Blighty slaps £100m spending cap on govt IT projects

The coalition's election pledge to cap IT contracts at £100m will become official policy from next month. A procurement policy notice released by the Cabinet Office sets 1 April as the deadline when the cap will come into force for central government departments, their agencies and non-departmental public bodies. The £100m …

[Mike Tree]

So, no cap then!

[Anonymous Coward]

the HEADLINE will be...

'Owing to costs all IT projects will be outsourced'

[jake]

How about ...

... spending a tenth of £100m on figuring out what IT actually is, instead? Surely that would be a better way to go about things?

[jason 7]

Re: How about ...

Indeed, also giving all MPs and ministers a 1 day course on IT basics could avoid a lot of costs.

If they had half a clue as to what's possible and what's not possible plus some idea of the scope involved could save a lot of money.

The viability of a potential IT contract/project cannot be gauged by the quality of restaurants/gifts the potential contractor takes/gives you to.

[Anonymous Coward]

Re: How about ...

Trust me I've met some and a days training is optimistic, several years would be necessary and even then....

They wouldn't listen anyway.

[jumpyjoe]

Re;How about

Nay! They'd hire someone from Microsoft to give the course and from then on it'd be business as usual.

[James Cooke]

So that's £100M for the big and triple that after the cost overruns and delays to produce a system that fits half the scope?

[PassiveSmoking]

Most IT projects that overrun on time or money do so because they fall at the first hurdle - requirements capture. Either the dev team don't know how to ask the right questions to find out what the client wants, or the client just gives rubbish answers to the questioning. Then then the product is a lot later on in development someone points out that what's being built isn't what they wanted (If you're really unlucky, this will happen on delivery of the finished system), requiring lots of work to be done over. This is a problem that affects public and private IT projects alike.

If you don't solve this problem then all the caps in the world won't help. You either abandon half-finished projects because of the cap, or you break the cap to get a project delivered (often due to the sunk cost fallacy).

Whether it's the dev team or the clients (or both) who are crap at communicating is pretty much irrelevant. Until the communication problem is fixed, this will remain an issue.

[Stuart Castle]

V. True

Thankfully, I've never been involved in a major government IT project. However, some friends have, and they say basically the same as you..

The main problem they encounter is feature creep.. Essentially the consultant managing the project fails to lock down the feature set properly (because, mostly, it's profitable to do this), and the clients suddenly come up with all sorts of features they want. This adds time and cost to the development and testing phases, and may also require extra hardware or software.

As I say, it can add extra profit for the consultants and companies involved (hence they seem to make little effort to stop it).

In the case of one of my friends, one of the problems he found with dealing with any government IT project is that he tended to end up dealing with management (who may or may not have an accurate idea of how things work), and often had little or nothing to do with the people who would actually use the system day to day.

I think that, as you say, the main problem (on both sides) is communication.

[despairing citizen]

You missed a major cause of public sector project failure

The elected twit(s) "in charge".

Based on their expertise dervived from winning a popularity contest, in which the majority of the eligible voters couldn't be arsed to vote, they feel they are qualified to tell international academic experts, and people who have spent their entire proffesional lives working in the relevant field, that they do not know what they are talking about, if they dare to comment on a project/proposed project. (e.g. Charles Clarke and the LSE report on ID Cards)

They also tend to do the classic of pick a politically acceptable price and timeframe, with either no clear scope, or a politically mandated scope, and then go tell the PM to deliver, not having done any of the sensible pre-project research, feasibility and prep work.

This isn't just in IT either, look at any defence project.

[Tony S]

So..

"IT projects will be allowed to exceed £100m on the grounds of national security or to ensure the continuity of a critical government service."

On that basis, you can bet that every single project will be classified as being either related to National Security or being a critical government service. So the limit is meaningless; simples!

Meme merde, jour different!

[despairing citizen]

Re: So..

OR

you suddenly have 120 NHS IT projects of £100m each, instead of one £12bn project.

...

see this is the £100m project to deliver patient records in Redhill, that is the £100m project to deliver patient records in Reigate, completely different project scope, as defined spatial extent of each project's terms of reference.

[tkioz]

I'm trying to figure out what projects short of national defence infrastructure should even come close to 100 million... let alone exceed it...

[amanfromMars 1]

Seven sevens is unrealistic and nine nines just too greedy hence the agreeable compromise

Virtual Reality AI ProgramMING is one project, tkioz, and an absolute steal with guaranteed delivery at £88,888,888, with not a penny more needed to be wasted and not a penny less to be supplied and well spent and invested.

But it is invariably only bean-counters who need such figures, to justify their positions, as those who are able to work the magic in such projects that deliver what they say they can do, can also easily do IT for others too, and for free rather than for a fat fee. But hey, that's the crazy capitalist system for you, ....... think of a number, add arbitrary subjective profit, factor in valued added tax to sustain parasites and cover sundry other self-serving collateral expenses and pretty soon one is talking serious paper wealth for spending.

[localzuk]

£100m?

Still amazes me how an IT project for government can hit £100m, let alone the multi-billion fiascos that have occurred over the last decade or so.

What is costing so much? Why does a database system for the NHS end up costing £1bn+? Its a freaking database!

