£90.3m - money well spent!

I can't resist it. I was going to quit, honest. Yet how could I have missed this Guardian article from December last year?

http://www.guardian.co.uk/government-computing-network/2011/dec/09/national-audit-office-digital-investment-unclear-directgov-government-gateway-businessgovov

Benefits of government's £90.3m digital investment 'unclear'

National Audit Office accuses government of failing to measure the benefits of Government Gateway, Directgov and Business.gov

The government has failed to routinely measure the benefits of its main portals - the Government Gateway, Directgov and Business.gov - which together have cost £90.3m over the past three years, says the National Audit Office (NAO).

In its report titled Digital Britain one: shared infrastructure and services for government online, the spending watchdog accuses the government of making investment decisions without sufficient information on costs and benefits.

In 2005 the government began converging online services on Directgov and Business.gov in an effort to reduce its public service websites, of which there were more than 2,500. Since 2006 1,526 government websites have closed.

The NAO found only one instance where the government had estimated the benefits of its investment in online services. Business.gov, which provides government information for businesses, was reported to have saved business £21 for every £1 spent in 2010-11.

Although there are likely benefits to providing business information in one location, the NAO found that it was not possible to say how much of this benefit would have been delivered anyway, if the information had only been available from multiple websites.

There's another figure for how many websites got closed as a result of convergence - 1,526 in fact.

Sounds like that NAO report could be jolly interesting reading.

Here's my earlier speculation about the projected cost of government supersites and the reality: Whatever happened to that 400 million? 

Betagov, Betagov - wherefore art thou Betagov?

With apologies to the Bard. As you might recall from your A-Level English, Juliet's lament means 'why are you Romeo?' rather than 'where are you, Romeo'? Indeed, over 2012 people might be asking where Betagov has gotten to. They've got a lot of user testing to do before they come up with something that isn't worse than Directgov itself. Combining a government supersite with accessibility is like ... damnit, Dorian ... Shakespeare would have nailed that with a simile. What's wrong with you tonight?

Nonetheless, the question that the test subjects, the big players like DVLA, DWP and Student Loans Company and eventually GDS themselves are going to be asking is: why do you have to be a Betagov at all? What was so wrong with government services having a website each?

GDS are doing their homework, of course:

http://digital.cabinetoffice.gov.uk/2012/01/27/search-data-user-needs/

There are over 300,000 searches carried out weekly on Directgov, with over 125,000 different search phrases. The most popular term, ‘jobs’, is searched for 4,500 times a week. At the other end of the spectrum there is a ‘long tail’ of 100,000 phrases that are only searched for once ...

Using this data to identify needs that are not being met by government can be daunting. However, it is possible to filter or group it to pick out emerging trends and unsatisfactory user journeys.

125,000 different search phrases. I would expect the number of different search phrases was somewhat lower on each of those 287 websites you killed and absorbed like some glutinous horror from a John Carpenter film. You're going to have your work cut out improving public services now.

It's OK though, you give one example of something Directgov's dramatic intervention managed to improve -

During the run up to a recent Christmas, a growing volume of searches that included the words ‘Christmas’ and ‘payment’ was spotted. We also noted that users were not clicking on the (irrelevant) results presented to them. Delving deeper, we were able to see what else those users looked for. This identified an unmet user need: benefit payment dates over the Christmas period. HMRC published an article on Directgov and, as we were able to supply the relevant keywords that users were searching for, the search engine-optimised article ranked well in Google quickly.

Do HMRC not listen to customer phone calls at all? Or check their own web analytics? I would have thought that people have been phoning up about their xmas benefit payment dates every single year since the dawn of the welfare state. Some local knowledge would have paid off there.

Luckily Directgov swooped and stuck their HMRC article on a seperate supersite; and thus another feel-good piece of anecdotal evidence about public services being improved by centralisation was born. [note before publish: explore Spielberg metaphor here].

I expect you've got your beady eyes on the vast HMRC website these days; not the complicated transactional stuff where you submit your tax returns, of course - just the pages of static content. They should be easy enough to prune, transplant and paint a non-branded Betagov white. You'll be needing some success stories in the next couple of years so converging HMRC might do the trick.

