Monday, December 18, 2006
A good albeit apparently justifiably anonymous comment on a recent post, "What's a Website Worth?". I've edited out the bits I thought worth airing and commenting further on: Let's leave aside the design/navigation issues the site still has to overcome - it's a big task, and the site is making progress against a tough objective. But even if Direct.gov were a fantastic, robust, easy to use site - is it still right to turn off all the others? Isn't there something to be said for promoting competition in approaches to government on the web along the 'let a hundred flowers bloom' line of thinking? With a single CMS, a single set of editorial guidelines, a single design template - how will central government sites keep pace, innovate, and become better still? Sure, having more sites will cost more money, and the argument for smart cross-linking is undeniable. There's also definitely a case for culling the existing jungle of government sites, but why not cull the weakest 25% year-on-year, and support the brightest 5% to see what we can learn from them? In my quest to keep things simple, I've always gone for "one site" because it cuts the debate down. As soon as you say "ok, we can have more than one", then everyone wants their site included and we argue indefinitely about what the criteria for inclusion are (believe me, I've been there). I realise that "one" is not the right answer just as "thousands" isn't. The important thing is to recognise where you're going to invest and where you're not going to invest. Everyone who runs a website in government believes that they can have the one of the "brightest 5%" if not this year, then next year, with just a little more money. I absolutely buy the idea of seeding some sites with funding. Let's say we take the top 5% by market penetration. We'll keep that simple and say it's the number of customers they could have times the ratio who are online. So direct.gov would be 2.8 million out of 36 million (60 million times 0.6, where that's the ratio of the country online, give or take), or 7.77%. That sounds good, but, Hitwise figures for November 2006 show google.co.uk at 8.06% and ebay.co.uk at 2.98%. A year ago, the same source had direct.gov at 2.8%. With direct.gov's traffic roughly doubling over the last year (I'm being only a little generous) then somewhere around 5-6% probably isn't out of whack. That works for established sites. It also works for sites that have a chunk of marketing money. It's hard to imagine, say, www.salt.gov.uk (one of my favourite domain names) getting much marketing or much traffic, so it should rule out the obviously banal sites. But what about a sexy new site? How would it get launched? How would it make its case so that it could be one of the 5%? And who would it displace? Nobody objects to the rule of let's fund, say, the top 50 sites only, until you meet the guy who runs the site at #51 - you'd better be ready for some stiff criticism at that point, because he's going to question your maths as well as your parentage. Let me go a bit further and look at 4 models in industry and government. The first is Amazon.com. Amazon sells everything you could ever want (ok, most things). It's easy to navigate, has a great search engine and lets you put everything you want to buy in a single basket. But not all the items that you buy are sold by Amazon - there are thousands of third party suppliers and, indeed, millions of customers who are selling previously bought goods (whether bought on Amazon or not). As a customer, I'm largely oblivious to this. There is plainly all kinds of integration and face-painting going on, but it doesn't matter to me. This seems to be the ultimate business model for aggregation of services under a single brand. Would anyone say that Amazon doesn't innovate? I doubt it - free shipping, new products, recommendations, gift wrapping services, buy new or used, 1 click purchasing (don't rag on about that patent stuff please). Then let's take itunes. Great interface, oodles of tunes of all genres, not so great at classical music sorting (something about composer/orchestra/conductor/soloist being hard to map I guess), ties in perfectly to the ipod (shuffle/nano/video etc), does podcasts, videos, even games. All the time you're oblivious to it as a customer. Except when you try and put your bought and paid for itunes music somewhere else - whether that's another player or even a Sonos. Never mind that I'm not trying to copy it, I just want to play it. This has turned out to be a genius business model for Apple but isn't that great for the consumer. Anyone who remembers how hard it used to be to switch from a Nokia phone to an Ericsson phone, if you didn't have all your phone numbers on the SIM card will recognise story. Eventually it breaks and all kinds of things emerge that let you get round it until the companies sort it out properly. Lastly, there's google. It creates lots of beta services, bundles them under its own domain name but makes them pretty hard to find unless you're dedicated (how many people even know google suggest exists, let alone orkut or google directory); Google buys new things that it can't build, like picasa or youtube and doesn't even bring those into its own brand structure initially. Eventually all of these things will be harmonised and just like the reason we fell in love with Microsoft Office was that we never had to wonder where the "open" button was or where the recent files are or whether spell check was consistent or whatever, it will all just work. And then there's UK government which did a sort of google thing early on, i.e. as you said, it let one hundred flowers bloom. If you'll forgive me a digression: Originally said by Chairman Mao Zedong"Let a hundred flowers bloom; let a hundred schools of thought contend", this slogan was used during the period of approximately six weeks in the summer of 1957 when the Chinese intelligentsia were invited to criticize the political system then obtaining in Communist China. It is sometimes suggested that the initiative was a deliberate attempt to flush out dissidents by encouraging them to show themselves as critical of the regime. Many of those who put forward views that were unwelcome to Mao were executed. So should we let them all bloom, or use a liberal dose of DDT, a la Mao? That isn't meant in poor taste. We need to encourage innovation - but we need to stifle the weeds. In the mess that is online government - and the UK is not alone in having this problem - we've grown a multi-headed monster that we can no longer control. Thousands of sites exist, new ones pop up most weeks with ever more bizarre names (I notice an increasing trend to use .com now to bypass government domain naming standards), they're rarely accessibly and probably rarely used. And, for the most part, we have no idea how much they cost - parliamentary answers provide the data that they have but most of the data is too difficult to get and, unless the question is very narrowly defined, the answers can be misleading and difficult to compare (I used to look at the data and try and draw comparisons; rarely did I get anywhere). Directing both new and existing sites towards direct.gov will help contain the problem - it won't solve it; no-one is going from 4,000+ sites to 1 or 2 anytime in the next 5 years. Figuring out the business case for which ones to let flourish independently will be difficult, but it is worth doing - perhaps using a venture capital style model (hence my interest in how much a site is worth, using the BSkyB link as a prompt) where money is provided for certain deliverables, perhaps market share, perhaps transaction count or whatever. At the same time, money is spent on direct.gov to bring it up to pace - we can agree