Public sector IT: Our own pig in lipstick?

conservative with a small c

What gets me time and again, in both public and private sectors, is IT managers who think that letting their environment become obsolete and out of support is the "risk averse" or "conservative with a small c" option. This is how you end up with hellish expensive support arrangements, as you plead with suppliers to allow you to add capacity for software which is off the price list.

If you are a bank running a central ledger on a mainframe, then running a 20 year old solution makes sense. If, however, you web enabled your services in 2003, then you should have started replacing them in 2008. Planning to "make to and mend" will cost you more.

Yes, yes, do-over always looks cheaper

Because it's usually tendered by people with no actual experience of the system that they're proposing to replace.

All problems look trivial to a salesweasel or bid manager, because it won't be *them* that's in the server room at 2am with an irate customer who didn't mention the vital importance of Despodulating the Flamcrack Rectifier because, well, wasn't it *obvious*? The last supplier knew *all* about it.

Couldn't agree more with parts of this.

Without naming names, the techies are just outgunned by the big suppliers.

On one project alone this year, I could conservatively save one subsection of one government department, 40 million a year, simply by bringing it in house.

The whole, "Government's job isn't to do the work" is rubbish. The same staff doing the same jobs, would do them better if there wasn't an agenda of "Forget delivering our commitments until you've got more commitments" bordering on what I consider the criminal, by senior consultants in outsourcers.

Progress meetings aren't even remotely related to meeting about progress, and in fact seem to me to be little more of catch ups where slick salesmen (who are technically illiterate but describe themselves as principal consultant or something equally facile,) turn up, flash beguiling terms and name drop, but don't let the government staff out the door until the question "how much have you got to spend?" has been answered.

At this point, they invent a new job role, or just anything to suck the money.

Need more staff on the job? Let's invent a new job called "Performance tester." Isn't poor performance self evident? Let's invent a new role, business analyst. (don't the staff know what their business is about? Why not ask them instead?)

Hey presto! 1200 quid a day please!

And it's not just government. On one job, I turned up and designed a Global reward system, (from scratch - having first had to learn the business,) in 12 days. the implementation of which, would have cost less than the Oracle licences. The big consultancy which had dozens of people on site, and were flying people who I could only think of as seat covering non-entities (charged out at over a grand a day,) down from places as far as Cumbria for meetings, had taken two years to fail to do it.

What they did make sure was done though, was that every move and inducement was signed off by an indigenous manager. The company thought nothing of it at the time of sign off because they thought the consultancy knew what they were doing, but they surely came aware of it, when they tried to sue the consultancy, and found the consultancy did know what they were doing.

That said, the government will never change the rules to allow genuine competition between the big consultancies et al, and the SMEs though, for two reasons.

1. Big companies hire care in the communty types, under the guise of diversity officer, cleaner, etc. Unemployed people vote the government out.

2. Big companies also hire ministers as non-exec directors, on six figure salaries.

I don't suppose the odd political donation hurts either, though can't say whether this ever happens. I really must look this up on Google.

The article writer, whilst almost perfectly correct, is wasting his breath. He may as well write an article about the injustice of agencies refusing to forward uncleared people for jobs requiring DV or SC. Both articles would be equally a waste of time.

Both wrong, both right. Both off the mark.

All over Europe, and that for once includes Blighty, we see the same old, same old: Looks, we gots intarwebz, sayeth the government in its wisdom, let's move over! This then fails spectacularly and we start pointing fingers.

How I'd like to do it: Choose. Not just what to intarweb-ise, but what to offer. What do we have a government for? What do we really want it to do?

Then we choose how to offer those services, how to support them. Example: So, you want to offer health care. To do that efficiently, you need to keep track of previous health, problems and all. Turns out you need some way of identifying the same person and attach a health record. Then there's privacy concerns. But you could solve both with clever use of technology. Then there's the fact that whereas making the things electronic would be very useful, the records need to be accessible --and updateable-- when there's no 'leccy, too. So how do we do that?

Start small, but in a way that means you can scale up. Identify those things that are useful elsewhere, too. Identify those things that you don't really need, then exclude them from the model wherever possible. Identify what needs to talk to what, and why, and when. If you can re-arrange things so that complexity of the things or the way they interact reduces, do so. And so on.

Yes, that means reinventing most of the wisdom that drove governments over the last couple of centuries, say since Napoleon, because he pushed through reforms that are still in use. We have to, because we're moving from paper to something that is entirely unlike paper, behaves differently, and has to be dealt with differently. No longer can we rely on large bureaucracies to remain inert and patient. We get instant fame once our details are leaked, so the systems have to stand up to that, not through patching, through inherent solidity.

