, resource requests and allocation, staff assignment, training, data gathering and reporting, custom reporting for partners. Everything was customizable to deal with changing requirements.
What we found when trying to scale, to departments doing the same in other authorities, or other departments within the same authority, is that not only are all authorities actually in competition with each other, but all departments within an authority are in competition, they are all little fiefdoms with their own set of legacy processes and requirements and external software to integrate with that they don't want to change because they are the 'best practice' and everyone else's way is wrong.
The costs to scale were more than we could afford, and we couldn't afford more without scaling. We couldn't go for more lucrative projects because we were 3 blokes in a shed, and no financial controller is going to sign off on a million quid project with 3 blokes in a shed when they can sign off on 10 million with oracle, salesforce, dynamics or SAP. We ended up doing about 70 POC/ pilots with various authorities and companies, most of which ended up as "thanks, now we know it works, we're going to take the detailed requirements you produced to oracle/ dynamics/ SAP/ salesforce'.
We made a comfortable living for 6-7 years, and eventually sold everything to crapita, who were running most of the integrations that got past POC into production, who from what i heard, quietly shut it all down and moved to their own custom stuff...
>> What we found when trying to scale, to departments doing the same in other authorities, or other departments within the same authority, is that not only are all authorities actually in competition with each other, but all departments within an authority are in competition
They're in competition for budget, mainly, but councils tend to attract people who want to build kingdoms. The type of problems you see in all public sector IT projects are almost always down to the pettiness of the people they're being developed for, but also...
>> they are all little fiefdoms with their own set of legacy processes and requirements and external software to integrate with that they don't want to change because they are the 'best practice' and everyone else's way is wrong.
What they mean when they say this is really that they've learned a way to do it and they don't want to have to relearn how to do their jobs. I sympathise with this enormously. Nobody wants to have to be retrained to do admin; after we've been doing it for a while we've learned exactly how much of it is essential and how much can be ignored. With a new system you have to start again. That's why people want a system that works exactly like the old one, and why you get so many late requirements when people realise they can't use all their old dodges to avoid the drudgery.
This is *exactly* what the Cabinet Office - formerly Crown Commercial Services - formerly CCTA - should be doing across council, education, NHS, government etc is having a straightforward *just works… aspire for well” standard solution that can scale from small district council to Birmingham sized… not endless pointless rounds of procurement frameworks.
Despite all the bullshit talk of bespoke … they all do the same thing, with the same customers, interact with the same other councils and other government agencies, collect the same council tax payers, the same local rates, the same education, social care and NHS interfaces, have the same suppliers, run the same HR and Payroll, are responsible for the same Highways, do the same statutory duties like Environmental Health, run the same parks services, run the same refuse services… etc
Standardise business processes, standardise workflow, standardise API’s, don’t have a behemoth ERP solution do it all - use standard best in breed solutions and interface them together, have a common website presence, have an app, true shared services.
Another train wreck - another council, another university… more on the way. Oracle, SAP, Workday and integrators enriched !!
annoyingly large private business has the same issue.
you'll get people who refuse to change. Who've worked the same way for 20-30 years and just won't change but CAN make enough of a noise to ensure that nothing changes.
to be fair, you get the same in IT. Try to get a Cisco house to go Juniper or someone else. A bunch of Cisco guys who've touched nothing but Cisco or a certain type of firewall or a certain type of database, they've worked on this for 20 years + and refuse to learn how to use anything else, regardless of cost or features.
It's bloody annoying. IT people are NO less stuck in their ways than any other department, regardless of what you read here on the Reg. Unless you have a VERY pushy architect or manager who'll say....I don't care that you've been working on that for your whole career...we're going with something cheaper, better, faster, more modern with better support.
Personally I would introduce new tech, just to keep an IT Department sharp and motivated, rather than letting them get lazy and used to the same old things. Better for their CV's and less likely to end up with the kind of person that you see in MANY large IT Deparments in their late 50s who are literally just there waiting for their pensions
Standardise, yes. But...