Haha, oh dear, I slipped into the second person again. I used that trope on my last post. Probably best I leave this one be. Maybe I'll go and comment on GDS a bit.

What happened to that £400 million? Part 2

Er, I'm stuck now. I Googled 'varney report savings' and nothing came up about whether the £400 million savings predicted by Varney was actually saved. Nothing whatsoever. Did they save only £200 million? £2 million? £2? 20p? A 2p chew?

You can put together the total running costs for Directgov if you check a couple of sources -

http://www.theyworkforyou.com/wrans/?id=2009-06-30c.282044.h&

Jim Knight (Minister of State (the South West), Regional Affairs; South Dorset, Labour)

The costs for Directgov in each financial year since 2004 are set out in the following table:

Financial year	£ million
2004-05	5.1
2005-06	10.5
2006-07	12.8
2007-08	13.9
2008-09	30.7

...

I believe after this the figures were approximately - 


2009-10 - £30 million
2010-11 - £23 million

... bringing the Directgov total to £126 million, over seven financial years.

Have I missed any millions here and there? According to Hansard, 'Expenditure on Directgov increased in 2008-9 in recognition to its increased importance in the Government’s strategy for online delivery of public services.' - in other words, the £83 million invested in Directgov since Varney, approved by the Comprehensive Spending Review in 2007, was to achieve the £400 million of savings by converging websites.

So, where's our extra £400 million?

Haha, only kidding, government is a big, complicated thing. You can't expect theoretical savings to appear on actual balance sheets. As Cross pointed out, Varney's figures were 'back of a fag packet' calculations. No-one seriously expected to save an actual £400 million.

Bit of a shame for Gubbins, mind you. Remember, our costs went up when we converged to Directgov. We have our own customer transactions which we maintain ourselves and are paid for out of Gubbin's annual budget.  They're painted orange with a Directgov logo at the top, but they're essentially built and maintained by us.

Y'know if Gubbins had pledged to save, say, £5,000 a year by automatically switching our office PCs off every night, HM Treasury would have the right to see these savings somewhere on a balance sheet.

For Directgov, normal accounting practice doesn't seem to apply. As I said in my first post, we have to make special allowances for government websites.

I’d go direct, guv

Directgov's spending was cut in 2010 along with other cuts in the public sector. At the height of Directgov's 'success' they were spending £7 million per yearon marketing. That's the figure I seem to remember, anyway. It's a bit hard to find on Google now.

As recently as February 2010 Directgov were commissioning a £2.05 million TV ad featuring various B-listers, earning the censure of the Daily Mail - 

http://www.dailymail.co.uk/news/article-1253496/How-star-studded-Government-TV-advert-featuring-Kelly-Brook-Helen-Mirren-costs-taxpayer-2million.html

A star-studded Government television advert - featuring the likes of Suggs from Madness and Kelly Brook - has cost the taxpayer more than £2million.

The eye-watering cost of the Directgov campaign includes the production, airtime and hiring the celebrities.

Quite a price tag for marketing a brand which would be closed down within a couple of years. If they do get rid of Directgov. But that's another story.

Hey, look. There's the Directgov problem again from a commentator -

The first time I saw the advert I was shocked by the self-indulgence of it and the ridiculous number of celebrities. I'd rather they spent the money on improving the clunky DirectGov website.

- Chris Z, Warwick, 24/2/2010 23:07

Sorry, I digress. The funny thing about this episode is that the Varney savings were still being cited three years after the Comprehensive Spending Review had started:

But Mike Hoban, the communications director for Directgov, said: "At a time of economic uncertainty it is essential that we give everyone in the UK easy access to important government information about taxes, benefits, job opportunities and education.

"Directgov will save the government £400m over three years. Therefore this is an investment that is important in helping the government save money."

Hang on a sec ... this was February 2010. The web convergence program had already been running for three years. Hoban was talking as if the £400 million savings were going to be made over the next three years.

