Monday, June 24, 2002

And, disappointingly, it looks like the US Govt don't want to use our Gateway for their web site, first.gov. That's a shame. Still, always nice to see how other countries implement this kind of technology, so I look forward to comparing notes in September when the first stage is live.
Very nice of Kablenet to spread the news about the upgrade of the Government Gateway. And they note it's the "UK's most important piece of e-government technology", which is always nice to hear. The first part of the upgrade was completed this weekend. We'll do the second part on the night of the 2nd and during part of the day on the 3rd of July. Then we can start using the new features we're adding - and services will start to come online taking advantage of them in August. One strange thing in the Kable article is about the Gateway "offering direct access from the Internet". It's always been directly accessible, but I think what they mean is that we'll be able to connect Local Authorities via the Internet instead of them needing to be on the GSI.

Sunday, June 16, 2002

Andrew Pinder spoke at GC2002, covering the need for more intermediaries delivering government services. It's a topic close to my heart. One of the big goals over the next 12 months is to extend the "network of trust" so that if, say, an online bank trusts you to have an account, then an e-government service can rely on the bank to provide the authentication. So, the steps for claiming benefit, filing tax or whatever are reduced. The SMS work that we're doing is another part of that - can we extend the network of trust to the operator that provides cellular service? Intermediaries need to be able to easily access government services through online channels, wrap additional value around them (e.g. a pension provider could show what the state pension will offer at retirement and then show options to improve that) and make the service simpler to use (so on buying car you can get the insurance, the tax disc and the parking permit for outside your house in a single transaction from the showroom). Some time to wait before we get that clever.
The GC2002 Awards. I sat on the judging panel which met a few weeks ago in Boisdale's for a few hours discussion of the pros and cons of a couple of dozen projects. The winners deserved their awards. But, one of the common aspects was how many projects undersold themselves by not directing us to the real benefits, the issues that they conquered and the potential for development. Next year, the message must be "work harder on the project overview" and look at the projects that won this year to get the hints.

Thursday, June 13, 2002

I spoke at the Electronic Government conference on 29th May. Dan Jellinek at Headstar summarised my presentation ably ... 'The government is in danger of reproducing its over-compartmentalised, department-based structure on the web, according to one of the leading architects of its web policy. Alan Mather, head of e-delivery at the Office of the e-Envoy, told delegates at a London conference last month that the government's current working methods resembled "lots of matchstick Eiffel towers dotted around - they look lovely but they aren't actually doing much. "Now we are in danger of repeating this on the web, with already some 2,138 '.gov.uk' sites, no sign of critical mass, no common design, limited transactions and no customer focus." He unveiled 'Mather's laws' of e-government which included "the number of technical solutions implemented by departments to solve identical problems quadruples every twelve months", and "80 per cent of the money spent on e-government to date has been spent on things the customer never sees". Some good things are happening but a far greater degree of consistency was needed across government, he said. "If you use Microsoft products you always know where the save function is. But across government web sites you don't know where to look to find forms, or search functions, for example." He said the way forward was for convergence across government on common systems and applications, with departments co- operating to solve problems. There was also a need for a far greater number of compelling, citizen-focused services to drive take-up, although "there is probably no single 'critical app' - you need more clusters or packages of useful, transactional services, perhaps with private [sector] services too".' I checked earlier and the slides were available for download if you'd like them.

Monday, June 03, 2002

Back on May 7th, Realnames went under. The story was widely reported from at least one side, that of the company head, Keith Teare There's another side to the story reported today (and tracked thanks to Dan Gillmor's site). The Realnames premise was that if you typed in 'coca-cola' in the browser address bar, then that's where you'd want to go. Funnily enough, I just tried that and where did I go? To a Realnames-powered page that takes you to the coca-cola web page. But if you type in "general motors" you go to an MSN-search page, where General Motors (the brand) is not the first entry. Of course, I'm using IE. I expect the results are different with Netscape. Similarly, if you typed in 'inland revenue', then you go to a Realnames-powered page that includes www.inlandrevenue.gov.uk. Neat, huh? I wonder if there's evidence that people type brand names direct into the tool bar looking for these pages? Isn't it more likely though that, rather than type in a brand name, you'd type in something more generic? What if you typed in 'pay less tax' ... would you go to a site promoting off-shore bank accounts, one that offered some tax exempt saving schemes or to the inland revenue's feature on ISAs or TESSAs? If you type this into IE 6 it takes you to another MSN-search page, where the first entry takes you to amazon.com and a book called "loopholes of the rich". Not what I was looking for. Trying this on google takes you, first, to paylesstax.org. Also not what I was looking for. Typing 'pay less tax' into ukonline.gov.uk gives you government's take on what you thought you meant - four or five pages of links to the Inland Revenue, none to Customs, one to National Insurance. The reason I'm thinking about this today has nothing to do with who is right and who is wrong in the Realnames case, but to wonder about how people find information on the web that they need. I'm coming from the angle of joined up government. Do people use search engines to help them find what they need, or would they rather a list of things that they can sort through, perhaps grouped by time of life or similar? Or is the ultimate tool a personalised set of topics based on what government already knows about the person, or derived from some information volunteered? Maybe it's a bit of all of them. What I don't think people want is a collection of links to various sites that they then have to sort through to find what applies to them. Personalisation must be the way to go. That requires content aggregation. That requires common taxonomy. And common standards for updating and managing data. And it means that the services offered must be useful enough for people to want to give up some personal data. That's a big step forwards from where we are. We are probably far enough ahead to collate data anonymously and add some data based on what we can assume (e.g. from post code), but certainly not smart enough yet to link that to what government genuinely knows about you as a person. So one size fits all is still our model and we're some way off one size fits one.

