UK.gov's cost-cutting 'shared services' went £500m OVER BUDGET

@Schultz

Spot on. Had they employed anybody with actual experience of setting up and running high performance shared services, then they'd have had a chance. As it was the majority of the staff were career civil servants who didn't have a clue about shared services design, strategy or performance management. AC because I have dealings with shared services management at both Research Councils, DWP and other - and they're nice, well meaning people. They just don't have the experience, they don't have a supportive management structure (ie the Civil Service is a management disaster), and employing highly paid consultants merely buys you some very expensive Powerpoint packs, filled with meaningless generalisations.

Re: I think I'm missing something

"So, in what way were they shared?"

Because before it was even more fragmented.

So, taking the Research Councils, there's five or seven different of them, and historically each did its own back office, wholly independent of the others. By bringing them together you should have standardised processes, economies of scale, and more professional management, and the management of each RC should be freed up from managing business but not mission critical acitvities.

There are problems that the article rightly picks up, but the maths is pure Graudian/MP standard rubbish. So the costs so far exceed the planned saving by the end of last year, well? Most business projects take years to pay back, and we accept that as normal. If these projects take five years to pay back, and it cost £2bn, then you'd be saving £400m each year, and as of year 6 that would be pure benefit. Maybe it could have been done in half the time and at half the cost, but at least we will see a benefit eventually? Ten years down the road the set up costs for the shared service will be irrelevant.

You might be thinking "but there should just be one accounts payable for all of government, etc", but to go to that in a single step from the preceding chaos would be madness doomed to failure, and there's functional differences that will probably always be different - eg you would always have different levels of clearance for staff processing defence invoices compared to run of the mill stuff, research funding payments would be a completely different kettle of fish to paying for office stationary. And unless you can imagine all of government running a single ERP, you'll always have multiple systems and different workflow arrangements.

I'm not defending all of it, but it is a start, and it it is saving money (although not so much when they involve the thieves and incompetents of the BPO sector).