Best of breed usually failed when they went for things that didn't integrate easily. Or at all. Irrespective of the combination of vendors. That's why most of the adoptions I see these days are single vendor for as much as you can possibly manage, and damn the compromises.
And in the SAP world, it was thought to be cheaper to change your company to fit SAP than to change it in any way, but alternatives were limited for a very long time. They're charging over to the cloud as fast as their anchor will let them.
As far as things like payroll are concerned, my expert colleague from Oracle uk developing a COTS payroll solution took one look at Australia's penalty rate legislation and decided the entire market was too hard to deal with. One size fits all isn't easy, especially in the public sector.
Finally, the biggest problem with Oracle's ERP for many, many 90's customers was that the instant you modified it, you couldn't upgrade it without redeveloping (not just retesting) all your mods. It was the underlying DB schema changes that screwed you. God knows how that plays in the cloudy Fusion world, but I won't be charging back there to find out.
Ahem, that's what the Cabinet Office *tried* to do with gov.uk! They tried to get systems rationalised, more cost effective and more flexible. And every little fiefdom kicked up an almighty stink and refused to play along. I know a few people from the Cabinet Office who were there for the early years and they said it was extremely difficult to persuade some mandarins to move with the times, while it was dead easy with others. Trying to persuade departments like HMRC, DEFRA etc to work better together by looking at their processes to see if there were things that people in different departments did the same and that could be shared somehow was like trying to get an oil tanker to turn on a dime (well, we've seen how well *that* works with the Evergiven, which, granted, is not an oil tanker, but you get the gist).
So yes, government (whether national or local) suffer from the same malaise of trying to break through decades of gathered cruft (or institutional memory) to improve things, make things better, and free up funds for delivering services, rather than being swallowed up by inefficient processes.
Yes, it's all so simple. Standardise it. Why has no one else ever thought of this?
Why not setup your own company and present your brilliant plan to local authorities? I'm sure they will be blown away as you present the standard processes that you will be implementing. Along with your strategy to get their entire work force on your side to make it happen.
Your book of standard best in breed solutions for all business functions will be worth it's weight in gold. You will have no difficulty in attracting investors and finance with such breath taking knowledge. With full multi vendor support to get everything working seemlessly together the implementation should be trivial. Throw in a single website, knock up an app and it will be done.
Can you deliver next Tuesday?
Birmingham in in the UK.
Brexit has happened.
Why reference Europe ? It's the UK's biggest city council, in yet another UK ERP disaster (starting to sound familiar - are there any ERP successes in the UK ?).
It's funny how there is a dearth of ERP disasters on the continent, compared to the UK. Actually, I wouldn't be surprised to learn that the UK has more ERP issues than all continental European countries put together.
Because this Birmingham is in the UK and in Europe as opposed to at least one that isn’t.
And the UK is still (as always) in Europe as Europe is a geographical entity distinct from the European Union - Brexshit or not
And obviously it must be the UK’s biggest city council as it’s Europe’s biggest city council.
If your tomato was recognised as being the biggest in the world you’d probably say so as opposed to saying the biggest in my back garden
As for ERP failures, I don’t think I have ever seen a successful implementation in 40 odd years
And no, I am not in the UK but I am in Yurp (respect to Steve Bell)
Usually I'll tell Brexiteers to stop being dicks, but seriously Pascal, you take the cake! Stop being one. You're not doing yourself favours.
The piece didn't say "The EU's biggest city council", so this isn't about politics. It's about numbers and size. And the last time *I* checked, Pascal, the UK is a country on several islands off the coast of... guess what! EUROPE. It is still, as much as you won't like it (and many Brexiteers don't either), a PART. OF. EUROPE. AS. A. CONTINENT.
