Government needs to bring IT skills in-house A parliamentary committee heard yesterday that government relies too much on outsourcing and needs better in-house IT skills. Parliament's public administration committee was told that the failure of many UK government technology projects is partly due to being tied into large outsourcing deals with a small group of suppliers …
Really sounds simplistic to me. Its surely down to tight contractural control, if you ask for X and supplier delivers X but then you realise you needed X1, delaying delivery and increasing cost, that is more down to the requestor not understanding their business.
Perhaps look more at culling mid/top tier management who seem to exist just to create meetings and let the "coal face" make decisions at the lower level.
Easy win to blame a supplier, rather than the £200k a year cake eater
... isn't that the problem being looked at? There is certainly an argument that it is better to have in-house resources that are a predictable cost than outsourced ones that vary depending on all sorts of things. There is also the level of commitment to doing the job properly - there is always a question as to whether an outsider with no continuing responsibility for what has been done might be less interested in doing the job correctly. Not to mention the fact that any problems can be dealt with quicker in-house, rather than waiting for some contractor to arrive from, say, Portugal (like the automated passport gates at Heathrow).
The answer is somewhere in between - have a core staff of your own techies, and buy in expertise when it is necessary (and make part of the contract include training up the internal staff so that you don't need to buy in the same expertise twice).
Yes, outsourcing can work and does work, but you need IT people to manage it. I know certainly in local councils that a small team of experts (even just in Excel) can make a massive difference to the efficiently of every department.
Usually the big issue the council has to deal with, however it stupid massive outsourcing that restricts what they can do on there own network and what there 'allowed' to develop.
So I would suggest that most IT fails in these cases are poorly written contracts for outsourcing, combined with no in-house IT capable of service development.
I know one department under a BT contract that was quoted £10k to add a 100Gb disk to there department server. Now I understand that there are more costs involved than the simple purchase of the disk, however, the department is still getting royalty screwed. (before you ask, yes the server had the capacity for the disks, and it was not a high performance requirement).
I am quite fond of this report, because I think the 'mass' outsource of IT takes control away from the council, which makes it harder to develop services and improve service. Multiple small outsourcing contracts are significantly easier to deal with.
Anyway, these are all IMO, will this report bring changes?, well I am not optimistic.
"A parliamentary committee heard yesterday that government (...) needs better in-house IT skills."
"Better"? they could start by having "some" to move forward from their current status of "clueless".
After that, I agree that they should get better to move from "incompetent" to "somewhat educated" by 2050.
A large part of the problem is that many British managers - not just civil service mandarins - have an ignorantly snobbish attitude to programmers (and indeed anyone who does real work). Alan Turing himself came in for massive broadsides when he crossed the line by sitting down beside the "oily rags" and "mechanics" to do actual soldering work on one of the first computers. It was fine for a professor of mathematics to write books and papers about the theoretical basis of computing, and (at a pinch) even to write actual software. That was "white collar" work. But to soil his hands and mix with working-class technicians - that was unthinkable! As Andy Wells, founder of Systematica, used to lament 20 years ago, far too many people in Britain think an engineer is someone "who washes his hands before he uses the toilet" (to get the oil and muck off).
This artificial and completely illusory class barrier results in a completely hopeless approach to IT projects. Programmers are thought of as essentially like navvies or bricklayers - unskilled, poor people who are paid wages rather than salaries and who should be hired and fired from week to week. Most British managers would rather die than admit that any programmer should be paid even half their own salaries - no matter how vital his knowledge or how great his contribution.
If civil servants and executives came to understand that the best programmers are more akin to architects, engineers and lawyers than to bricklayers they might be more receptive to the idea of retaining them as in-house staff. And that would be the first step to getting some good, reliable, maintainable software systems.
That is valid for any technical personnel.
It is the standard question of a British manager when encountering a salary structure typical of a USA technical outfit: "Why did you allow your report to earn more than you? That is unacceptable"
In any case, there used to be a way around it - the skilled one were contractors so the UK managers could have a justification of paying them proper money. However Labour killed that via IR35. So the honorable committee can rant for all its worth, but there will be no people and no skills to take over from failed projects.
Labor (and its paying sponsors) made sure of it.