or disagree about navigation and so on, but whatever the right things to spend on it, again subject to a good case (which need not be monetary), are spent on it so that it too is innovative. Believe me, when it was fired up, it was definitely innovative. All the work on accessibility, style sheets, single route to any device, search engines etc were all nowhere to be seen anywhere else (not claiming credit here, I was an interested watcher). Personally, I like the Amazon model. For the most part, I believe people want to come to one authoritative place - after all, it's unlikely that they "shop government", looking for differing views or better benefit assessments. The information there should be in a common voice with simple navigation that lets me get through the hierarchy quickly. The information that I have to provide, I want to give once and once only. I don't want to fill in 61 forms for government benefits - actually, I really don't want to fill in any forms, but that's a topic for another day. Certainly, the idea of a single central site would scare everyone else in the opposite direction at great speed. It would quickly be criticised as slow, unwieldy and lacking in innovation. And it would be all those - but, for the right aggregation of content and with the right adoption by the rest of government (the way the DVLA now routes you to it with every letter they send), it could be the new ebay, amazon or google. The place where everyone goes first - and where they branch away from only when it can't help, but using seeds placed in direct.gov to find where they need to go. So, I'm with you, almost. Kill the bottom 75% or relegate them to policy and position paper locations only; invest in the 5% of new business cases that you see every year, with a finite innovation budget; keep direct.gov in line with the best of the innovative ideas after they've been proven in the wild, again, on a finite budget.
Posted by Alan at Monday, December 18, 2006
Saturday, December 16, 2006
Just wanted to thank the North American company, who shall remain nameless at this point, who sent me a shiny new ipod Shuffle as part of their latest marketing campaign. They've thoughtfully pre-loaded it with some podcasts discussing innovation under the tagline "Innovative Ideas Can Come From Everywhere". I'm intrigued by the promotion and, of course, delighted to be sent a Shuffle (as a gadget fan it's important to have one of everything). Whilst I'm certainly completely in agreement with their chosen tagline (and it's not very different from the one at the very top of my blog right now), I was more than a little disappointed to have to pay customs duty on the shipment before I was allowed to take delivery - and, at that point, I didn't even know what was in the packet. £14.75 certainly isn't much to pay for a Shuffle but it's a funny way to get your customers endeared to your thinking. I've asked them to hold off before shipping me anything else - imagine the duty on a grand piano - as I think it would prove they had great wisdom.
Writing the business case for a government website is a difficult job. I suspect few governments have made much of a science of it, preferring to go one of two routes: (1) it's common sense (everyone else is doing is so we have to do it) or (2) it's part of the government's strategy (so we have to do it). Does BSkyB's purchase of 365 Media on Friday provide a new benchmark? Purely transactional sites have a pretty easy job: cost of doing the transaction online/cost of doing it online*total number of transactions forecast is about as far as you go. There are (at least) two things that complicate this calculation: a) What is the true cost, i.e. do you include capital costs (depreciated over what period) and, if you're carrying multiple transactions, how do you separate out operational costs? and b) How do you figure out what your transaction count is going to be? Early on, many (includng me) thought that e-goverment would be a slamdunk - people hate dealing with government anyway, so give them an easy, automated way of doing it and it would attract folks in droves. Poor implementations, negative publicity (repeat until false) and too many me-too sites quickly laid waste to that view. Only now are volumes reaching levels that give a foundation for a business model. Top down business cases failed dismally, as Kawsaki advises. Pure content sites, however, are more difficult. BSkyB paid somewhere between £94 million and £103 million pounds for 365 Media's portfolio of sites (such as Teamtalk.com, PlanetF1.com, Golf365.com and Cricket365.com) which, together, attract some 2.2 million unique visitors per month. These sites are all advertising supported. The buyout suggests that one monthly visitor is worth around £45-50 which seems like quite a lot of ad dollars. Direct.gov's October viewing figures (page 23 of the eDt monthly report) show 2.75 million visitors, up 68% on the same time last year (a pretty good growth rate). Only 320,000 of visitors use the search function, either proving the original design point that good navigation and great content brought together from all over government was the main factor or that, over time, people have learned that the search (which doesn't cover all of government but is restricted to the site itself with occasional offsite recommended links) isn't very useful. So, if direct.gov were valued on the same basis, would BSkyB pay £124 million for it? It seems unlikely given that there's no advertising-supported model (eDt's 2004 efforts to syndicate advertising involved placing ads on a network of a dozen or so government sites, hoping to refer business between them. The most popular ad led to the "search government" page). Would it be possible to put commercial ads on a government site and so drive revenue? Would it be possible if government outsourced its sites to commercial entities? It seems unlikely that ad dollars are going to be attracted to government sites (can you imagine, "dealing with a death in the family, phone Stanleys, leading undertakers in the UK"? or "need somewhere to spend your benefits? try 888.com"). So if you're going to invest a few million in a content aggregation site - a "super-site" - you have just a few numbers to play with: - X number of other websites will be switched off at an average cost of £Y, meaning there'll be an absolute save of £X*Y, less wind-down costs and any write-offs. - £Z pounds of capital investment in existing sites will be avoided, some of which will be redirected to the new site (and then those sites will be switched off). Such capital spend might be required for routine upgrades, provision of accessible services (although this doesn't seem high up anyone's list with over 60% of sites not meeting the standards) Without hard numbers and cast iron commitments backing up these assumptions, the business case is unlikely to have much credibility. The business case for direct.gov included a line that noted that other government websites would wither and eventually die, because all the visitors would first go to direct.gov. It's hard to test that now, 2 and a 1/2 years after launch, but I suspect that it hasn't proved as true as many would have liked. One thing I am sure of though is that the myriad government websites that exist today lack a collective business case, so I am heartened to see the Varney report reiterate the need to consolidate down to two (one for individuals and one for businesses, leaving the self-employed with a slightly split web personality), even if there's nothing new in that. Disappointingly, Sir David didn't have a go at the business case for this recommendation. That would have shown some interesting numbers but instead we're still working on the "leap of faith" business model.
