The only thing that's remotely unusual about this is the scale. It's a large scale version of Libra, CRB, CSA, Passport Office and all the other IT disasters, and it was an obvious flop from day one.

Where do they go wrong ? These errors are almost mandatory in every system at every level of the public sector.

(1) They don't ever talk to people who are going to use the system. If you are (for example) going to produce a Police IT system you should talk to the low level Plod (ideally anonymously) about what they actually *do* and how the systems actually work, ideally spend time on shifts and "shadow" the various jobs.

The implementers talk to the people who are in charge of the workers, so they hear what they managers *think* is happening, not what actually does happen. The two are rarely the same.

(2) They don't ever try small scale first. The reason for this is money. You get more £££ for a massive roll out.

What you should do is implement it a small step at a time (always bearing in mind that it will be part of a bigger system), trialling it in a few places, learn from the inevitable mistakes. Then you scale it up ; from a hospital, to a local network of hospitals, then to a county, then nationally.

These implementations are always mega-new-system crash-bang launch and it never ever works. (Choose and Book's figures are fradulent achieved through bribes and fiddles).

(3) They build way over the top systems which aren't necessary, again it brings in more money for the IT firms. An example of this is the "Spine", a huge network which is used to transmit the hospital data. If this goes down, the whole thing fails.

No-one else designs systems like this. The usual way is to have lots of interconnected systems (which could even pass information over the internet, encrypted) and achieve their power by those links. Look, for example, at how the p2p Piracy networks efficiently move masses of data about. Google is a classic example of this ; it's not one massive machine with every website on it, it's lots of little ones cleverly connected together.

The perceived advantages - that if you are on holiday in Scotland they can access your records in Cornwall - are not worth the money put into it. This *hardly ever happens*, so using a slower internet based link on these odd occasions is fine.

(4) Almost all Public Sector people are lazy morons. But they think they are really hardworking, clever and really high powered business types (how else do they get to the top, they think ? (Buggins turn, brown-nosing, etc.).

As a result, people who do have a clue about running businesses buy them a few meals, tell them a few stories, and take them for the biggest ride ever.

Unlike the private sector, there is no punishment for failure, as long as it's kept quiet. In extremis there is reward for failure in the form of a fat pay off to keep schtum.

Private companies exploit this outrageously, as rather than have their incompetence publicly exposed, the public sector desperately throws money at them in a desperate attempt at a cover up and in the vain hope that something, anything, can be made to work.

No amount of public money is too large to cover up Public Sector incompetence. Hence the massively accelerating budgets (also down to obvious loopholes in contracts)

(5) They always build from scratch. There is more money in this than buying and adapting systems which already exist. Systems already existed here and in other countries which could be improved and adapted, this is what CfH was originally going to do, estimated cost a couple of billion.

Instead it's ALL NEW with the concurrent massive expenditure and risk.

All these projects go through the same four phases.

The Pride phase
===============
At the start, when they have massive claims and expectations, "this system will do wonderful things". They spend money on all kinds of stuff, which is construed as a worthwhile activity in itself. Sometimes, early progress can actually be made (or in the case of things like C&B, faked to look like they are working). This usually never lasts because no-one thinks of the scale up.

The Panic phase
===============
This occurs when the public sector bods realise it isn't working. The normal sign of this is escalating budgets as money is thrown at people in a desperate attempt to get it back on track. It fundamentally fails because it is on the *wrong* track to start with. This is internal, you can't tell this is happening, other than the budget escalating as suppliers run rings round them.

The Coverup phase
=================
CfH is just coming out of this. This is when it all goes very quiet, reports are delayed and rewritten, and nothing much seems to be happening. Getting information about the project is very difficult, and what is provided is very positive, but short on specifics "Our systems are wonderful", but no information as to when it will actually work. Things like Computer Weekly being banned because it writes critical articles also occur.

The Blame phase
===============
This one is optional ; the public sector takes responsibility for nothing. There's inevitably a lot of spin to blame everyone but the people responsible. In CfH or NPfIT or whatever it's called this week, this is about to start.

Posted by: Paul at April 17, 2007 11:17 AM

To be fair, IT projects on this scale have a very poor track record in the private sector too. Of course, that has made the private sector very wary of undertaking such projects, while the government remains enthusiastic.

Posted by: Andrew McG at April 18, 2007 05:53 AM