- A good day to fly
- A good day for industrial diseases
- Bandwagon of minor achievements
- The evidence mounts
- The cost of IT failure
- Getting the IT to work
- Digital talent or digital madness?
- A reader quotes
- A policeman fights back
- Dumbest targets of all
- The minister’s on TV
- Fiddling while people suffer
- Pretenders duck the big lever
A good day to fly
Concerned that the public-sector strike would mean blocked airports, the government pulled together thousands of civil servants to man the border. But on the day, passengers were delighted with the speed of getting through immigration, remarking it had never been so fast. As well as more desks being open than usual, thousands of people were on standby at the airports just in case a queue developed somewhere and there were reports of marquees and queue-management plans in case queues developed. But there was none.
It is incredible, this is a highly regulated service, we know how many planes are landing and no plane can fly without knowing how many people are on board. For many years I have told the same joke as I wait in the queue for immigration along with thousands of other passengers: ‘they didn’t know we were coming!’; methinks its funny because of course they do. But now we learn they don’t!
It should be a simple task to resource immigration desks to cope with known demand; but ministers just threw money at the problem. When we get back to normal it will be costs that return the desks to being a constraint on getting through the border.
The government has announced that VAT will no longer be a requirement for shared service deals in the public and voluntary sectors. For private-sector shared services providers, VAT has been the barrier to entry, but now it is lifted. Clearly ministers want the private-sector to join the shared services feeding frenzy. See the news here:
Recently in Dublin I asked the civil servant responsible for driving shared services in Ireland if he’d seen the evidence. He said he had. I asked who gave it to him. His answer was the UK Treasury. So I sent him our collection of evidence. IT-led industrial shared services designs fail, and fail big time.
The Local Government Association has published a summary of shared services ventures amongst local authorities. It amounts to a measure of compliance; sharing services being the current ministerial mantra. But look at the projected savings: over 143 shared services projects, the savings expected amount to £153m, that’s about a million each and that is extremely modest and, I suspect, can only represent savings through less-of-a-common-resource, always the minor saving.
At one level we can take that as good news, for it means they will not be going for the big IT-led industrial designs which always make large but erroneous claims for lower transaction costs. But it is also a lost opportunity, for local authority services can be massively improved through studying them and then redesigning them to manage value. The bandwagon of minor achievements represents compliance steering people away from opportunity. And bear in mind these savings are projected, unlike the savings achieved from studying and redesign which are delivered, sustained and thus real evidence.
Meanwhile more evidence of failure, this time from the NHS where shared services are creating lots of errors – a normal phenomenon in industrial designs because they try to shoehorn everything into standardised processes. Read the news here:
These problems don’t go away, they just settle down, eventually, as high levels of failure demand. But as the industrial designers and out-source providers treat all demand as work to be done the phenomenon remains not understood and as out-source providers are paid on transaction volumes, they wouldn’t want it understood.
Readers may recall my criticisms of the UK Research Councils’ shared services venture. It was blogged and the blogger followed it up by interviewing with the man in charge, David Delpy. Delpy claimed, in the summer of 2010, that all would be fine in the end. If you want to review the arguments follow this link:
RCUK Shared Service Centre ‘a fantastically expensive failure
But now we have the National Audit Office’s recent report telling us that the only way efficiencies will be delivered is if the UKRC takes on more work, a sure sign the NAO doesn’t understand the problem and a strong signal that the shared services venture is up the usual creek.
The NAO reports that the project was due to be completed by December 2009 at a cost of £79 million. But it was not completed until March 2011, at a cost £130 million. Further, we learn that ‘delivered savings are adrift of the business plan by at least £73 million’ and most damning of all: ‘implementation of this project does not represent good value for money’.
You can read the news story here:
Meanwhile, Ian Liddel-Grainger MP, who has campaigned about the costs of SW1, a joint-venture shared services between IBM and two councils in the south west, reports on his blog that many of the services are being returned in-house:
The local television station reported the beginning of the review in March this year:
And once the report was out, it made the news again. One alarming revelation is that the councils simply could not afford to walk away from the arrangements, as to do so would incur ‘enormous’ costs. It ought to alarm every council leader who is being asked to sign up to a similar industrial scheme. See the TV report here:
One of the reasons for shared services programmes failing is large-scale IT. Quite simply, most large-scale IT projects fail. At the risk of boring you, I think every politician should read ‘Dangerous Enthusiasms’ by Gauld and Goldfinch, for the evidence.