It would have been an ideal opportunity for Hoban to have shared a spreadsheet of actual savings from the various government organisations who had closed down their websites to join Directgov; seeing as by then, Directgov had already swallowed £96 million of taxpayers money in pursuit of this £400 million.

Digital by default

As far as government white (orange?) elephants go, Directgov is a snip. Compared to £2.7 billion reportedly lost on a replacement NHS system, £96 million on an unloved orange website, and £400 million of vanishing theoretical savings will soon be forgotten by the general public.

The point is, when a government organisation runs its own web services, its expenditure shows up on balance sheets. After each financial year, the government can audit them on what they've achieved with their budget.

When you centralise web services into a government supersite, it becomes more difficult to account for expenditure. The ideals are too lofty, the aims are too vague, and the politicians don't want to be bogged down in the details. It happened with Directgov.

Let's be fair, though. If we haven't saved £400 million, at least we cut down on the number of government websites out there. According to Computer Weekly in April 2011:

http://www.computerweekly.com/blogs/public-sector/2011/05/500-rogue-gov-websites-nabbed.html

Sharon Cooper, director of strategy and innovation for the Government Digital Service, told a recent Inside Government conference the unit had achieved Varney's target off shutting all unnecessary public sector websites and subsuming them into DirectGov by March 2011.

It had shut 287 websites by 5pm on 31 March, converging 95 per cent of all public sector information into DirectGov. But it had found another 500 websites that must be axed.

"There are still another 500 out there because we found a lot more in the process of trying to shut them down and that work is still going on," said Cooper.

Oh dear, were they hiding? So, we're still investing public funds to reduce the number of government websites, even though after five years, there is no evidence this causes any savings.

Again, as she mentioned Varney, this would have been an ideal opportunity to share a spreadsheet showing £400 million in savings.

Erm ... I'm confused again. How do you measure 'information' unless you work in quantum physics? They're reported that statistic as if it's a real number.

Real numbers versus Directgov numbers

287 is a real number, mind you. It wouldn't be hard to email 287 organisations and get them to say how much Directgov saved them.

Presumably the total comes to at least £83 million pounds - the running costs of Directgov, post-Varney. That means £289,000 per website.

A professional website can be built for £10,000. I'm no expert on web infrastructure but I imagine the hosting needs of a government organisation would run to roughly the same amount of money. So, a new information-carrying website could have been built in 2008-09 and hosted for three years for £40,000.

Seven such websites could have been produced for each website which was converged onto Directgov. I would expect that if you ran user testing on said £10,000 website it would perform better than the same information being presented on a government supersite.

Y'know, a broke graphic design graduate would probably build you a decent information-carrying website for £100. Maybe I'm being naive about how much websites cost in the public sector? This is all getting terribly confusing.

Remember, Directgov only hosts pages of text. It has no interactive functions, unless you could COTA boxes and the email service. The transactions (people ordering a new driving license etc) are still hosted by the organisations themselves.

So, like the Varney figures, this all sounds ever so slightly fishy.

That '500 websites' number sounds a bit made-up too. Did Varney have 787 websites in mind when he came up with those £400 million of savings? By the sounds of things he forgot to tell Directgov.

Post script

Returning to the Daily Mail article:

But Shadow Cabinet Office Minister Francis Maude called Government advertising and marketing spending ‘out of control’.

He added: ‘Labour seem more focused on squandering our money on vanity PR projects rather than actually addressing the pressing problems of the country.’

It's interesting that the £2.05 million on an advert attracted Maude's condemnation at the time; but £126 million on a government supersite didn't influence the direction of government web services under Maude's control after he took power.

Hands up who likes Directgov?

Not Gubbins, that's for sure. Back in 2007 we began the process of shutting down our own website and moving the content across to Directgov. We had our own website, just as we have our own forms, letters, marketing, call centers, IT, mailroom, lawyers, accounting, printing, reception and tea lady. We clean our own offices. None of these things are run from Whitehall. But it was deemed that our website should be.

