Another one bites the dust

The Department for Works and Pensions has decided to ditch a major IT project. The streamlined payments system was supposed to deliver savings of £60m by 2008. The project cost to date has been £141m. Anyone still believe a National Database Register and ID scheme is deliverable and within budget?

  • Occasional Commentator

    Writing good, reliable, cheap software is quite easy. The problem is that nobody in government knows how to buy it, and therefore nobody has the incentive to deliver it. The private sector is little better, but it is at least a bit better at pulling the plug earlier.

    Do the public not realise that when a government department announces a tender for IT work, the IT sector rubs its hands with glee? It’s the easiest money for old rope and these coyboys are sucking up your tax as quick as they can before the public wise up. The systems are pretty straightforward (compared to the money paid out) and would be much the same throughout the EU, if not the world.

    The correct approach is for the industry to develop software in advance for this type of thing. Then they should try to sell this premade product to governments around the world. The governments could test it, or investigate other governments’ use of it, very cheaply and just not buy it if it’s not good enough.

    The only way this’ll happen is if the government announcements it’s only going to buy premade software in future. They might be ignored at first, with all the companies just concentrating on other idiotic governments, but if all the EU states got together and said to the industry “enough is enough, you’re only getting our money after you’ve made the product and we’ve decided we still like it” then it’d work.

  • kensei

    How many times must this happen before government learns to write contracts that hammer companies that do this?

  • Occasional Commentator

    kensei,
    I remember everyone saying how “cute” and “smart” the FA were for having a contract for Wembley were the FA paid a fixed price for the entire project no matter how much the builders had to spend on it.

    It wasn’t really that “cute”, it’s just the obvious way to do it. Same with the software.

    These companies should pay a windfall tax (I normally don’t like retrospective things like this, but this behaviour of the software companies is borderline criminal in my opinion)

  • Michael Robinson

    Wembley is not really the best example of the obvious way to do things! The problem with fixed cost contracts for major projects is that it is impossible to forsee all the problems that may occur. When things go wrong and costs rise, the lawyers get involved rather than people trying to solve the problems.

    [url=http://www.airport-technology.com/projects/heathrow5/]Heathrow Terminal 5[/url] is a much bigger project than Wembley and is on time and on budget. It isn’t been done using fixed price contracts as BAA (despite their incompetence elsewhere realised that the “obvious way” doesn’t necessarily work work for major projects and it is more impoirtant for the contractors to solve issues than covering their ar$es. I’ll see if I can dig up more on the approach used.

  • Michael Robinson

    .. however I agree about the staggering waste and incompetence of government IT projects!

  • nmc

    I work for an IT firm in Belfast, and think that the government could, through the creation of their own IT department, potentially make money while creating their software. Instead of tendering large IT projects, hire ten IT graduates and pay them 30k a year. Provide them with the resources required, good experienced leadership, and then set about creating the required software. The software could then be sold to other governments or private firms, and if it were the desired outcome, the department could make a profit by taking on other projects for private companies.

  • Occasional Commentator

    Michael Robinson,
    So what if problems occur? That’s what the (software or stadium) developer is meant to look out for and cover themselves. It’s called doing their job – the management of that company find the money to fix their problems. Imagine ordering a car and being told you’ll have to pay more because the first engine they put in didn’t work? That’s not the way business is supposed to work. The car company should factor in the cost of the odd failure of a part and charge the customer the agreed price and no more.

    If BAA are good at pretty much running their own construction firm, then bully for them. They probably have more experience than the FA, what with constantly extending airports and terminals at various airports. The FA probably don’t have such experience (as they would have no involvement in the average club’s stadium).

  • Gooseberry

    “Writing good, reliable, cheap software is quite easy.”
    OC, please explain how this is so.

    I agree with rest of what you say about company’s rubbing their hands with glee when a government body is looking for software. ‘Of the shelf’ approach is not always a workable option but a basic framework should be there with ‘bolt on’ bespoke modules designed for the specific customer.

    Good, reliable software is never easy to develop and is never ever cheap.

  • Michael Robinson

    errr, no…a £4.2B project with 60 major contractors isn’t quite the same as buying a car.

    If something goes wrong, it isn’t a matter of leaving it into the garage to get it fixed.

    Saying “its your problem” to a contractor or supplier is missing the point – if something goes wrong on a major project then it is the client’s problem as well.

    It was precisely because of major issue in a previous project (the collapse of the Heathrow Express tunnel) and the blame culture arising from this, that BAA decided they weren’t going to do fixed price contracts.

    The results – compared to the Wembley fiasco – speak for themselves.

  • Occasional Commentator

    Michael,
    I don’t know what you mean when you say “a £4.2B project”? Was it Wembley or Terminal 5 or some software?

    Anyway, a project is the same project regardless of the price tag assigned to it by one person or group. If it could be done with 42 million, then it’s not really a £4.2B project.

    It’s easier with building because most buildings are probably just variations on existing buildings, and the building trade is millenia old. Software development is much newer and changes more quickly (although the principles of sane software management are the same as ever).

  • Gooseberry

    Move over Bill Gates here comes ‘Occasional Commentator’

    So OC, how many Software Companies do you own?

  • Michael Robinson

    LHR T5 is £4.2B, Wembley a mere £750M (so far)

    I think you are being a bit dismissive about modern construction challenges – T5 will be one of (if not the largest?) single-span buildings in the world to it is hardly a copy of something else.

    Sane software management is not that different from any other sort of sane project management although I think software is a bit more guilty than most of believing the latest language, tool or methodology will be the silver bullet to magically solve all problems.

    In my personal experience, managing techy coders is a hell of a lot easier than managing bolshie builders!

  • Occasional Commentator

    There’s no need to start the “who are you?” routine. I may be a nobody, I may be somebody. There are plenty of people who own companies who are idiots. And in the current climate of idiot government purchasers, a successful company doesn’t need to know much about software development to make the money.

    Seriously though, some of the best software costs literally nothing to buy and you can be sure the cost of the time to develop it is much lower than the cost of much equivalent commercial software.

    I don’t think anybody familiar with any part of the software industry would deny that the difference in productivity between motivated and competent programmers and managers can be orders of magnitude (i.e. 10 or 100 times better). Some of the most obvious truths about software development have been known for over 30 years, but you can be sure the average software developer isn’t aware of them because he/she doesn’t need to know them.

  • Occasional Commentator

    Michael:

    I think software is a bit more guilty than most of believing the latest language, tool or methodology will be the silver bullet to magically solve all problems.

    Too right. Rewriting bad software in a new language is just like translating a bad novel into another language – it’s still a bad novel. That’s not to say there aren’t vast differences in between computer languages (I’m currently hooked on Haskell), but good software design, just like bad design, shines through no matter what language is in use.

    Good tools make good programmers better, they don’t really improve the bad programmers. When dealing with mediocre programmers it’s better to use good processes and management than it is to change tools.