In Australia, where, as I have mentioned before, it is easier to get information about the costs of failure, the state of Victoria’s Ombudsman reports the ‘most damning assessments of public sector IT project governance in Australia’s history’. In the last five years alone, there have been cost over-runs of .44 billion, extensive delays and a general failure to deliver on stated aims.
The Aussies, like our ministers and IT vendors, think it is all due to problems of governance and the like. It is not. Even where you are lucky enough to get the IT to work, you get massive costs associated with industrial design.
We have been doing more work with IT developers in our private-sector clients. As many readers will know, our approach is: understand (study the service as a system), improve (don’t touch the IT, except to turn it off), then ‘pull’ new IT into the better design. The result is IT that works and you spend massively less on it.
We think we might need to give this approach a name, because that’s how things go in IT. Big IT projects used to use the ‘waterfall’ method, now the big buzz is for ‘agile’ (doing the wrong thing faster); so to provide a frame to explain how our approach is delivering far better results, maybe we should give it a name. Waterfall’s method is design and push down, Agile’s is design through successive iterations, ours is study, improve, pull. Got any ideas for a label?
In response to last month’s item on government hiring more ‘digital talent’, a reader writes:
The ‘digital talent’ announcement ‘says that it’s ‘a move away from long-term, gold-plated IT projects’; funny, it doesn’t look like that to me, in fact just the opposite.
This is ridiculous, especially as it’s all supposed to have been done with ‘BVPI157′ (getting all govt services online by 2005); then transformational government; then the Direct.Gov portal through which all govt services were supposed to go but don’t; and now they’re working on the alpha.gov portal which will be so good it’ll do everything. Except perhaps explain why it looks like all it does is duplicate Direct.Gov which was itself supposed to do everything…
When did I become so cynical?’
I guess when he realised it exemplifies the definition of madness.
‘Here are two quotes from Goethe, that I recently come across, that may be of interest to you, if you have not already heard them.
‘There is nothing more frightful than ignorance in action.’
‘There is nothing worse than aggressive stupidity.’
These come to mind as I read your articles on govt approach.’
Regular readers will be familiar with Inspector Guilfoyle, a blogging policeman who does great work. Two recent blogs: The first where he is taken to task by a senior officer on the need to work to targets (which he won’t): http://inspectorgadget.wordpress.com/2011/11/06/we-are-still-obsessed-with-targets-shock/
The second where he argues against specialisation:
Divide And Don’t Conquer
Both are worth a read.
Simon Caulkin was at our recent policing event where Inspector Guilfoyle spoke and he has summarised the presentation in an article: http://www.simoncaulkin.com
We should take our hats off to Inspector Guilfoyle.
A leader in the health service writes:
‘ Two years as a director of an NHS trust, I have only just realised just how dumb the ‘treat 85% of admitted (90% non admitted) patients within 18 weeks of referral by their GP’ is. The performance against the target is measured as the percentage of those treated within the month (or quarter) for which performance is being measured who had waited less than 18 weeks. Assuming that being treated in under 18 weeks is better than being treated after 18 weeks (not always the case – if you have mild knee pain and don’t want surgery before you go on a world cruise) this target may have some proxy value for good care (relative to some others at least!).
The issue is that if you have a large backlog of patients who have waited over 18 weeks, then this can go unnoticed UNTIL, you try to clear the backlog (arguably a good thing to do for patients) when these patients, some of whom have been waiting 12 months unnoticed by regulators, suddenly trigger as an 18 week breach dragging performance down.
Therefore ‘smart’ trusts (who put targets ahead of patients) who want to hit their targets will continue to ignore their long waiters and hit their targets. UNTIL someone notices that certain trusts now have very large numbers of patients waiting very long times.
We as a board spend so much time ‘holding directorate managers to account for hitting their targets’ and debating how we can make a case to arbitrators around how targets are measured (e.g. if another hospital refers a patient 40 days into a pathway, should we still be accountable for the overall 62 day target or should we split the target with the referring hospital?)’