In 2007 Directgov were already running a Product Review. direct.gov.uk was going to get new colours and layout. It was a brave new world. Directgov accepted they weren't very good and needed to change.

We didn't have anything planned for gubbins.co.uk. Everyone agreed it was fit for purpose.

We went ahead and converged anyway. There were a few tense meetings about it at Gubbins board level. One of the Gubbins board also sat on the Directgov board, although I'm sure that had nothing to do with it.

The migration of Gubbins

We turned the Gubbins web content into Directgov web content. It wasn't easy. Because of The Directgov Problem we couldn't trust our customers to navigate Directgov properly. So, we had to put everything on big pages. There are no anchor links to help people find anything. We had to make the most of it. Our service is now run from Directgov and our letters and publications were updated to send customers to direct.gov.uk/gubbins.

I work in the web team. We don't like it. As a rule of thumb, an update which took an hour now takes a day on Directgov. We can't make our own updates. We have to send everything to a Directgov editor in a word document.

We keep getting different Directgov editors to work with. When a new one starts, they don't know anything about Gubbins. They don't seem to know much about writing for the web either. When we send them something to update, they rewrite it. When we explain why we want something written a certain way, they argue. When we want a new page, they have to submit a form to the shadowy Publishing Board, who we've never met.

According to Government on the internet: progress in delivering information and services online By Great Britain: National Audit Office 2007, the UK is the only country who has tried to 'centralise content' across the public sector web. It's funny that no-one's run research to see if centralising content actually works.

Directgov gave us access to the Stellant CMS and Speedtrap analytics, but these are slow, old and tired. Chances are they haven't been replaced since 2006.

Moving to Directgov was supposed to save Gubbins money, but our costs have gone up. We have our own web infrastructure, so we haven't saved any money on hosting. We have the same number of writers and developers but they spend their time negotiating with people from Directgov.

We still run the same online services. Directgov the website doesn't really do anything. You can't book an appointment with it. To do that, you have to leave Directgov and move to gubbins.direct.gov.uk. We still run these screens but they're painted orange and have the Directgov logo painted on them.

Every time we build or fix something, we have to run it past the Directgov editors, the Directgov design team, the publishing board, and usually some other executive with a clever job title who we haven't heard of before.

'Why can't we get our website back?'

Dear me, if I had a penny for every time I was asked that.

Our directors still don't like it. One of them worked out the move cost us an extra £200K of avoidable phone calls. Half of the people who phone try finding the answer to their question on the website first.

Our call centers don't like Directgov either. They keep asking why we have to use it.

Our writers and marketing people don't like it. They keep writing 'we' when we have to say 'Gubbins'. We have a teeny Gubbins logo on our landing page of Directgov, rather than a website built by our own designers.

Our customers don't like it. When we show them a gubbins.co.uk prototype build by our own designers, they prefer it.

Our stakeholders don't like it. Even people who don't work on the web for a living know about things like local navigation, breadcrumbs and tabs. 

The product review in 2007 didn't come to anything. Or Project Austin, which came after that. Apparently everyone is excited about Betagov although Gubbins aren't so sure.

Directgov don't like Directgov

You'd think there was someone out there who liked Directgov. But Directgov the organisation don't like Directgov the website either.

You'd think there was someone out there who could say - we built Directgov. We've shown that it's improved the public sector web. The design is fit for purpose. It's as good as any website that was built by the private sector.

Such a person would be a web designer. But Directgov doesn't have any web designers. At least none that have ever visited Gubbins. It has plenty of publishing, PR and marketing people. In fact, that's all it seems to have. They outsource anything technical or anything which involves building websites.

In late 2010 Directgov paid Gubbins an actual visit. They didn't meet our web team, of course, just the board of directors. It was about the time Jayne Nickalls and the top brass all resigned or retired.

They admitted Directgov wasn't very good, just as they had in 2007. However it was 'not Directgov's fault'. Plus we'd had the benefits of cross selling from the Directgov platform. And there was a radical new Directgov on its way called Alphagov.

Let's just hope it happens this time.