Europe's Largest Local Authority Slammed For 'Poorest' ERP Rollout Ever (theregister.com) 71
UK government-appointed commissioners have labeled Birmingham City Council's Oracle Fusion rollout as "the poorest ERP deployment" they have seen. From a report: A report published by the UK council's Corporate Finance Overview and Scrutiny Committee found that 18 months after Fusion went live, the largest public authority in Europe "had not tactically stabilized the system or formulated clear plans to resolve the system issues and recover the operation."
The city council's cloud-based Oracle tech replaced the SAP system that it began using in 1999, but the disastrous project encountered a string of landmark failures. The council has failed to produce auditable accounts since Oracle was implemented in 2022, costs have ballooned from around 19 million pound to a projected estimate of 131 million pound and, because the council chose not to use system audit features, it cannot tell if fraud has taken place on its multibillion-pound spending budget for an 18-month period. In September last year, the council became effectively bankrupt due to outstanding equal pay claims and the Oracle implementation.
The report from "best value commissioners" appointed by central government to investigate struggling councils said that following the Oracle implementation, "a serious lack of trust had developed between members and officers driven by the failed implementation and subsequent lack of progress to resolve the situation."
The city council's cloud-based Oracle tech replaced the SAP system that it began using in 1999, but the disastrous project encountered a string of landmark failures. The council has failed to produce auditable accounts since Oracle was implemented in 2022, costs have ballooned from around 19 million pound to a projected estimate of 131 million pound and, because the council chose not to use system audit features, it cannot tell if fraud has taken place on its multibillion-pound spending budget for an 18-month period. In September last year, the council became effectively bankrupt due to outstanding equal pay claims and the Oracle implementation.
The report from "best value commissioners" appointed by central government to investigate struggling councils said that following the Oracle implementation, "a serious lack of trust had developed between members and officers driven by the failed implementation and subsequent lack of progress to resolve the situation."
Oracle Software (Score:3, Insightful)
Re: (Score:3)
I've been involved in dozens of database implementations, two of the Oracle. Both were multimillion dollar multi-year fiascos, and both were replaced within a few years (by SQL Server and Informix).
Re:Oracle Software (Score:5, Interesting)
I have worked on Oracle ERP implementations for over 20 years
There are a few things that can result in poor outcomes like this
1. Incomplete requirements
2. Demanding more customizations than the software can deliver without coding changes
3. Incomplete test coverage
4. Unbending timelines that do not take items 1, 2 and 3 above
While I have worked with Oracle products since the mid-1990, I have also been part of SAP delivery teams
In the late 1990's. the rules for SAP implementations was NO CUSTOMIZATIONS, change your business to fit the software
That was a good rule back then, and Oracle ERP offered a lot of customizations (from a large range of configuration changes to accommodating custom coding at the pl/sql level), it was a big selling point for Oracle over SAP...
HOWEVER Oracle Fusion is a different beast, and most existing Oracle ERP implementations that move to Fusion have to be stripped of customizations, particularly at the pl/sql level
It is hard to tell what happened here, maybe the people selling Oracle ERP over SAP oversold customization, maybe the customers were tired of being in the SAP no-customization trap, and demanded more customizations than were necessary
However, in most cases like this, requirement changes were identified late in delivery, testing was cut short to meet existing timelines and after each failed delivery, the project was extended to meet these needs, over and over and over again (each time adding more requirements)
Frankly, Oracle ERP is a solid product that is used widely without these sort of errors, and poor project management and unknowledgeable leadership MUST be considered as root causes
Re: (Score:1)
Oracle ERP is a solid product that is used widely without these sort of errors
The software "just works". It is the customizations and shitty implementations by VARs that make it suck donkey balls. The whole notion of a "value added reseller" is a false name. Oracle has ruined its' reputation by letting others implement its software. It should have maintained an international team of implementers to keep fiascos like this from occurring. That being said, some of it is absolute shit. I recall an EBS implementation sold with Oracle Transportation Management where the middleware was comp
Re: (Score:2)
There was a lot of that going on in the biz world. 20 years ago I wrote something that was supposed to compete with MS and IBM and Oracle etc etc, which had a lot of BPML and stuff that MS BizTalk was supposed to be able to do, except it was all open source, written in Java, ran on anything, had satellite integration and IOT integration and integration with any protocol/format you could imagine. Sounded like a good idea at the time. WTF did I know, I was like 24. I rage quit when they tried to make me a tra
Re: Oracle Software (Score:2)
My experience is that off the shelf unmodified software solutions that require your business to change how it operates often result in a working but not optimal business.
