Birmingham City Council gone bust (failed ERP implementation)

viz

Fascinating story about Birmingham CC's failed Oracle Cloud ERP system; I'm sure there are plenty of folks on here with experience of ERP implementations. I wonder what's going to happen now - we still need a city council.

monquixote

This is what happens when non technical folks get tied in knots by enterprise IT folks.

viz

Should BCC have their own in-house permanent digital / enterprise team? I understand Evosys / Mastek are the SI.

theatreanchor

Woking too. What a mess. 

monquixote

I'm sure it would help.

It's not my area of expertise but

I find the whole idea of these single vendor IT transformation mega projects weird and set up for massive cost overruns / failure.

Sporky

theatreanchor said:

Woking too. What a mess. 

Woking was weird. It looks like the previous councillors spaffed several billion on a vanity project and then all quit. 

stickyfiddle

I feel like I've heard far more stories of utterly bungled Oracle/SAP/etc implementations costing obscene amounts than I have successful ones. I realise there's obviously a massive selection bias there but still. Shouldn't this stuff be relatively well-understood at this point?

And who is signing contracts that commit to spending multi-millions without a clear plan for delivery (and importantly..  evidence of delivery)?? 

exocet

stickyfiddle said:

I feel like I've heard far more stories of utterly bungled Oracle/SAP/etc implementations costing obscene amounts than I have successful ones. I realise there's obviously a massive selection bias there but still. Shouldn't this stuff be relatively well-understood at this point?

And who is signing contracts that commit to spending multi-millions without a clear plan for delivery (and importantly..  evidence of delivery)?? 

The overruns are certainly not unique to Public Sector, I’ve worked in Private sector for 25 years, current company has been spaffing  millions with Oracle for years and yet the system can never deliver what is required…always one more upgrade required….

theatreanchor

Sporky said:

theatreanchor said:

Woking too. What a mess. 

Woking was weird. It looks like the previous councillors spaffed several billion on a vanity project and then all quit. 

Arrr. And the Tory Council were recently ousted by the Lib Dems. So much fun. 

ewal

viz said:

Should BCC have their own in-house permanent digital / enterprise team? I understand Evosys / Mastek are the SI.

I'd be 99% sure they do have and they'll be the ones getting the blame, even though it's likely decision will have been made by business leaders.

Business leaders have raised the ERP subject at the council I work at, normally because someone like an Oracle sales executive has told them they should.

We tell them to go away because 1. Council management culture is 'inappropriate', 2. Need to fully understand the cost of change.

Up until now they've listened to us

CHRISB50

I’m an SAP consultant. 

I’ve seen projects drag (particularly implementations), and costs increase, but never anything like this. 19m to 100m is totally mad. 

I remember the BBC canned an SAP implementation as it was taking too long and cost too much. 

Someone had their head in the sand. 

monquixote

stickyfiddle said:

I feel like I've heard far more stories of utterly bungled Oracle/SAP/etc implementations costing obscene amounts than I have successful ones. I realise there's obviously a massive selection bias there but still. Shouldn't this stuff be relatively well-understood at this point?

And who is signing contracts that commit to spending multi-millions without a clear plan for delivery (and importantly..  evidence of delivery)?? 

What you have to understand is that this is by design.

As a certain type of supplier you sign off a project for a nice low number knowing the client hasn't understood the complexity and they will find themselves in 12 months with a system that exactly meets the terms of the contract but doesn't do anything useful. At this point they are several million pounds in the hole and have no choice but to continue because you have total vendor lock-in.

On the vendors side you have technical sales and legal teams who are on £300k+ / year and do this kind of trick for breakfast lunch and dinner and on the other side a bloke from the council.

Unfortunately because of the way procurement works at big institutions they always want single vendor systems than do everything delivered in one massive lump which is is exactly what every book on software development tells you not to do.

MrSwanson

When I was working in local gov finance, within 9 years I saw 3 flawed ERP implementations. For prominent London Boroughs too....

stickyfiddle

monquixote said:

stickyfiddle said:

I feel like I've heard far more stories of utterly bungled Oracle/SAP/etc implementations costing obscene amounts than I have successful ones. I realise there's obviously a massive selection bias there but still. Shouldn't this stuff be relatively well-understood at this point?

