A Correction To The Title?
West Sussex’s Debacle rollout pushed back again as costs balloon 15 times…
West Sussex County Council has once again delayed the implementation of Oracle Fusion for HR and payroll – set to replace an aging SAP system – following a series of setbacks that have seen expected costs swell to more than 15 times the original estimate. In a notice to schools, the council said the program team had reviewed …
"This revised timeline reflects the scale of the change, impacting over 20,000 users, and the need to ensure a safe, stable system with a reliable and accurate payroll, particularly for schools and term-time employees,"
Are they saying that they started the project without the need to ensure a safe, stable system with a reliable and accurate payroll, particularly for schools and term-time employees?
The situations here and in Birmigham are absurd. 90%+ of what all councils do is the same and if it isn't it should be. The government should hire** and run a central, dedicated unit to define, spec., procure, install, train, learn from best practice, etc. This way they'd have the clout to get better pricing and terms from suppliers. Councils that genuinely need customizations could be handled by the same team which would rapidly build up the expertise to judge whether they were really special or just needed help to change their processes to the standard ones.
</dreaming>
**and, obvs, pay the going rate.
This way they'd have the clout to get better pricing and terms from suppliers.
It's all very well defining, procuring, installing, shouldn't the remit be to ensure that local and/or national government own the IP and the design? The customer base for UK local government is (in the longer term) simply the intended unitary councils that the current slow moving reorganisation is moving towards, so we're talking about about only 45 organisations across England. Trying to procure for that number of customers isn't going to hit any economies of scale with the potential vendors, who will be all the usual suspects.
Whilst it's even further wishful thinking, I'd adopt your proposal in its entirety, except rather than "procure" I'd want to see the design and purpose build of a system in house that meets these authorities common statutory duties, and then mandate that councils have to adopt it. Whilst that mandate treads on a few "local democracy" toes, councils have proven time and again that they can't manage these projects.
My thoughts as well. Best ERP implementations I’ve worked on were small, focussed, affairs with a drive to go live with as little customizations as possible. Same product with less user focus? typically drawn out messes.
What these, fascinating, ERP articles should delve more into is the exact composition of the cost components.
If the overrun is due to vendor invoices - upfront and maintenance fees - then maybe change vendors & products (as some suggest here).
If on the contrary the bulk of invoices are from the implementation team (as seems likely here) then look at product fit, implementation team competence and user specification processes.
If the cost accrues because product demonstrably needs workarounds and customization to what’s in its core functionality, ban its use for govt until it matures.
That all said £2.6M for 19k employees seems like a massive underestimation to start with, so bad planning is also part of this.
Is your place a council, in which case, business as usual presumably. If it's an independent business that's more interesting as (self employed so making big assumptions here) they'd have a bit more of an interest in not going "Birmingham" on a pivotal part of their IT, & won't have the option of selling off a load of stuff to finance the shortfall?
Colour me shocked that this has happened. Not.
The issue is again a system that isn't fit for purpose without extensive customisation never mind the additional training for operators.
It seems to me that it's *almost* a deliberate ploy by the software vendors (Oracle, SAP, Sage / whoever) that they make most of their money by customising systems rather than just delivering something that works. Never mind that SAP users are struggling to justify the migration to S/4HANA due to the massive additional costs of re-customising their customisations of the on-prem system.
I would completely agree with previous posters that what is needed is a single system that is applicable to all councils as they all do pretty much exactly the same things in exactly the same way and have the same fiscal and reporting requirements.
I have some skin in this game as our local council has just been amalgamated with another which uses a different ERP system. The shit is heading towards the fan as I write.
they make most of their money by customising systems rather than just delivering something that works.
was chatting with an IT integrator once and they admitted that they bid "at cost" at the initial tender stage and then screw the customer for contract variations as they know customers will always need contract variations.
You are not supposed to customise COTS based systems; change your business processes first, especially if your ERP of choice is Cloud based (Fusion/SAP Cloud) it will be easier (migrations/upgrades) and cost (support, run) far less over a 3/5 Year TCO
That said, if AI (the coding variety) even remotely actually worked as advertised the cost of these projects (implementation, migration) should be coming down; its not AFAIK, so...
That's tail wagging the dog. Your business processes got you to the scale where you can afford this stuff. Changing them now just because you bought some software is its own risk
You're meant to work with some sort of functional expert to get the software aligned with your process without having to customise the thing into a unique artifact that no one can then get you out of
Agreed theres a lot of things you can functionally adapt (configure/setup) in common ERP platforms, to your business processes; they each implement a business process i.e. "procure to pay" using (their own view on) best practice - but if your functional consultants can't get you most of the way there without breaking out an IDE/compiler (what I mean by customisation - Oracle Visual Builder or SAP BTP) your "pre-sales" functional consultants/business have chosen the wrong software - admittedly you may not have a choice - see the relevant Oracle/SAP support forums on recommendations; the base SaaS versions have limited options for customisation anyway - though common in the past, current general guidance is to adapt your business processes if you can
"You are not supposed to customise COTS based systems; change your business processes first,"
I worked for three Oracle partners before retiring. I have tried to convince clients to change their business processes many, many times. Have never succeeded. Usually management and the project team see the benefit but the actual users have objected to changing the way they work and management decide it's not worth upsetting the users so request customisations, that were never part of the agreed requirements. As someone at te coal face, so not getting any commission from the extra revenue, it can be a wee bit frustrating as I had to gather the user's requirements, write the spec, get the user and project team to agree, cost it up and update the project plan, often meaning a delayed (and unpopular) go-live date.