Software solutions built from the ground up bit by bit adapted to how the business operates are more efficient, but harder to achieve.
I'm unfortunately in a business where SAP is used and the UI is clunky and not very intuitive or user friendly. You can do what you want - mostly, at least enough to fulfill the basic needs. But it takes time
Two Things (Score:3)
1 - All ERP requirements are incomplete / inaccurate. No exceptions.
2 - Oracle will NEVER tell you this until the ink is long dry and it's time for the change requests and extensions to roll in.
Pushing the blame downstream is both honest and dishonest. The PMs and business experts can be held responsible, but Oracle knows it has a 60% chance of failure no matter what. But they'll get paid.
Re: (Score:2)
>>1 - All ERP requirements are incomplete / inaccurate. No exceptions.
Are you familiar with the Oracle AIM methodology?
It comes down to there being a finite number of configurations and well structured processes, BUT most organization will not spend the money on completing them, and instead shove off the responsibility to IT team members, which results in the user shock when they set down to the new system and cannot use it to perform their jobs
As a counter argument, I worked at a FDA regulated compan
Re: (Score:2)
There are cases where the cost resembles the estimate, the timelines remain under control, the benefits are realized, and the business evolves positively. When you capture that unicorn, build it a beautiful paddock and admire it every day, because you aren't likely to see another. Even "success" usually means you dragged the carcass over the line, got the signatures in place to close out the paperwork, and left a bunch of unaddressed processes and gaps that will haunt the org for years.
I'm not saying Oracle
Re: (Score:1)
Re: (Score:3)
Re: (Score:3)
That bus left the station long ago, the time to **start** firing people for recommending Oracle would have been in the mid-'90s, anyone recommending it since almost certainly has a financial interest in its implementation.
Re: (Score:2)
Re: (Score:2)
IMO, Oracle simply deprecates older software versions and forces customers to upgrade, although this is done a relatively long timeframe
When Oracle _really_ wants to increase their income, they institute 'software licensing audits' that typically generate a few million in new licensing fees from customers
Re: (Score:1)
Why am I not surprised?
Because of your observer bias? Big projects tend to Oracle or SAP. Big projects tend to end up in the news when they fail. Governments tend to fail harder than anyone.
The problem here isn't Oracle, it's the government. Yeah nothing is ever smooth, but a 500% project overrun isn't the fault of Oracle, that is a clear fundamental failing from the very VERY inception of the project all the way through to execution.
Re: (Score:2)
Re: (Score:2)
Scott Adams said that that no matter how absurd he made a storyline in 'Dilbert' someone would email him asking whether he had been corresponding with people in their company.
Not a failure (Score:2)
Re: (Score:3)
As someone who has watched (by being employed as an EE that sometimes had to get info into or out of the respective ERP system and worked closely with the finance departments) multiple companies with differrent ERPs from completely different vendors sink ungodly amounts of money and effort into subpar software customizations, workarounds and absolute dumpster fires of barely working processes... the common factor here is most definitely the ERP software vendors. Even the smaller ones are a hellscape of upse
Re: (Score:1)
I read the title and my brain suggested Oracle. Two lines down. W00t!
Re: (Score:2)
After all (Score:5, Informative)
Company - "How much will this cost us?"