But you don't want to rip out the old system and enjoy the fallout until you can get the new system working. You have to work your new system so that it can adapt to the old system, take over seamlessly, in fits and parts if you have to, then once we're "over", it can morph to its real form. And so on.

And I could come up with the first plans for well under a million. What the final cost would be? Less than the running costs of the paper mill we have now, with more function. What it'd look like? I can only see the tip of the ice berg yet. In any case, it won't be free, but it won't come with the absurd costs that the current pork barrelling fest is incurring.

Do I have the final answer? Fsck no. Do I see where the road needs to go? Sure. Does the government? They can't, they're not flexible enough. Part of the medieval set-in-stone mentality that drives bureaucrazies. Do the politicians? If they've thought about it at all, they'll believe they don't need to; the next election is in a couple years, not decades.

Public Sector Capital Cuts

1. Public Sector IT development is often treated as capital expenditure, precisely that category of expenditure highlighted by our new coalition government for cuts. Does anyone seriously believe that these cuts are about the overpaid, incompetent bureaucrats in the civil service and local government? No, it is about ignorant and overpaid consultants in the IT industry, the private sector, who will also be losing their jobs and consequently reducing their demand for everyone else's products and services. These cuts are not about waste and inefficiency at all - they are about contracting the economy and mass unemployment. That is the agenda.

2. Small firms struggle to get in to the racket as a consequence of the doctrine that they need to be big enough to sue if it all goes pear shaped. Can we have this doctrine re-examined in the light of the history. Just how successful was the capacity to sue ITsoft when it failed to deliver, or Accenture when it walked out of the health contract leaving the NHS to unpick the remains of the IT soft deal. Does anyone recall Accenture handing over shed loads of wonga? It seems to me that when a contract goes wrong, getting compensation out of a big powerful contractor is impossible?

Outsider's thoughts

First of all, let me state that I am a complete outsider to these big projects. However, perhaps I can bring a different pov to the discussion because of that.

My issue is that if I were in a position to be running a project, to see the phrase "radical innovation" applied to a key part of that project would have me running for the hills quicker than a yeti in a heatwave. First of all, it sounds like marketing BS, to which I am extremely resistant in all areas, and secondly, even if I got past that, what I am hearing is "I think you know so little about this subject that you will pay me to experiment with my half-arsed ideas". However, if the existing supplier came to me and said "Oh, well, we can keep this thing going for another few years - why change?", I would think that they are merely trying to avoid a tendering process that they might lose.

To put it in a nutshell, neither "radical innovation" nor "conservatism with a small c" fills me with any confidence at all. I definitely think that the big companies have messed up government IT projects so much that they should be excluded from any further tendering (and spare me the "but the client keeps changing the brief" moans - it doesn't hold water), but there is a significant risk of the same amount of fail by going with SMEs, especially if they are pushing "radical innovation".

Somewhere, the tendering process needs revising - but how? Who can you trust to give a sane, honest review when so much money and ego are in the mix? It is all but impossible to get a truly independent review from people with enough experience that they are outside the tendering process.

More soapbox time from a different view

Architects - there's a phrase that keeps coming up. Gov IT has loads of them, but they're seperated from the actual delivery of projects. Nowadays all the senior management are interested in any branch of gov is writing 'Strategy Documents' completely divorced from the day to day business of the Departments. Frankly this probably not a bad thing as their strategies are mostly unimplementable and often extreme and stupid.

Public service culture and advancement within it is generally based on two things:

1. Brown nosing

2. Conformity

It's not a place for innovation and its resistance to change is immense. Big bang solutions will always be subverted, mostly subconsciously, as the basic culture is so entrenched.

Which of course makes it pretty closely related to big business as well ;-)

The reason other than arse covering for consultants get so much say is that even the people within the public service have a very low opinion of their colleagues and the people in nicer suits give them shiny new buzz words to play with.

Two things

a) There is a need to radically reduce the number of systems, most of these are supporting concepts that can be scrapped - for example the BBC licence fee and the database backing that up (fund from general taxation as most of us have a TV, and all are assumed to have), the road fund licence and the database there (just raise the fuel tax and therefore tax those who use most fuel and cause most polution), local taxation (fund from central government per head of population and use democracy to decide how it is spent).... There are more depending on how radical you feel but those small things woudl be a great start.

b) Buy British, its British tax payers money and if it is to be used to support profits and jobs they should at least be British profits and jobs. This applies not only to IT systems but all government expenditure. Sort out the current mess which is government procurement, it doesn't need to be complex or difficult, and certainly shoudln't be expensive or impossible to tender. (I've tried and believe me it really is impossible).