Sunday, December 10, 2006
Rejected. No London Marathon 2007 for me. A magazine says "sorry" and I get a zip-up top with "London Marathon" on the outside and a big "2007" label on the inside that no one but me need ever see. I've decided not to run any more races for charity so will only run when I'm granted a place through the ballot. The spirit of generosity towards Macmillan Cancer Support has been enormous over the last 18 months but I think I've pushed it enough. I'll carry on donating personal money to Macmillan as I have done for the last 10 years or so. This winter will be about running for fun rather than for a race although I'll slot in a couple of half marathons in March or April. I now have a half dozen marathons under my belt in four cities - London, Paris, Vienna and New York. I've learned a few lessons from those and I'm logging them here so that I'll remember them next time. Maybe they'll help some others too. 1. Don't take the training plans published in books and magazines as gospel. Most of the plans will have you running 40 miles+ a week at peak. Art Lieberman's plan has you doing 36 miles a week for 12 weeks, with a 16 mile run every Sunday, and this for those looking for a 5 hour finish. Hal Higdon's Intermediate plan sets you off at 24 miles in week 1 and has you at 43 miles 15 weeks later. Now these guys have, of course, trained hundreds of people and run dozens (hundreds in Hal's case) of marathons themselves. They know stuff that I don't know. But I do know that I've run sub-4 hours on a total of 231 miles training (70% of which was in the last 2 months) and I've run sub-4 on 453 miles (those are figures for the 6 months before a marathon). You've got to find something that works for you and that fits in with what else you've got going on, recognising that some other things are going to have to take a backseat if you want to finish safely and not spend the next few days walking around like a 90 year old. If you can stick to a plan, go for it, but know when to back off and don't feel bad about it (except for Rule 5). 2. Don't believe the pace tables in the books. I can run a 10k in about 47 mins. The books tell me that should put me on track for a 3h 20m marathon. Likewise, I can crack a half marathon in 1h 41m, which marks up to 3h 23m. Yet the fastest I've done is 3h 45m. There are perhaps a couple of reasons for this (1) I don't follow the training plans (see above) and do enough long runs that would improve my stamina or (2) I'm not a natural long distance runner (no surprise there). But few people who are running marathons are naturals and fewer still have the time to do the long runs necessary. So plan for a slower second half than the books say and set your target time based on that. If you're running a half in 1h 41, you're going to come in between 3h 40m and 3h 50m for a full, if you're like me that is. Plan also to run your first half a good 8-10 minutes slower than you can run a half marathon. So a 1h 41m half becomes a 1h 50m first half followed by a 2 hour second half. 3. Do wear a pace strap, but use it wisely. Pace straps assume you'll run your race at an even pace. You might in which case, I think you're rare. I've so far not managed to run an even pace, let alone get close to the famed "negative splits" (where you run the second half faster than the first). In NY the second half was 8 mins slower than the first, in London earlier in the year it was 20 mins slower. A smart pace strap will force you to run the first couple of miles more slowly than plan - but the ones I've seen don't think that way. I picked up a Nike strap in NY which had the first mile (on a target of 3h 50m) at 9:39, before moving to 8:14 in mile 2 and then drifting back to 8:45. Times varied, per mile, by plus or minues 5 seconds from that pace until the end. Perhaps this argues for a personalised pace strap (conveniently, unlike the Lucozade ones you see at UK races, the Nike one is a velcro strap that lets you replace the pace chart inside a plastic strip)? If I was starting at the front, mine would have the first three miles pretty fast and then slow me down; it would recognise that the second half will be perhaps 7-12 mins slower than the first and it would also know that I'd speed up in the last 2 miles. You have to know how you run to build a pace strap like this, but I think I'm there now. The strap would also have a big font. 4. Run in the right units. Marathons in the UK and North America run miles, in Europe they run km. I do all my training runs and pace analysis in km which means that in every UK/NA race I have to do the maths in my head for how I'm doing. That's fine at the "round" numbers, 5, 10, 20 etc but you try figuring out 17 miles in km when you're just coming to the wall and you're trying to figure out how you're doing against your plan. 5. The bulk of the preparation is in your head. When you're out running and it's cold or wet or dark or all three, you want to quit. You want to take a shortcut. That slightly nagging pain in your knee or your foot soon becomes a focus and you think you'd better cut it short in case it develops. The moment you do that is the moment you're selling yourself short. You will get many of those moments during the course of 26 miles on race day and conquering them during training is the best way of handling them so that they won't matter when it really matters. 6. Don't run a race in November. If you're like me and you shift to a healthy diet for the 6 weeks or so before a race - lots of pasta, no drinking - you'll really suffer in December. All of the people you didn't see in October and early November will book to see you in late November and early December. The Christmas season then comes and you'll be booked every night. At the best of times the siren call of Great Bordeaux is hard to refuse and, after weeks without, your shields are down, phasers are set to stun and you'll know about it every morning for a month. I imagine running a race in January would be even worse. 