Oracle - "How much money do you have?"
Re: After all (Score:4, Funny)
Company: we want a system that will send messages between employees
Oracle: okay, thatâ(TM)ll cost you £1,000,000
Company: here you go
Oracle: hereâ(TM)s your system.
Company: hang on, this doesnâ(TM)t work, none of the messages are getting through!
Oracle: ohhhhâ¦. You want to be able to *receive* the messages as well as send them? Thatâ(TM)ll cost you £100,000,000
Re: (Score:2)
Not really. Yeah most Oracle projects will overrun (most ERP projects in general do) but Oracle can't account for the 600% overbudget in this case. This requires coordinated idiocy at every level of the project by all parties.
Re: (Score:3)
Not really. Yeah most Oracle projects will overrun (most ERP projects in general do) but Oracle can't account for the 600% overbudget in this case. This requires coordinated idiocy at every level of the project by all parties.
Well, you might be correct - my experience has been universal (of those I know). And when it takes 5 years with the meter running on a project that was budgeted for 2 months, you'll forgive my people for agreeing with me Businesses and organizations expect at least to be within a year of the promised finish to function.
\ I'd be a real fool to recommend Oracle knowing that.
Re: (Score:2)
My experience is universal as well... with all vendors. Even the best projects tend to overrun. The biggest problems usually are that people select an ERP solution that doesn't fit their corporate work processes. The second you try to customise everything. ... how do I say this gently with the maximum sugar coating ... YOU GET FUCKED! Yes that was the gentle version.
When doing a major ERP solution you need to customise the business processes to fit the product, not the other way around. Otherwise your proje
Re: (Score:2)
you need to customise the business processes to fit the product
Then implementing ERP products as a class are contrary to a business's best interest. Businesses have processes that work a certain way because that's what works best for the business, and frequently this gives them their advantage over the competition. Having to do things the same way as everyone else reduces or eliminates this advantage for minimal or no (or negative) payback.
Fusion? (Score:3)
I thought Fusion was 30 years away...
(of course it doesn't help that they don't say what an ERP is) Executive Retirement Plan maybe?
Re:Fusion? (Score:5, Funny)
I'm reasonably certain (but not 100%) that it's not Erotic Role Play.
Re: (Score:3)
Re: (Score:3)
It doesn't even sound as sexy as Enterprise Resource Planning.
Re: (Score:3)
I thought Fusion was 30 years away...
(of course it doesn't help that they don't say what an ERP is) Executive Retirement Plan maybe?
Yeah - even the original article doesn't. But as soon as I saw Oracle - I thought Come on, this is how every Oracle install I ever saw went. Where I was at last, the Oracle implementation was supposed to be 2 months. It still was being worked on 5 years later.
And it is in Oracles monetary interests for it to take as long as possible.
Re: (Score:3)
The very first Oracle ERP implementation that I was involved in had been grossly underbid by the original consultant, and we had to step in, clean up their failures, and then follow a mature implementation plan
The biggest problem is that any Tom, Dick and Harriet can CLAIM to be an oracle implementer, get on a bid list and low-ball the estimate to get the job
Then, they realize just how deep they are in and take their money and walk out
It ALWAYS takes years to implement Oracle ERP, and the organization getti
Re: (Score:2)
T
IMO, and attempt to compare Oracle ERP to a simple software install plainly ignores the complexity of the systems involved
But they are happy to sign you up.
Re: (Score:2)
That was an example I watched awhile back. The 'consultants' (husband and wife team) ended up being hired on full time after the first two years. the project went on so long the head of IT retired and one of the consultants took over the entire IT department.
When they first showed up, the business bought them a house on the beach. I wonder if they still have it?
I'm sure it exists... (Score:1)
It seems this is the SOP; "task failed successfully".
Re: (Score:2)
Since my IT career began years and years ago, I can't recall a single time I've read about Oracle ERP installation or conversion success stories which came in "on budget" or "without utterly destroying something". It seems this is the SOP; "task failed successfully".