Mind you I was once interviewed for a role as an implementation consultant (not for Oracle) where I was told that I could earn commission. When I asked how, I was told that the sales team often under-spec'd the system, I would then identify these gaps, get the change request from the client and get commission on the additional revenue. I didn't get the chance to tell the company to go forth and multiply as at 37 years old, I was too long in the tooth to work with all the youngsters.
Not sure of the truth in this rumour but when a very large government department decided to upgrade their ERP system all of the quotes were many times their budget. So one of what was then the "Big 6" took the requirements and drastically reduced the scope of the project, with the agreement of said department. Thus the contract was signed at the udgeted cost and the huge additinal costes were all via change requests. A stitch-up at the highest t levels of both client and integrator.
At these cost levels, besides using a public sector implementation team rotating through councils I would also strongly consider putting in place go-live/no customization bonuses for the users.
It’s, to an extent, a stress to change some habits. Although I question how much the worker bees care as long as it works and is fit for purpose. As opposed to users - typically mid/low management, not say payroll clerks - sitting in those meetings and claiming they need something special.
Anyway, put aside a training budget and a bonus for a job well done and a Birmingham narrowly averted.
One gig I started out doing (got canned for refusing to falsify acceptance tests) had the consultants trying to rewrite a legacy AS400 paysheet in the new ERP to be a pixel-perfect match. For 200 unionized employees. Much easier, IMHO, to come clean: “same numbers, different looks, here’s $500 bonus at go live!”.
Put peer pressure in place to stop these pointless dick-waving examples of “I am an important thinker in payroll business process theory, have never run a payroll and our council can’t possibly use the same scheme as the others”.
Reality is the business processes for non-core functions like HR and payroll are probably driven by a) unions and b) whatever the previous IT system mandated. There are probably some things the worker bees really don’t like and if you can improve their life there, they will be happy too. If not: bonuses. Just make sure your replacement actually works for them, tho!
as non-core disappoints me.
(Providing work and currency to people who on average working the area is very core for a local administration, apart from the more obvious aspect.)
I could even argue that Personnel is necessary to look after the personnel who do everything else.
A council is in the biz of providing services to residents. That’s its core biz and they may have special ways to do it.
A payroll is a means to an end: paying employees who work for that council. It’s very important that they get paid correctly, but how it happens is not a unique business feature of that council.
No one’s claiming paying employees is not important. Phoenix federal ERP in Canada $1.2B fiasco had people getting mispaid for months/years. That’s unacceptable.
But a working payroll, as long as it works, is not core to govt business. It is to say a payroll as a service provider.
Not going broke and screwing residents - cough, Birmingham, cough, IS a core function of those councils as well.
Having seen the damn straitjacket that the ERP enforced at a former employer, I’m not sure if any “off the shelf” product are fit for purpose.
They are so hideously complex that no one can understand them. Not what business functions they should support, nor how it works underneath. The business processes require so much customisation to fit the business, it’s better to fit the business to the function (tail wagging dog). Users (non IT, non tech savvy-financial ones) don’t know WHAT the buttons mean. Literally, “for my job I do to this menu and enter this, because the other system told me to”. They really don’t know WHY they are doing it, or even what information they are really giving it. Maybe the person employed in the role when the system went in knew what the data means. But, you know, human nature means the job transfer wasn’t complete, or documented, and no one read the documentation and its now lost. In the end, garbage data mangled by a system no one understands.
Suddenly these mainframes (and IBM is the only game in town, now) and ancient COBOL codebases don’t seem such a bad idea. Business process that haven’t changed in decades.
For a very short time I was entrusted with authorising purchase orders on our SAP system. The slide deck (!) I got to guide me bore no relation to what a modern UI user would expect, eg pick My POs, select one for $project, review and click confirm. Just like Amazon does with business payments authorisation. No wonder the replacement ERP needs customisation!
why the people who negotiated all of this had to be paid so much ?
Seems to me you could have achieved the same outcome at a fraction of their wages if you had asked a local school to do it as GCSE coursework.
Oh well, at least they can cut the services to the vulnerable and needy. Which was the plan all along.
One does wonder how so many assets become serendipitously available. As councils inform us that they have more and more statutory requirements it seems strange that they can meet them with fewer and fewer assets. Could it be that the council has been previously profligate in acquiring "things" which were not, in reallity, required?