So, yes, it is entirely ok to be referencing Birmingham's size in relation to its *gasp* geographical neighbours. Get it? Geographical? See? No politics? Just geography, you know, the study of land?
And it's not about the size of ERP disasters, or the number of ERP disasters. It's about the size of the local authority, in terms of population.
So, kindly, Pascal, stop being one.
'Oracle Fusion, the cloud-based ERP system the council is moving to, "is not a product that is suitable for local authorities, because it's very much geared towards a manufacturing/trading organization"'
That was the problem with oracle's erp modules back in the early 90s. Sequent computers were the driving use case early on (almost exactly unlike nearly every other enterprise scenario, like food manufacturing, or publishing, or ...) and it was very difficult to get beyond that mind-set. Well, that plus the fact that the various modules were developed largely in isolation, eg customers and suppliers were two completely separate things, which was fine until your customer was also a supplier, when it was a major headache.
And here are 30 years later...
My wife used to work at Next Retail. They rushed in Oracle HR and chaos ensued. Problem included shifts not being correct, lack of shift visibility when you could see it. Unable to offer shifts for swaps. Timesheet errors, pay errors and leave requests being lost. And all of that whilst on minimum wage. They got a grovelling email apology from HR but by then a lot of staff moved job. She did so too as not being paid the hours you had worked is not a good look. She's been left a year now, and is still in touch with her colleagues who say it is no better.
The city budget is a tad under £4k per head if my mental arithmetic is still ok
They are still voted in because of the alternatives...
Brummies ain't thick and don't vote for Christmas if you take my meaning.
PS: previous system paid my wages ok. Current system pays my wages OK.
Most of the overruns ceasing abruptly when I asked one simple question.
"Are you comfortable with the phrase fundamental breach of contract?"
Suddenly, the cost overruns ended, date slippage ceased and progress ensued, as under our laws, a fundamental breach of contract is not only fully recoverable, but damages can be awarded. The contracting company then has to figure out, free product and services *and* pay damages or absorb their illegally massive underbid, as such is simply fraud to acquire a contract. Plus, awarded damages, losses and oh yeah, getting blacklisted as a vendor in perpetuity.
Our UK division went through he same thing back in the 00's. The (rather stupid) decision was to move from separate Construction specific Finance systems (yes we are construction) to Oracle Business System. Just like this municipality OBS had no presence int eh Construction industry and ABSOLUTELY no expertise in out business. The project was an unmitigated disaster! A decade later it is still not (as my British friends like to say) "Fit for Purpose!"
Here in the US we are on JDE, another Oracle acquisition. Now they are blackmailing is with Licensing fees! If we move to to their Cloud ERP (not gonna happen!) or move our workflows from AWS to Oracle Cloud Infrastructure they will cut out cost by 40%! After much negotiations now they will accept if we move "some" of our workflows to OCI. So we are moving out least important workflows over their, certainly NOT our JDE infrastructure, (I am sure we will regret it) and they will give us the discount! I am quite sure our costs will go backup very quickly.
It's like dealing with the Mafia! Shitty products and services and deceitful pricing!
Just one point:
Birmingham is the largest lower tier local authority by population, but the département of Paris and the Île de France region are larger.
Île de France has several départements, including Paris, and Paris is made up of a load of arrondissements which are the lower tier local authorities.
The SAP ECC solution was not customised by BCC rather delivered as close to standard by Axon Solutions Ltd. Most of the customisation was integration to other BCC applications, as usual. Service Birmingham was responsible for application management so presumably would still know a lot about how it works.
It didn't take five years to implement SAP ECC so clearly the approach is not well suited to Oracle software and the implementation partners. Similarly conversation from SAP ECC to S/4 HANA doesn't take five years so given the standard Migration tools this was likely a political decision to use another software vendor.
Clearly S/4 HANA private cloud edition ticks the cloud box on a range of hyper-scalers. So freedom from legacy hardware and data centres is readily available.
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2024