7. Set yourself a target then leave it behind. Despite my very best efforts, once I've started a marathon, the pace is the pace. In NY I tracked the 3h 50m pace guy all the way to the half way mark and then gradually fell behind - he was running a steady pace, I was slowing down in the second half. Eight minutes behind over the second 13 miles may not sound much, but it's an enormous amount to make up when you're already tired. Likewise, if I'd been heading to finish NY just outside of 4 hours instead of just inside, I doubt that there would have been much I could have done about it after mile 20. So, push yourself by all means, but don't burn up. Faster runners than me tell me that they've been physically sick at the end of every marathon. That's never happened to me and maybe it's because I haven't pushed myself to the very end of my limits, or maybe it's because that doesn't sound like a fun way to finish a race. 8. If you're running NY or Paris, practice drinking out of cups. London's drink stations are the best I've seen - bottles of water with the tops removed and squeezy lucozade cartons, also with the tops off. NY and Paris (I can't remember Vienna) are not quite to easy - cups filled with liquid. I defy you to drink anything like enough without slowing to walking pace, and a slow walk at that. Paula Radcliffe has written that you can practice it. Well, in 26 miles I couldn't figure it out, so take her advice and practice before you go, if only to stop yourself from being covered in sticky juice. NY did have one thing over London, PowerBar Gel at mile 18/19 - and not in just the horrible flavours either (although you wouldn't want to touch the stuff if you weren't running 26 miles). 9. It will hurt. Running 26 miles will hurt, hurt and hurt some more. Get used to it. Be consoled during the race by the fact that the pain during the race is unlikely to be as bad as when you try and walk down some stairs the following day. I haven't yet found the amount of training that stops that pain, although I have reduced it from 5 days of feeling 90 years old to just one.
Posted by Alan at Sunday, December 10, 2006
Saturday, December 09, 2006
With a quick trip to Chamonix only a couple of weeks away I've been studying the resort snow reports looking for signs of hope. Over the next 7 days the forecast say there could be as little as 21cm of snow. Even the local reporters are sounding somewhat downcast.
Dubai may now be the destination ski resort, with the best snow east of Aspen. Their SkiDubai resort may be the only place you'll be able to ski inside a decade. Imagine the queues as all of Europe's ski fans descend on Dubai, mixing a few days in the snow with a sun tan, a splash at Wild Wadi and the purchase of a few luxury watches. Even if the dollar falls to three to the pound, Dubai will still look attractive with its currency pegged to the dollar. There won't be enough hotel rooms or airline seats, despite all the expansion, to handle the influx of tourists. Better build some more palm islands.
Apparently in 2008, they'll be opening a new resort so as to cement their lock on the European ski market:
A second Dubai indoor ski slope is the Dubai Sunny Mountain Ski Dome in Dubailand which is due to open in 2008. It will provide a great attraction, featuring a revolving ski slope, an artificial mountain range, an ice bridge, a cable lift, a snow maze, an ice slide, polar bears cold water-aquaria and special sound and light effects.
I'm surprised they felt the need for a revolving ski slope - I've found that most slopes do that as I tumble down them, head over heels.
Jokes about the Dubai Sheikhs being responsible for the very global warming phenomonen that will make their resort the only place in the world that you can ski will not be welcome.
Posted by Alan at Saturday, December 09, 2006
Tuesday, December 05, 2006
What seems like years ago - in fact it was 2 years ago - I heard about a service called SpinVox that promised to translate voicemails into text messages. The idea of never having to listen to a voicemail again fired me up considerably. I hate the things, especially the long, rambling ones that people leave as they're driving somewhere; the poor souls have nothing to do but catch up on their calls so you end up with a baffling, unstructured message that seems (and often is) never-ending. Well, no more. SpinVox is available to all those who want it, right now. You can sign up for a free trial - one week or 50 messages - at their site. After that you'll pay upwards from £5 to have all of your messages sent to you either as text (hmmm, get enough of those already) or email (better still, you can forward the message back to the person who left it for you and give them the answer they need in the context of the message). Tests over the last week show that SpinVox does pretty well, even with difficult voices. It stumbles over surnames and seems very confused by the word "hospital" (don't ask), but I'd say it's 95% accurate, even with Wayne, my Brummie friend. FOr instance, here's the kind of mail you get (I've changed the names to protect the guilty): Alan, hi it's Joe. Alan just wanted to see if ya had availability on Fri, for us to get a few folks together to talk with Bob about what needs doing on the commercials & how that fits into the wider architecture with you, also want to pick your brains about how you wanted ___ to play into that. Might be worth talking about the ___ sorting ___ & focus on the required outputs You can see it's not quite perfect - and it's mostly smart enough to know when it doesn't get a word at all. Neatly, if you want to listen to the vm, it gives you a short code to hop straight to this specific message to save you having to sort through old ones. This is just such a good thing. I used to dread the end of the day, dialling up my voicemail and hearing that dreaded dis-embodied voice intone "you have more than 20 messages". Now I get them as they come, complete with the number they originate from (no more scribbling down numbers when I'm between places) and I can choose to respond by phone, email or text. You can bet that it will mostly be email or text though, so get used to not hearing my voice. Genius.