Sunday, June 02, 2002

Let me take you back to June 1999. Folks have said that government IT typically runs perhaps 2-3 years behind the commercial sector. That would make the timing of that last article about right - it notes a significant loss of service on Ebay. It also says that Yahoo is using the opportunity to gain business from Ebay. That's the same Yahoo that closed down a bunch of its auction sites in Europe a couple of weeks ago and said that it would promote Ebay instead. But then again, that wasn't the first time and it wasn't the last. And then there was Yahoo's DOS attack. Marc Andreessen says that sites only get famous when they go down ... AOL, Ebay, Yahoo, Etrade etc. Outages on commercial Internet sites aren't new. They aren't even new to government sites. Internet technology is new technology when technology fails it's usually not trivial or easy to diagnose (especially after it's been live for a while). The examples above are just a few of many. Typing "outage" into Google gives more than 460,000 answers and most seem to be noting hardware, software or network outages, planned or unplanned. The most bizarre, scanning quickly was at freshmeat.net - it's not at all what you think! In the corporate world, just as in the government world, if your internal systems fail then no-one but your staff notices - but the moment you put those systems on the Internet, you expose yourself to a whole new set of risks. Flaws in your internal process that can be covered up by "work-arounds" are immediately exposed. Outages that mean work piles up in the office for a while until the system comes back online affect your customers. And that means stronger controls, more testing, more rigour and, in government's case, better partnership - within government and with the private sector. Commercial sites have recovered from these issues and they did that by addressing the problem quickly and by working hard to regain the trust of their consumers. That seems like a good message to me.
The Inland Revenue's online service for the Self Assessment tax return has been taken offline. Kablenet had the best story. I can't comment on the whys and whats, but I do know that the IR are taking this apparent security breach very seriously. There are a lot of smart people working on it and when the issue is sorted, SA online will be back. In the meantime, if you have a 3rd party software package (like the one available online from Digita) you can use that to file with no issues. On a broader front, this should be sounding some alarms. If the IR can have issues with an online service given the expertise and the experience that they have, then what chance do the smaller departments, agencies and (especially perhaps) the local authorities have? For the last couple of years I've had a pretty consistent theme in my presentations about how hard this Internet stuff is, how unlikely it is that any single department can cope with it and, therefore, how necessary it is to work together. There aren't enough checks in the present process to direct departments to work together, so it has to come from the departments themselves. Natural groups should form around client sectors (students, mothers, etc) or trigger events - (need a loan, looking to change jobs, having first child etc). Trying to take these on in isolation will result in certain failure - whether it's because the service won't meet the client need or it's not interesting enough or it just plain doesn't do what it's supposed to do. I've usually led with the Government Gateway as a prime example of this working together and, more recently, with the UKonline platform that we've developed to host many departments. That's a bit of the puzzle, but it's not all of it. We need to bring focus to these major events and sector groupings so that joint working becomes the norm, expenditure is reduced and all the expertise is concentrated in the right areas. When you visit Amazon, you don't go to different sites to buy DVDs, CDs, or books; when you go to the Sunday Times, there aren't separate user interfaces for the money section or the news review ... how else could it work? So, rather than make people use their brain power to figure out how to work through the government process, we need to harness our brain power on making it easy for people to find what they need - consistent interfaces, simple descriptions, common text and so on. This won't happen by accident. It needs sustained effort and a real willingness to work across government. That means that the constraints that the organisation imposes on us now: accountability, risk aversion, scale and complexity need to be addressed rapidly.