It is tampering on a massive scale.
A reader writes:
‘The minister for health appears on a non-stop message on hospital TV. To turn him off you have to register, and when you are ill, in hospital, do you really want to register on some godforsaken screen?! The screens are hard to operate too – not at all intuitive.
One patient said ‘It was eerie. Everywhere you looked there was Andrew Lansley. My mother-in-law had to keep topping up the machine just to escape him.’ It reminds me of when I visited a patient in a short stay ward and heard a nurse tell a patient that ‘it’s cheaper if you pay for 10 days’ – in a short stay ward!’
A student was given access to an ambulance service. His dissertation was concerned with a new fuel-monitoring scheme for ambulance drivers. I expect you know the sort of thing: a device in the vehicle that gives feedback on fuel use, in common use with haulage firms because careful drivers use less fuel. There was conflict between ambulance drivers and management over the scheme, described by management as ‘resistance to change’. The new monitoring system imposed fuel-use targets and drivers were subjected to one-to-ones and ‘performance management’.
It was immediately obvious to the student that fuel usage was largely outside of the drivers’ control. Studying what was going on he could see that ambulances were often sent to buildings that were new and thus not on the database, meaning they’d have to drive around to find the location; multiple vehicles were often dispatched to the same call, causing unnecessary journeys; vehicles were dispatched before problems were understood, leading to unnecessary journeys and longer response times for incidents that did require an ambulance.
Digging deeper he saw that the software used to standardise decisions of call-takers was compromising the effectiveness of dispatching decisions, standardised responses could not accommodate contextual circumstances, causing unnecessary journeys and compromising safety.
He made a report to management. It was ignored. His advice was that managers would be better to spend their time sorting out how to get new buildings on to the database and improving decisions about what ambulances were sent where. In a short time the student had seen the way to improve this system. He should be given an A grade, but his university lecturers admonished him for his impertinence, his critical dissertation jeopardised their relationship with the ambulance service; which for them was merely a place to send students.
I told the story to a friend working in health. He said: ‘It gets worse. When the paramedics turn up their decision-making is shaped by the priority which the dispatch team has given the call. If dispatch said it was urgent then the paramedics tend to go into a ‘just in case’ mindset, conveying into A&E completely unnecessarily. It’s just like the schedule of rates driving the wrong repair to be done.’
Meanwhile ambulance managers are fiddling around with their new management toy trying to control drivers’ driving. You couldn’t make it up.
We only have ourselves to blame. We have put lots of our public-sector case studies in to the public domain. So they get copied by the people I call pretenders – people who pretend they do what we do.
I have just been shown an example: Consultants taking our intellectual property and using it to help local authorities redesign homelessness. In summary, it amounted to no more than clever process improvement work; all references to changing management thinking were removed and the intervention design treated management as a steering group (what a joke), not leaders of the change. Obviously that way you avoid difficult territory but you also fail to gain the greatest lever. You can’t change the system without changing thinking.
By contrast, I was told what our newest private-sector client had to say about Vanguard. Addressing an audience who were about to start work studying their system he said that they chose Vanguard for two reasons. The first was Vanguard is brilliant (pardon my blushing) and the second: that Vanguard can sometimes hack off senior managers because we challenge them. Actually he didn’t say ‘hack’ but if I used the work he used, this newsletter might not get past your firewall.
Just to be clear: we don’t arrive all-guns-blazing, in fact our first interventions are with top leaders to help them ‘get it’; without which we won’t proceed. But in big companies where you can’t get everything moving at once, you always find things coming out of left field that represent the old philosophy and it is our responsibility to help prevent the inevitable conflict, primarily through briefing leaders who have ‘got it’ on how to tackle the contradictions. We know from bitter experience that if you don’t you can only look forward to a train crash down the line, and that won’t help anyone.
So yes, if you want to help leaders change their thinking, there will be times when you have to challenge them. Thinking is the big lever.
And despite the growth in pretenders stealing our intellectual property and doing their safe and often wrong-headed versions of our work, we will continue to publish public-sector cases. My advice to people in the public sector is: if you don’t use our experts, you’d be better off going it alone.
Next year I shall be publishing lots more to help public-sector leaders do it themselves.