Posted by Alan at Tuesday, December 05, 2006
Saturday, November 25, 2006
And, sadly, this isn't a good top ten to be in. November's PC Pro magazine claims that the PM's announcement in 1999 that "all government services will be online by 2005" was one of the worst IT predictions ever. Being a historian of this kind of thing, I know that in November 1999, the PM actually announced that things would all be online by the end of 2008 but that he revised that statement in March 2000, bringing it forward 3 years. I believe it may have had something to do with the French or the Canadians announcing an end of 2005 target. The folks at PC Pro believe it sounded like "groundless bluster" then and now. Their primary source for this is the "Better Connected 2006" report where just 13% of local government sites are cited as being transactional (where that is defined as "adding more than one type of online interation"). They also note that even the Cabinet Office has reported that 97% of official sites are unusable by disabled people. So, after £10 billion (their figure), apparently the government admits that only 12% of adults have visited a local authority website (as opposed to zero % children I wonder?). Perhaps the PM should have stayed with the first prediction. He'd have a good chance of being right if (and ony if) there were now a consolidated and well-led effort to kill off poorly visited sites, aggregate citizen-focused content in key sites (and those stand out - the Revenue, DVLA, direct.gov, department of health etc), drive the online identity debate with the government gateway at the centre (not ID cards, but then I'm biased) and leverage off third party platforms and operations, e.g the citizen's advice bureau, accountants and payroll providers, the post office etc. No simple task but I'm absolutely sure there is more appetite for this now than there ever was before.
Posted by Alan at Saturday, November 25, 2006
Wednesday, November 22, 2006
The big story this morning, at least in this part of the world (NW England), was that the local police are being issued with fingerprint checking devices. Reporters and civil libertarians alike reported breathlessly about the police's new ability to carry out roadside checks to check identities. Some seemingly conflicting assertions and figures were bandied around on TV and on the web: - 60% of people stopped by the police give false identity information - The automatic number plate recognition (ANPR) teams, who already cross check vehicle number plates against databases, will be able to verify a person's identity within five minutes without having to take them to a police station - More than 6.5 million fingerprints are cross-referenced - The device has an accuracy of 94-95% and will be used for identification purposes only, say police, and there are electronic safeguards to prevent misuse - The encounter can be 15 minutes on the roadside rather than three hours in the police station - Fingerprints will only be stored onto a mobile device until it is full, which is at around 100 fingerprints. - Inspector Steve Rawlings said it takes two sets of fingerprints and the fingerprints are not retained - The device could save police forces £2.2m a year in resources wasted pursuing false identities So let's get this right: The police stop someone driving a car, perhaps acting suspiciously. They run the number plate and find out who the person driving it should be, based on insurance or registration information (or both). If the car doesn't have insurance, they arrest them and take them to the police station for driving without cover. They can find out who they are later I imagine. They ask the person who they are; if the details provided don't match the owner listed on their system, or the person isn't able to name that person and explain why they're driving the car, their fingerprints are taken. The fingerprints are checked against a database that contains about 10% of the UK population's records. There are something like 38 million drivers so we could be generous and say that if everyone who has committed an offence is also a driver, then something like 15% of the population is in the fingerprint database. If they're not in the fingerprint database, we have to trust they are who they say they are and either take them to the police station to find out more or let them go. But 60% of people who are stopped apparently provide false information. That's a stunning figure. The assumption must be, I suppose, that 6 out 10 are guilty of something and so tell lies about who they are, risking the police taking further steps. The next assumption must be that if you're prepared to tell a lie to the police, you're more than likely to have committed an offence for which you were fingerprinted and so you have your record on file and it will show up. The missing figure is perhaps how many people are stopped per year so that I could compare the 60% false identity giving with the 6.5 million records in the national fingerprint database and the £2.2 million cost of followup. Further, rather than delete records right away, the police want to hang on to "100" (or until the machine is full - must be ZX81 style memory), hoping that they'll catch you again and fingerprint you with the same machine. I can see these statistics being trotted out again, pretty soon I suspect, to justify ID cards and fingerprints being stored upon verification for later checking. All for £2.2 million in follow up costs by the police forces? I think I'm more interested in what the 60% are up to giving false identity data and whether more bad folks will be put away.
Posted by Alan at Wednesday, November 22, 2006
Sunday, November 12, 2006
The NY marathon is safely behind me and normal life has resumed for a little while at least. I finished in a little under 4 hours, slower than I'd hoped for but with the way my training had gone in the month before the race, better than I expected. The logistics were masterfully organised. Over 30,000 people are bussed to the start between about 5am and 7am - the start (see the picture on the left) at the Verrazzano bridge is over 16 miles from Manhattan and there's no easy way to get between the two unless you're into double marathons. But the consequence of this early start and a kickoff time of 10.10am is that a good 3 hours is spent hanging around. The weather, at least, played its part - the skies were clear, the sun shone and the wind took took the day off. The race itself is tough. The start is slow and it felt like I was running at half pace as I dodged in and out of runners all around me, many of whom were busy kicking off sweat pants or jerseys as they ran, leaving piles of clothing on the floor for the unwary to get tangled in. I was on the top of the bridge so avoided the oft-told (apocryphal?) story of golden rain from above. It took a while to get into stride and my pace, even then, was a lot slower than London through the first half, but then it didn't slow as much in the second half. I ran the first 21k in about 1h 55m, the second in about 2h 3m. The main difference between London and NY was my pace at the start. I was on the white start line for London and was able to hare off at a speed that I probably shouldn't have done, spurred on by the 10,000 or so people who overtook me in the first 20 mins. In NY I started in 16,646th place (assuming that the numbers on the front correspond to starting position) and finished 11,786th. With the two graphs side by side like this, it's really quite startling how much I slowed in the second half of London. The average pace times aren't that different - around 5:20/km in London and 5:31/km in NY, but the volatility is very different. Justgiving tell me that I've raised over £12,000 online in the last 18 months and I know there's a further £3,000 or so from offline donations. £15,000 to a great cause for running a couple of great courses. Seems a fair deal to me. Best moment of NY? Coming off the dreaded Queensboro Bridge - a mile and half of uphill slog, mostly in the dark, eyes fixed on the guy in front looking out for any change in pace to avoid smashing into him or tripping over his legs, hoping that whoever was behind was doing the same - and seeing Macmillan supporters, friends and family lining the street cheering and waving, before making the turn onto 1st Avenue where the crowd was easily 10 deep on both sides of the street for several miles. If you see my time for km 27 on the graph, you'll see how much faster good support can make you run. The support throughout the whole race was awesome - they say 2 million people turn out to watch it. Worst moment? Coming down any flight of stairs the day after the race. Oh, and that sodding hill running alongside Central Park from about mile 22 to mile 25. A big thank you to all those who supported me through the training, made generous donations to Macmillan and who encouraged me on race day, whether I knew you or not. Your turn next year folks?