And who is signing contracts that commit to spending multi-millions without a clear plan for delivery (and importantly..  evidence of delivery)?? 

What you have to understand is that this is by design.

As a certain type of supplier you sign off a project for a nice low number knowing the client hasn't understood the complexity and they will find themselves in 12 months with a system that exactly meets the terms of the contract but doesn't do anything useful. At this point they are several million pounds in the hole and have no choice but to continue because you have total vendor lock-in.

On the vendors side you have technical sales and legal teams who are on £300k+ / year and do this kind of trick for breakfast lunch and dinner and on the other side a bloke from the council.

Unfortunately because of the way procurement works at big institutions they always want single vendor systems than do everything delivered in one massive lump which is is exactly what every book on software development tells you not to do.

Oh of course - I was a consultant for 9 years! And now I work in infrastructure procurement for (non-UK) government. I understand it far better than I'd like to tbh! 

Sunk cost fallacy is a huge thing... 

nero1701

It's often down to the specification provided to the ERP provider, once implementing the specification begins, they often realise more is required. 

TimmyO

I read that the Oracle project was approved based on the fact that procedures were to be changed so as to fit a pretty OOTB oracle implementation.

but they then left the processes the same
and started having the solution customised to fit. Cue money pit. 

danishbacon

nero1701 said:

It's often down to the specification provided to the ERP provider, once implementing the specification begins, they often realise more is required. 

The world needs more BAs. BAs + adequate time allocated for thorough analysis = better requirements. Then you go to tender.

Chalky

TimmyO said:

I read that the Oracle project was approved based on the fact that procedures were to be changed so as to fit a pretty OOTB oracle implementation.

but they then left the processes the same
and started having the solution customised to fit. Cue money pit. 

The old "we need the system to match how staff do things" instead of change your staff or their mindset.

Sporky

danishbacon said:

The world needs more BAs. BAs + adequate time allocated for thorough analysis = better requirements. Then you go to tender.

At which point people like me (but evil), who respond to more tenders a year than most BAs will create in a career, look for any opportunity to cheat and win.

Doing one for a university at the mo; by ignoring the BOQ (created by someone with no significant expertise) and going for the functional spec I can save them 10% with better quality and future proofing. Careful reading of the tender documentation, and careful answering of the questions means my response counts as compliant even though there isn't a single item they specified.

If I was evil I could game the scoring system giving them something useless. Tenders are not a good way to buy anything. 

monquixote

danishbacon said:

nero1701 said:

It's often down to the specification provided to the ERP provider, once implementing the specification begins, they often realise more is required. 

The world needs more BAs. BAs + adequate time allocated for thorough analysis = better requirements. Then you go to tender.

Totally agree.

Sorely undervalued skill. I have a brilliant one on my team and he's worth his weight in gold.

danishbacon

Sporky said:

danishbacon said:

The world needs more BAs. BAs + adequate time allocated for thorough analysis = better requirements. Then you go to tender.

At which point people like me (but evil), who respond to more tenders a year than most BAs will create in a career, look for any opportunity to cheat and win.

Doing one for a university at the mo; by ignoring the BOQ (created by someone with no significant expertise) and going for the functional spec I can save them 10% with better quality and future proofing. Careful reading of the tender documentation, and careful answering of the questions means my response counts as compliant even though there isn't a single item they specified.

If I was evil I could game the scoring whole giving them something useless. Tenders are not a good way to buy anything. 

I may have oversimplified my statement. The BA is a piece of a wider puzzle and not every BA is created equal, a junior with tool knowledge shouldn’t be speccing up big projects solo. If the puzzle doesn’t have all the right pieces (proper PM, Senior buy-in with appropriate guidance and evidence, resourcing for implementation and a technical reading of the tender responses to identify and sift through the bullshit) then it’s going to fall apart in one way or another. Whether it falls apart big time or not depends on the complexity and dumb leading the blind ratio.