We can safely assume that British managers will not change and would still more likely commit harakiri than pay someone technical a proper salary. That leaves a restoration as pre-IR35 landscape as the means of getting where the committee wants us to be with viable SMEs and individuals set up as SMEs to pick up work.
It will take at least half decade if not a whole decade to get back there. Labour working on orders from its sponsors was quite thorough here.
It is true that contractors used to get away with not paying the NI contributions, and that was IMHO wrong (yes, I was a contractor, but paid my money as salary with PAYE and NI deducted, much to the disgust of my accountant), but IR35 was draconian, and effectively prevented a 'service' company (IR35 terminology) from transitioning to a SME, because you were not allowed to accumulate money in the company to allow you to start employing other people.
It meant that you could not compete effectively with a larger company which was exempt from IR35 merely because of its size, as you ended up with a much higher tax burden, and very strict controls about what you were allowed to expense.
Is there such a thing as a "vocational manager"? I don't think so. The typical British manager has got to that position by climbing the greasy pole through sheer determination. Let's be honest - how many of you managers reading this can say "I love my job as a manager - it is so creative and I get a real buzz from looking after my staff. I would freely give my time to ensure the well-being of the people and department I look after." No? I'm not surprised.
This results in that most if not all managers do not really have the relevent skills in the subject that they are trying to manage. The standard procedure is: "We don't have the skills (especially me - in a quite voice), so let's outsource the lot and if the supplier f*ck up, we can beat them up and blame any failure of the project on them."
True, decent managers are worth their weight in gold but fortunately for the coffers of most companies, they only have to pay for cheap, lousy imitations. Not so good for the drain cleaners unfortunately.
before being sold a 'solution'.
eg: it's useful to have a clue about what you actually need and what it should cost.
Some might even say, "Is handling large amounts of information a core activity of government? Yes it is, so why even think about outsourcing?" Perhaps because we have a political class that can't stand being told anything by people who know what they are talking about. Much more pleasant to deal with salesmen: always so eager to please and, hey, after all, it's only money from the taxpayer!
Yes you do need IT people within Government to manage IT, but you need them to manage rather than meddle. If you entrust someone to do a job then let them do the job. Challenge them, make them work harder, drive down costs, but let them do the job without interfering. An awful lot of people in the outsourcing companies were once employed by the civil service prior to being TUPED out. They were probably allowed to do the job previously, so why not now? You could probably argue that it was outsourced due to cost / project failure, but the meddling managers are still the meddling managers,they weren't tuped.
As a long-since TUPE'd ex-Civil Service programmer, I well remember that it was the managers who'd scrabbled hard to remain in the CS who were left behind to manage the contract. The managers who had something about them followed us into the private sector and began to learn the arts of contract farming.
As an aside, there are now more people in the goverment department for which I used to work managing the contract with the suppliers than used to provide the entire IT function.
My first reaction when I read the article was to feel the urge to laugh to the point of incontinence. What are they going to do? TUPE the private sector staff into the CS? Bwah ha ha ha ha ha!
...albeit many moons ago. It would help if the civil service actually paid something akin to market rates for developers.
Anyone with any talent left as soon as soon as they realized they could double their salary by moving into the private sector. I enjoyed my job there but succumbed to the lure of the lucre.
And as per my comments on Bury doing their outsourcing...
You need to be able to control your suppliers. You need to be able to set targets, achieve targets, that sort of basic management. Whether you do it in house or use a subby to do it what you need is a clear goal and a plan to get there. This is what is lacking in Government IT programmes much of the time.
Too much consultation too. A racehorse designed by committee and all that.
Having seen a lot of Govt contracts they all had very strong penalties in them that could be applied to the supplier if they did something wrong. However, like all good things in law in order to be able to enforce the penalty you need to be able to provide evidence of the wrong doing: in house Govt procurement people can't do that and in-house IT staff are so uninvolved they have no idea what evidence to look for and of course the supplier is not going to self incriminate.
I once sat in a meeting where the customer was telling me that we had screwed up, made an absolute hash of the project, etc. When I asked them to provide one single piece of evidence to back up their allegations they went straight back to to telling me that we had cocked up, didn't know what we where doing, etc.