Posted by Alan at Sunday, November 12, 2006
Wednesday, November 01, 2006
I had dinner on Monday night with a newly minted senior chap at a major US-based technology company that you would instantly know. Sadly, it being the run up to the New York Marathon 'n' all, I wasn't drinking. The name of the wine chosen for the rest of the table was, however, particularly apt.
Posted by Alan at Wednesday, November 01, 2006
Sunday, October 22, 2006
The New York Marathon is in just 14 days. Today's run, the last long one at 22k before the big day, had me splashing through puddles for the bulk of it. Passers-by shake their heads in dismay wondering what kind of fruitcake would be out in a pair of shorts and a t-shirt in that kind of weather - torrential rain. Other runners, it seems, recognise a kindred spirit (fruitcake) and nod or say hello more frequently than usual. With my training all but done, it's either in me or it isn't. This time round, the second marathon this year, I've certainly down the distance, 651km since May 1st (versus 655 in the run up to London over the same period), but my pace is slower - my fastest half marathon was in Nottingham at just under 1h 47m, against 1h 41m (from memory) in Reading in March. Nottingham was definitely a tougher course, but I can't use that as an excuse as my average pace has been down on every run except the Nike 10k a couple of weeks ago when I ran my fastest 10k in about 8 years. Here's the monthly distance chart for the last 12 months And the monthly average pace graph for the same period Comparisons are a bit difficult - London came near the end of April, New York is at the start of November. But I ought to be hitting 5:45/km or faster this month and, apart from the Nike 10k, I'm at 6:00/km. Still I can live with that difference, this one isn't so much about the time or the finish quartile, it's about the fun and experience of running the streets of New York with no cars and with millions of people out to watch. I'm looking forward to it. If you haven't yet thought about sponsoring me, all contributions will be gratefully received by Macmillan Cancer Support at my JustGiving page.
Posted by Alan at Sunday, October 22, 2006
Saturday, October 21, 2006
Is this true or just another gullible story? The Software Publishing Association estimates that on average the chance of a software patch causing crashes or serious compatibility problems is seven times greater than the likelihood of running into the bug or security issue which the patch was intended to fix. If it's not true, it does what all the best gullible quotes do, makes you stop in your tracks and consider it possible. And, like the very best, if it isn't true, you want it to be. Which may go to show that I have no idea whether gullible is truly for the gullible or absolutely true stuff. I can live with either.
I met Richard only once, just a few months ago. He was wildly articulate and passionate about first Egg and then Zopa, his most recent creation. He spoke about the focus group sessions where the name "Egg" had been tested and where, consistently, it came out top with potential future customers, despite no one inside Prudential believing in it. I'd played around with Zopa, fascinated by the model and the managed exposure it seemed to offer lenders. His thinking was inspirational. Today I learned that cancer, the scourge of perhaps 1 in 3 of us, has taken him at the incredibly young age of 44. Cancer is a cruel, hardened, street fighter. The news stories say Richard's battle was brief and I hope, therefore, relatively painless. I have lost family and friends to the dreaded cancer, it's never easy. My thoughts are with his family and all those who knew him.
Saturday, October 14, 2006
As if we needed further evidence that we shouldn't believe what we read in the press, except when the Daily Express does yet another front page of the Diana murder conspiracy, this week's fun suggestion was that HMRC would soon be checking eager employees' use of Blackberries to make sure that they weren't be used for personal email (God forbid - I do enough on them already). 'Twas The Sun that first aired this, suggesting that tax inspectors would perhaps confiscate your gadget and count the emails, levying tax accordingly. By today, the story had crossed to several papers including, if I'm not mistaken, the Independent and The Standard. HMRC, responding quicker than a mail flying through the ether from a blackberry, say "Garbage!" according to several sources, not least The Register.