Councils often own things like leisure centres which can be an important source of revenue. When cash is then needed immediately they're forced to sell off these assets, removing the future income. For example, after the 2012 equal pay ruling against Birmingham City Council, they sold the NEC group.
No. Blame central government. Central Government has been reducing the funding they give to councils, telling the councils not to raise council tax and telling the councils to do more. The councils are stuck between a rock and a hard place.
The councils are having to do whatever they can do to get cash today, knowing full well that down the line they're going to be even more screwed.
Not to mention that local government are often the victim of Westminster Political Theatre, whereby central government announces £xxxm to fix potholes/ support social care/ address SEND costs etc etc, but the money has to be spent by 1 April, and is usually announced with insufficient time to bid for, plan the works and enact the spending of the claimed windfall. Similar goes on with the NHS when it comes to promises of more money from government.
Jeebus, hasn't any council a list of priority works (aka "wishlist"). When that shit happens, retitle the last wishlist and resubmit.
Hell, when SWMBO asks "What are you thinking" ... the answer is always "Who would win a battle between zombie orangutangs and vampire wolfhounds on the moon". You just have to be ready.
Software that is a framework *always* requires extensive software development to make it work. Whether that is in the financial sector, SAP, Oracle etc. Some also require you to modify your organisation to fit the product.. 8 digit "cost codes" anyone.
When using a framework product, you also have the "benefit" of paying top whack for "integration consultants" (university grads getting tiny percentage of the fee)
You could put together a lot of this stuff yourself given a few small teams.. but everybody *keeps* falling for the "oh, doing it yourself would cost £15m, FIVE TIMES our quote", but then its £50m, and nobody ever reviews it. Sunk cost. Time after time.
Anon, because.
Worked in all elements - see 10 box model - of the lifecycle of Oracle ERP inc Fusion, PeopleSoft, but also for SAP S/4HANA , Infor etc for all of my career inc in Public Sector
I also worked on ODOO (OpenERP), OpenEHR and related and whilst its not perfect, I'm pretty sure if you put all the money being spent on Oracle Fusion and SAP Cloud ERP (forced) migrations into say a UK based "Open Source Public Sector Hubs" - run it where you want/with who you want - the cost to the Taxpayer would be an order of magnitude less and you may even get a "Public Sector Open Source Ecosystem" supporting a wider UK/EU Digital Sovereignty ...
They should start looking into contractual mechanisms where 3x implementation cost overruns should result in consequences:
- void vendor support fees until it goes back under 3x
- fire top decision makers for project at council: project director, project manager. audit specs for needless bikeshedding and stakeholdering and fire those who wrote them. and those who undersized the training budget.
- blocklist implementation consulting firm from govt gigs.
This is £2k implementation per employee. What on Earth could justify that, unless the product was a bad fit to start with?
p.s. Yeah, all about as likely as The Donald getting a clue, becoming an empathetic leader and not asking those he insulted to get his chestnuts out the fire. Or streaming services becoming good value. Still, I dream, therefore I am.
walesonline.co.uk: .. “since 2016, Infosys has earned £15million working for the Care Quality Commission, the regulator of UK care homes. In 2019, it won a £5m contract with the Medicines and Healthcare Products Regulatory Agency, and this year secured a £25m IT contract from the Tory-run East Sussex County Council.”
“Infosys is among companies that shared £100m in public sector contracts between 2015 and 2021. It was also one of nine partners in a £10m contract with Tory-run Westminster City Council last year.”
The council tax payers should be asking for thie CTO/CEO etc heads to be put into a pillory and have rotten tomatoes thrown at them for a few days/weeks, Take their bonuses for the last few years to recoup their losses.
DXC = Delivering Xtra Cash
As for The Oracle - there's a reason she was called a prophet :)
I can't help but feel that the vast majority of the problems are caused by the tenders and contracts being written by people who lack the experience and expertise to avoid the common pitfalls.
Now, I'm not saying that this is necessarily true across the whole of the public sector, but I briefly worked for a public sector organisation myself a few years ago. Within a few weeks, I realised with a creeping level of horror that every single one of my colleagues could be bucketed in one of two ways:
1. They were competent, and had less than 2 years service in the organisation.
2. They had more than 2 years service, and were so staggeringly incompetent that you wouldn't trust them to put the bins out on the correct day.
This rule was 100% true of all of my colleagues regardless of their seniority level. Basically, anyone who was employable elsewhere bailed at the first opportunity.
These council ERP failures smell to me as though it's the second group who were instrumental.
"I can't help but feel that the vast majority of the problems are caused by the tenders and contracts being written by people who lack the experience and expertise to avoid the common pitfalls."
This - worked on both sides of the "contract" fence but its not just Public Sector
The responses to the tenders and the evaluation are also often poor because there’s no actual competition or the required solution is so complex there is only one answer... no bid...