Suppliers don't get hit with penalties because the customer does not have in place th contract management skills and techical managers who are needed to monitor performance and put together the evidence to show it is the supplier's fault.
Most recently I was thinking about the Typhoon spare contract and the relevant govt committee report stating there were *no* penalty clauses for late (or even non) delivery of parts.
But I'd certainly believe without *clear* measures of "success" and the methods to *monitor* how near (or far) the project is from them the question becomes impossible to answer*
*Of course if success were measured that *might* imply some civil servants would not get their annual performance bonus (IE not even *met* their baseline standard), which *might* explain why some projects don't have performance measurement built in.
in the UK (although I believe Scotland is different) it's just the fact that anybody who *does* anything, is regarding like shite.
Went to a presentation evening at my sons school last year. 3 local universties were there, and one guy, bigging up the benefits of a university education when on to tell us of a student who graduated 2 years ago, and was now earning £40,000 a year. What did they study ? Engineering ? Medicine ? Science ?
No, they studied "media and politics" and were working for a bank.
My brother jumped ship after getting his BSc, and is paid a lot more in the US (working for Uncle Sam !) than he would ever have made here.
Successive governments have been told that for 20 years, what makes anyone think they'll listen now?
Outsourcing is the same incompetent beancounter-driven false enconomy as the fad for replacing secretaries with self-service tools. Get the technical staff to book their own travel (or whatever) with a web app, and hey presto you can lay off a secretary and "save money".
The fact that you now have 10 engineers, on 60+K each, separately spending several hours doing what one 20K/year admin used to do for everyone in an afternoon goes straight through their empty heads. The net salary bill has gone down, and any loss of productivity is clearly the fault of the lazy engineers.
The same is true for outsourcing where you pay less money on the surface, but get less-competent people, who take longer to produce one of those fiascos that end up on page one as another waste of taxpayers money.
Come the revolution, accountants should be second against the wall, just after the lawyers.
It seems that *no* one does a perfect job of managing outsourced IT for government.
But some do it a *hell* of a lot better than the UK and there are lessons to be learned.
One *very* interesting point I had not considered up till now is that made the AC who wrote
"Some might even say, "Is handling large amounts of information a core activity of government? Yes it is, so why even think about outsourcing?"
On that basis *no* server//mainframe database system should *ever* be (or ever have been) outsourced, although desktops and desktop services like email should be a more reasonable task.
As pointed out time and time again business change *only* happens with *board* level buy in to the idea in the first place (UK Govt departments do have a "management board" at their top so this is not an abstract term relative to the UK civil service).
And of course that requires the *board* (composed mainly of I'll hazard a guess male ex-public school and PPE at Oxford types) to understand there *is* a problem, not just a committee of backbench MP's.
Parliament might start by tattooing on the forehead of *every* senior civil servant with IT responsibility this message
"NEVER trust an IT salesman"
Thumbs up fro giving *hope* that more effective IT is *possible* but I doubt much will change.
1. Get some people who know what the business process is in a room with some IT guys and write down the detailed requirements (don't get the supplier to do this as they don't know your business)
2. Write down a clear description of what you want the supplier to do, include milestones, deliverables and acceptance criteria and then ask supplier to price it up (if the supplier writes this make sure you read what they are saying and match it back against your requirements).
3. Make sure the business is bought in to the solution and the description of what the supplier is going to do. Both the business and Procurement team need to be joined up on this.
4. Make sure you link failures by the supplie to meet milestones, provide deliverables and achieve acceptance tests to liquidated damages but don't penalise the supplier ('cos in English law you can't) and take into account that a demand for £20 million is not going to be acceptable on a job that earns only £500K
5. Put in place a joint business and technical team to manage the contract and have a clear view of how they are going to measure supplier performance. Do not do anything that is not in the contract and do not forget to do anything that is in the contract. Keep a diary of when delverables are due and make sure that they are received.
6. Manage change, just 'cos its a small change does not mean there is no cost applicable. A day' activity will impact the schedule by a day even if they supplier does it for free. Lots of changes will result in exta charges and delays
Do all of the above and you may actually have a successful project which come in on time and if you're really lucky +10% above budget (which is pretty good considering)
If *most* govt IT projects were <10% over budget they'd be handing out Knighthoods.