You've got to read the odd spam email, every so often, just to see what's the latest thing. This one, picked entirely at random, had me wondering. It's sent as an image embedded in the mail rather than as text - I don't think I've seen that before. Plainly that seems to stop the spam filter in its tracks. Outlook is pretty good - I lose about 95% of incoming spam between runbox (at the server end) and outlook (client end). Anyway, here are some excerpts (I only wish I could add in all the colours used throughout the message): Get ARSS first thing Monday If you've been alive over the last few years you know that any stock can move given the right circumstances ...goes BOLISTIC! This is your chance to get in before it BLOWS Who even knew there was a stock with the ticker ARSS. I had to check, just to see if it was real (and also to make sure I was nowhere near when ARSS decided to BLOW). Needless to say, the chart doesn't make for pretty reading. It goes on, in plain text: The project work involved the expansion of the apron, the construction of an asphalt shoulder, associated drainage and electrical work and the replacement of airfield signage. That should mean something. php - FOR PERSONAL AND NON COMMERCIAL USE ONLY! He was joined by local students and scientists at the unveiling of the seismograph showcase at the Geological Survey of Canada. Monitor displays speed, distance, calories, and time to keep track of progress. There are still places in the Tampa Bay area that. Implant features circulo-trapezoidal neck design that allows for wide range of motion. Conformable, lightweight tape has passed primary skin irritation, cytotoxicity, and skin sensitization product safety testing and does not contain latex proteins. well enough to become a criminal. And then ends, fittingly, with: Is this a good idea or throwing good money after bad? how pathetically cute. com - all the latest news from EU. how pathetically cute. You can preserve your capital and jump back in when stocks begin moving up. I wonder if they spotted "throwing good money after bad" before they pasted in the random text?
Wednesday, October 11, 2006
The Bush Administration sounds fed up with criticism of its e-government initiatives - Congress has blocked funds proposed for several projects in the past. They've published a rebuttal to some FAQs, including this one, under the catchy title of results.gov: MYTH: E-Gov initiatives haven’t improved services to citizens. FACT: E-Gov initiatives have made it easier for Americans to access government services and find information about federal programs. For example, GovBenefits.gov provides citizens a single point of access to government benefits and services. To date, the initiative has provided information to more than 22 million citizens and generated over 5 million referrals to agency benefit programs. Prior to the launch of GovBenfits.gov, citizens had to search multiple agency websites in efforts to locate federal benefits. E-Gov initiatives have also made it easier for citizens and businesses to pay their taxes online. With IRS Free File, eligible taxpayers can prepare and electronically file their tax returns over the Internet using commercial software for free. IRS has received more than 15 million tax returns through the initiative resulting in processing cost-savings of more than $32 million. The Expanding Electronic Tax Products for Businesses (EETPB) initiative aims to reduce the tax-reporting burden on businesses while improving the efficiency and effectiveness of government operations. Since the inception of EETPB (and through July 2006), employers have filed over 5.2 million SS-4 forms online to obtain Employer Identification Numbers (EIN) electronically. Furthermore, businesses can access a range of tax forms online (e.g., 94x, 1120, 990), resulting in reduced burden and increased government efficiency. Grants.gov is the Federal Government’s single website for over 1,000 grant programs that award more than $400 billion annually. In FY 2006, 75 percent of all Federal discretionary grant opportunities (2,011 out of 2,678) were available for electronic application submission via Grants.gov, and grantees submitted more than 80,000 applications through Grants.gov. This is a 31% increase over the number of opportunities that were available for electronic application submission in FY 2005 (44%). Whilst I was looking at this factsheet, I was pretty surprised to come across another site, ExpectMore.gov. It's a reasonable facsimile of the UK's OGC reviews except for the fact that you can see a list of programmes that are peforming and those that are not. Here are the findings from one, General Services Administration - National Information Technology Solutions: Results Not Demonstrated A rating of Results Not Demonstrated (RND) indicates that a program has not been able to develop acceptable performance goals or collect data to determine whether it is performing. * The assessment found that the program is useful to federal agencies that do not have in-house expertise to acquire information technology (IT) products or services. However, the program must better demonstrate the value it provides to customer agencies. * The program must develop long-term outcome goals and efficiency measures which are comparable to other federal agencies or the private sector. While the program does have annual goals, it must develop annual goals which measure the savings and quality improvement that agencies achieve through use of this program. And, for comparison, here's a programme rated as "not effective" as opposed to "results not demonstrated" Ineffective Programs receiving this rating are not using your tax dollars effectively. Ineffective programs have been unable to achieve results due to a lack of clarity regarding the program's purpose or goals, poor management, or some other significant weakness. * Performance goals have been created, but progress has not been adequately assessed. The program measures performance by estimating additional natural gas reserves attributable to technologies developed by the program. The program is working to more clearly demonstrate the validity of its performance assessments, including through regular use of periodic, independent program evaluations. * Actual additional natural gas reserves attributable to technology developed by the program have been relatively small. Moreover as noted by the National Academy of Sciences: "It is difficult to separate the contributions made by the Department of Energy and contributions made by industry and others." * This program and other applied research and development programs at the Department of Energy are working to improve their estimates of potential program benefits and use that information to guide budget decisions. The program has made progress, but estimates are not yet comparable between programs because of different modeling methods and assumptions. Who would have thought that the US government would be more open on this kind of thing than the UK government? There are 5 pages of "programmes not performing" and 12 pages of "programmes that are performing" - which would make a mockery of the Standish Group's Chaos Chronicles if it's true and might give us something to think about regarding public sector project success. What I think is also fascinating is that you get to see programmes that you wouldn't even think to look for - Missile Defence (The Missile Defense Agency (MDA) mission is to defend the U.S., deployed forces and allies from ballistic missile attack. MDA is researching, developing and fielding a global, integrated and multi-layered Ballistic Missile Defense System (BMDS), comprising multiple sensors, interceptors and battle management capabilities), or Advanced Computing Simulation (The program provides leading edge, high-end computer simulation capabilities to meet nuclear weapons assessment and certification requirements. The program uses code computer models and experimental data to predict the effect that aging and other changes will have on the U.S. nuclear warhead stockpile.).