It is designed to fail, as long as the suckers H^H^H^H^H^H^H companies keep sending money to Oracle.
Customizing ERP frequently fails (Score:4, Insightful)
From the article: The council initially customized Oracle but now plans to reimplement the software out-of-the-box, adopting standardized processes.
From what I've heard, you conform to the way the ERP does things, not vs/vs. Attempting otherwise for anything other then cursory detail is where you end up fighting hideously complicated software into oblivion.
Re: (Score:2)
Hush, let the kids play with their anti-corporate narrative.
Re: (Score:3)
Microsoft Dynamics 365 implementation specialist here.
From what I've heard, you conform to the way the ERP does things, not vs/vs. Attempting otherwise for anything other then cursory detail is where you end up fighting hideously complicated software into oblivion.
A good ERP implementation is about rethinking how the company does business, then remaking the business processes to work with the ERP system. Understanding how different ERP systems work is just as important when selecting a system as the final price is.
Regarding customizations, no single OOB ERP is a 100% match for any company, so things need to get added or tweaked. For example, a company may have a preference for a specific tax processor, so the nece
Re: (Score:2)
Yep. Used to work for a Navision integrator before moving to embedded development. Any ERP works only as well as the guys who do the ERP customisation did their job. If they suck, the resulting ERP environment will suck, too.
Re: Customizing ERP frequently fails (Score:1)
It isnâ(TM)t always that simple, in many cases you are bound by law and regulation to certain procedures which vary from company to company. For example if you are a union shop, your rules on how much can be processed on any one day by any one person must be encoded, and those are negotiated and often contradictory. Bring in your HR system and youâ(TM)re now stuck with hundreds of codes for vacation time, as if anyone cares (none of my peers, underlings or bosses do) why I take PTO, well, apparent
Re: (Score:3)
From the article: The council initially customized Oracle but now plans to reimplement the software out-of-the-box, adopting standardized processes.
From what I've heard, you conform to the way the ERP does things, not vs/vs. Attempting otherwise for anything other then cursory detail is where you end up fighting hideously complicated software into oblivion.
I work for an ERP vendor. A common thing we see if that customers often want their new ERP system to behave like whatever their previous system did, or what what some employee got used to at a previous job. This leads down the path of extensive customizations, which are often not only unnecessary, but lead to continuing problems going forward such as making upgrades far more difficult than they should be.
We always advise customers to test thoroughly with out-of-the-box functionality before concluding they n
Re: Customizing ERP frequently fails (Score:3)
The customer wants to dictate what the software theyâ(TM)re contracting you to write does? How dare they dictate the specs!!!!!
Re: (Score:2)
The customer wants to dictate what the software theyâ(TM)re contracting you to write does? How dare they dictate the specs!!!!!
You're missing the point. Contracting people to write custom software is extremely expensive, and we encourage customers not to undertake that unless proven necessary. It's human nature to be resistant to change, and understandable why people may want things to look like their previous system. But it's the job of any responsible ERP implementer to sometimes push back against customizations to prevent the scope and cost of an implementation from ballooning. Contrary to what you may think, customers don't alw
Re: (Score:2)
Customizing ERP frequently COSTS YOU. (Score:2)
From the article: The council initially customized Oracle but now plans to reimplement the software out-of-the-box, adopting standardized processes.
From what I've heard, you conform to the way the ERP does things, not vs/vs. Attempting otherwise for anything other then cursory detail is where you end up fighting hideously complicated software into oblivion.
It likely wasn’t just the complexity of customization that scared the shit out of them. It was the licensing practically begging for a violation. Using Oracle modules in a very vanilla way? No problem. But the instant you step out of bounds even slightly, the other side of Oracles face emerges. The Litigation Department. Also known as where the real money is made.
Naturally the software is designed to allow you to easily violate all that shit you never read in the contract.