BTW I dimly recall that the Prince II methodology says most of the same at some length.
I'd still say without Sir Humphrey and his chums feeling it's very important (and HMG IT cost overruns suggest they don't) *nothing* will change.
But I like your approach (especially item 4).
...to get in a couple decent technical architects to give independant ideas of what the project should cost and write a proper justification for the estimate. Hell you could probably get a team to play Marc Cohn's estimating poker to come up something.
First off it would show up the woefully woolly and inadequate requirements that come out of the Public Sector. Even the ones that are written by big con artists ^W^Hsultancies. If you can't at least derive some use cases they need looking at again.
They could also be able to test what is soaking up the costs and redo the MOSCOW analysis a bit better. Most of the projects I've worked everything was mandatory. Big Bang or nothing.
Then procurers could detect those quotes that are obviously too low because they are likely to be attempts to buy the work and then play the change request game 75% of the way through.
You'd also have a baseline for the responses to the requirements. Sort of model answers.
I've worked client-side and supplier-side in the Public Sector for quite a while and most of the procurement teams I've dealt with are like lambs to the slaughter because they are in the position of someone taking their car into a dodgy garage for an oil change and having to listen to the sucking of teeth that really means "KerCHING"
It won't happen for reasons already alluded to. They won't trust techies to give some idea of costs because it's a PM and business concern.
Outsourcing projects is a total nightmare, never get what is required, the out sourced company doesn't understand what is required, the specification is not detailed enough, more time is spent trying to sort problems out, when all is said and done, more time and money is saved by having the IT skills in house, that way the requirements are more clearly understood, the bod programming it knows what the client wants, and is closer to the customer, everyone wins. Only issue is that government do not pay decent IT wages, so be prepared to take a hit there, oh and don't expect job security or indeed any decent pension, that's all been taken away.
No outsourcing is not bad, it is a tool, and is therefore only as good as the people who use it.
Unfortunately it is also a board room fad, so it is massively and inappropriately used.
You also need to manage the outsourcers, but usually the idea is to ditch all in house staff who could do this as a "saving"
It also gets used to try and get rid of problems. CFO/CEO doesn't understand IT, so they outsource it, thinking that the problem will go away. It doesn't it just get more expensive.
That said, I have seen and used a number of good outsourcing projects, it doable, you just need to think about it before hand.
funnily enough the in house IT staff (back when IT development was all in house) created some benefits agency software more than 15 years ago. It's still running now and is the most reliable system the department use.
The government would save millions by removing all senior decision making IT staff and just letting the guys who do the work and support each office do their job, and even make up the contracts, because it will sure as hell be more informed and based in reality than they are now.
The problem with govt IT is:
1 too much politics, too many politicians and decision makers being wined and dined by salesmen working for the IT service vendors and so on
2 too many of the supposedly senior IT pros on the public payroll have never worked on a live system implementation
3 too many ex forces and the like working internally and for the suppliers, many of the "top security experts" are helicopter pilots by profession (I kid you not) and they are allowed to write the rules to which everyone else must comply, far too few people who studied computer science! Or who worked man and boy on systems delivery
4 govt buyers and sales folk from vendors always get in the way of senior users/sponsors talking to senior IT pros who actually know what’s realistic. Often the senior pros in the vendors know what’s possible but are prevented from talking to senior customers by procurement rules and wishes of sales folk to keep everything happy and positive. No brownie points for predicting accurately, he who over promises tends to win sadly.
5 top politicians "designing" IT solutions on whiteboards with bogus experts a la NHS IT shambles
6 far too many govt ITT's written by the same small group of people in the same consultancies PA consulting etc, half the time the ITT is the same regardless of the project, and marking system on which winning bids are selected is badly flawed
7 too many public school %ӣ$%$ ers, especially now with Cameron in charge, any fuckwit who went to Eton and pissed around on PC's is able to "advise" the govt while people with years putting in real systems are ignored
Etc
You missed one;
too many users think that their ability to logon to the internet qualifies them as an IT expert.
Much the same way as their annual holiday flight qualifies them to fly a 747 out of heathrow, and give design advice to Boeing on the finer points of aerodynamics
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2024