[Anonymous Coward]

Re: £100m?

As a civil service IT procurement officer, I can catagorically assure you that the cost of a project scales linearly with the number of users or records it has to store. It makes absolute sense that rolling out something used by 60 million people costs at least 6,000 times that used by just your average large company with 10,000 employees. Put that way, I think a mere billion pounds for a nationwide database is quite reasonable, considered!

And since it's 11 that makes it time for lunch at the Ivy... those nice people at Capita are buying today. :D

[N2]

Re: £100m?

It makes sense to me to rid Govt IT projects of the Prince2 strait jacket they seem to constantly jump into, that way it probably could be made 6,000 times cheaper.

[despairing citizen]

Re: £100m?

There is an assumption in your statement....

"Why does a database system for the NHS......"

it is never A (singular) system, it's usually 460 systems and working practices which they attempt to stuff into 1 (or 5) systems (e.g. magistrates court case management project)

At which point the complexity factor is now 460^460, not 1

Until national projects are "National", we will still have 460 systems, or 1 expensive defective bug riden system, that is either late and/or cancelled.

[jason 7]

Govt IT projects getting out of hand

The best example I remember was the refunding the folks who bought a National ID card.

Someone had the balls/bare faced cheek to say it would cost around £43000 a time to refund £32 to any person that wanted their money back.

[El_Fev]

well as a tax payer

at least they are trying to stop the constant trains wrecks that IT projects in government have turned into and cost us, unlike the last bunch of fuckwits which just spent everyones money like water!

[Crisp]

At least Government are heading in the right direction.

Now all we need to do now is to start scaling back that £100m figure.

[smudge]

At last the truth?

The article says:

"Projects will also be allowed to exceed £100m where "a strong case can be made that doing so increases the overall cost to the taxpayer, notably increases the risk of failure or increases the security threat to the public body or government as a whole", the notice says."

Read that very carefully. Is this, at long last, an admission of how things really are? Or have the little words "by not" been omitted?

[Measurer]

Re: At last the truth?

Damn, beat me to it! If it's anything like government engineering projects then the £100M will be spent writing reports on how a keyboard and mouse actually works.

[Scott 62]

why impose arbitrary spending limits at all?

i'm pretty sure your average goon on the street wouldn't have a problem with an IT project costing £500m (assuming its necessary, not random shit like giving every civil servant an alienware PC...), as long as it comes in at £500m when it's finished, not £2bn

[jason 7]

KInd of makes all those job adverts....

...asking for PRINCE2 to be a load of cobblers.

Anyone ever actually applied PRINCE2 to a Govt IT Project? If so why did it fail and end up costing £130 million rather than the £20 million originally budgeted?

Oh and why were you and your team not fired/hung drawn & quartered/tar & feathered/shot at dawn?

[Tim #3]

Re: KInd of makes all those job adverts....

Has anyone really applied all of Prince2 to anything? A former colleague of mine used to work at the OGC (who devised Prince2) - even they do not use it for their own internal projects.

[Anonymous Coward]

In standard Government units -

- that can be expressed as 0.5% of an aircraft carrier

Footnote for the younger generation: aircraft carrier is legacy terminology for a large flat-topped ship that could be used to carry, launch, and land aircraft, back in the days when we had aircraft.

[Anonymous Coward]

Re: In standard Government units -

Who needs aircraft.

The way our security is being eroded countries like Iran and the North Koreans are going to see us as role models and not enemies.

[Measurer]

Re: In standard Government units -

Let's get the aircraft carrier bill first (just the one), before we go and make silly comparisons!

[Will Godfrey]

Hark, Hark!

Is that the flapping wings of a flying pig I hear.

[amanfromMars 1]

Worthless, is the high price paid in your ignorance for their arrogance?

Come on, ladies and gentlemen, just think about what you know of promises given by politicians to know of the value of promises given by politicians.

IT isn't rocket science.

[jake]

@amfM (was: Worthless, is the high price paid in your ignorance for their arrogance?)

Rocket science has been depreciated because it never really was exactly what the proverbial man in the street thought was rocket science ...

More important is to ask yourself "How do I know a politician is lying?" The answer is simple: They exist. Never trust a politician. They don't give a shit about you.

THAT said, there occasionally exists an elected official who actually gives a rat's ass about their constituency. These folks are not politicians. They are humans with a clue. Said constituency should cherish these people. Sadly, they are rarely re-elected.

[Is it me?]

The £100m IT project

Sorry, but, it's very difficult to spend £100m on just IT, and that's where the problem is. This kind of project is not just about IT, its also about business transformation, and long term support.

So many government organizations believe that their shiny new IT system will cure all their problems, start with high hopes, and then decide we're not going to change the way we do business, so this great new system has to do what the existing one does, and then they start to change it, before it has been implemented, often contradicting established requirements.

£100m buys you a 10 year deal for a government organisation of about 6,000 under a your mess for less, type deal, of which about £3m will go to SAP, Oracle and Microsoft without them lifting a finger.

£100 will also buy you a shiny new national system to do something for a large department, like the NHS or HMRC, to get that on IT alone, it really has to be a Lot of IT.