Re: (Score:2)
My company writes ERP software (business accounting software if you are unfamiliar with the term ERP - my company website has a good description https://www.eaerich.com/ [eaerich.com] ). While changing business processes can be required, it is also important to remember that often those very business processes are what set a business apart from the competition. I'm not talking about creative accounting, either.
In order to survive a business needs a competitive edge. That can include products and services, and can also
Re: (Score:3)
Tesla was looking at implementing a mandatory SAP upgrade, estimated to cost gazillions and take a stupid amount of time. Instead they wrote their own in under six months and tossed their standard ERP software "solution" out the window. I imagine the conversation going something like:
IT Director: I swear, we could probably write our own software to replace this junk in half that time and a tenth of that cost.
Musk: You think so?
IT Director: Umm, probably.
Musk: OK, go do it.
IT Director: Oh, shit.
But it hasn't killed anyone yet... has it? (Score:3)
Maybe a poor rollout, but so far, it doesn't appear to have killed anyone? Unlike say, the Horizon accounting software the UK Royal Mail rolled out a couple of decades ago?
Oracle... (Score:2)
Coulda googled (Score:2)
Re: (Score:2)
You linked my absolute favorite bit of Oracle sleaze. This is a shady win beyond the wet dreams of hustlers everywhere.
As part of a settlement, Oracle gives Oregon 6 years of unlimited licensing for their biggest, most expensive tools. That's 6 years to hang yourself with a true up that's going to strangle you for decades. I wonder what happened in 2022?
There are two kinds of people... (Score:3)
Re: (Score:3)
Anybody can google ERP and get 'Enterprise Resource Planning', which still is waaaaay to high-level to be useful
Oracle ERP is Financials (General Ledger, Accounts Payables, Account Receivables, etc...), Human Resources (Base HRMS required for all other functions, but can be extended to include Employee self-service, Manager self-service, Time Management, Recruitment, Internet Expenses, etc...), Purchasing (iProcurement, Fixed Assets, etc...), Manufacturing (Bills of Materials, Inventory, etc...), Grants Acc
Re: (Score:2)
I totally believe there is a huge need for people who can support it. In my case, that stuff is way outside of my wheelhouse. That would be (as Monty Python would say) "time for something completely different".
Re: (Score:2)
We got in there, and up on the whiteboards were pasted approximately 200 to 300 sticky notes, with marker lines drawn between them. I mean, LITTERED across the whiteboard covering the long wall of the conference room.
"Is that a project schedule?"
"Not exactly, it is t
Re: (Score:2)
Re: (Score:2)
Once upon a time, I was a land survey tech. Did everything from boundary surveys to construction staking.
Forty years later, wondering how long I can ride the Oracle gravy train until I want to retire
In between were steps in CAD, GIS, CASE, SQL, PL/SQL, ERP and all of the environments and technologies to support them, eventually going back to school and getting an MBA so I could understand accounting software, and the customers who use it, better
Along the way, I found out that I dislike modern business pract
We need healthy competition between Intel and AMD (Score:1)
AMD's ascendency is no better for us than Intels was back in the days of Bulldozer and friends.
Roll your own! (Score:2)
20 developer x 200k / year * 5 years? 20 000 000 or 20 million.
Even if I add in another 10 people, that's 30 million, where is the other 100 million? You don't need auditing to realize you're either committing fraud intentionally, or committing it through gross incompetence. You can stand up this kind of system in ~1 year, generally, and you don't need 30 people, so even with ideal numbers, that are government padded, where is the 100 mi
Re: (Score:2)
Re: (Score:2)
Re: (Score:2)
Uh, oh, maths. You're going to confuse the 8-digit UID posters.
Re: (Score:2)
Typical Oracle (Score:2)
And, who goes to prison? (Score:2)
For massively wasting taxpayer money and possibly corruption? Nobody? Then crap like that will simply continue. Apparently money is plenty and there is no reason to be careful with spending.