For the last couple of months I've been playing around with my latest ipod and the nike+ gadget - it sticks neatly between my laces and seems to stay put. I haven't really taken it too seriously: Mostly I listen to audibooks and they seem to be recorded at a lower volume than music (or maybe the nature of the content means that it just seems that way) so I have to listen on loud, the Nike+ voice owner, pleasant though she is, therefore shouts in my ear every mile or every time I press the centre button for an update. Usually, it freaks me out because I'm really not expecting to hear anything interrupt the book and I'm often lost in my own thoughts not really counting the miles. Nike may have just changed that and made me want to take this thing more seriously. Today they sent me a mail inviting me to sign up to a group challenge. This is the first one: RUN WITH LANCE 5,394 miles donated to date Lance Armstrong is training for the marathon in NYC, and he's raising money for the Lance Armstrong Foundation as he goes. Donate your miles to reach the goal by 11/5. Lance is running the NY Marathon - November 5th - and is expected to do a near world-record time. So over the next 3 weeks I'm going to run for his cause and make sure I wear my ipod and nike+ for every run and I'll just have to push myself a little harder to make it come good. That said, I haven't figured out how to sign up for the challenge so if you've figured it out already, I'd really like to know. The email links to the "sign up now" part of the site, but not to the "already a member, press this button to donate". There's no sign of a donate button on the site and nothing about "group challenges" which is how Nike refer to it. Maybe a screw up on Nike's part here? I think this is just a fantastic idea. Kudos to Nike for pursuing this in the week that the Independent's front page carried a story noting how the Americans are eating themselves through the world's resources and figures that show that 30-40% of Americans are obese. One way to make people exercise more is to make it worth their while; a far better way is to let them make it worth far more peoples' while.
Saturday, September 23, 2006
I've had to turn on comment "word verification". If you leave a comment you'll be asked to enter a word that appears on screen as an image. I've resorted to this only because someone is leaving multiple triplets of comments that link to whoknowswhere (I daren't look). Don't let it put you off.
I lived in Paris for about 4 years, from 1997 to 2001, with a few gaps here and there. Driving there was always an experience - the first time I ventured on to the roundabout at Place D'Etoile forever sticks in my memory; it's one of the roundabouts where traffic coming on has right of way. This video, all 8 mins and 39 seconds of it, puts my driving experience in Paris to shame. The sound track is that of a Ferrari or maybe a Lambourghini; the visuals are a bumper level view of Paris - the Arc de Triomphe, the Louvre, Sacre Coeur, a few pidgeons scattering, one or too shocked pedestrians and a couple of near misses. My favourite moment? The driver turns left into the Louvre (the pyramid on the right) and the engine noise reverberates off the archway. That and shooting more than a dozen sets of red lights - the only time I did that in Paris was when I ran the marathon and I wasn't going quite so fast. Watch it with the sound up loud. All tours of Paris should be this much fun.
Sunday, September 17, 2006
The Government Gateway won an award this week ("another one?" I hear you cry). It's an IDDY award, or perhaps an IDDYIOT award. Apparently it's only for those deployments of Liberty technology. Here's what was done to win it: - Deployment -- The Government Gateway Authentication Service has been designed as the authentication server for all e-government services in the UK. Nearly eight million citizens in the UK are registered to use the gateway service. - Circle of Trust -- The Gateway provides authentication services on behalf of multiple other public-sector bodies, based on trust principles established in UK e-government legislation. The Gateway also supports a "tiered" authentication scheme according to the level of assurance provided by the user enrolment process and the type of credentials issued. - User-Centric Capabilities -- The project has been developed to provide citizens and businesses with ease-of-use capabilities for accessing a variety of UK government services; not only does the Gateway provide a single authentication and entry-point for online government services, it now supports the predominant open standards on the market, making it easier for public sector bodies to integrate its authentication capability with their own service provision systems. - Highlights -- Deployment supports all federation standards to allow for complete interoperability between government agencies nationwide, there is less need for each local authority to develop or implement its own secure authentication mechanism. The Gateway provides local authorities with a single, consistent and robust security mechanism at minimal cost and effort on their part. - Interoperable Federation Technologies -- A principal aim of this project was to reduce the cost and complexity experienced by government departments and other public sector bodies (such as Local Authorities) in making use of the centralized authentication service. To that end, the Gateway was enhanced to support both WS-Federation and the Liberty Alliance Identity Federation Framework standards. This delivers a level of interoperability and protocol-independence which greatly simplifies the task of integrating service-provision systems with the Gateway's authentication functions. It also means the Gateway can deliver consistent authentication to its users without requiring them all to adopt a single standard, which could potentially alienate a substantial segment of the user-base. All those apart from the last bullet have been true for a while. I didn't know the Gateway did (used? incorporated?) Liberty so I asked the guys back at the Cabinet Office what was with that. Jim (It's life but not as we know it), replies: This year we built a single sign on portal as part of the Gateway UI. The business objective was to deliver a white labeled common authentication page that would manage the authentication calls with the Gateway. In order to do this we had to implement single sign on to mange the user’s authenticated session between the Gateway domain and the participating portal domain. We did this by implementing an interoperable SSO protocol handler that allows the portal to select whether they want to use one of the Liberty, WS-Federation or SAML protocols. The security token that they receive is a SAML 1.1 token but each one can be customised per portal. So that's all clear then. Congratulations. Nearly 6 years on, it's great to see the Gateway still be recognised as leading the way.
I must be 5 years behind everyone else on this but I discovered Goggles today, the Google Earth flightsim. Setting it to start in London, I was delighted to recognise Hyde Park "beneath" me, and then to fly down Knightsbridge and to my own house. Wonderful stuff. Next up please an online version where you can shoot your friends. What's a bit of flying where you can't shoot other planes?
Monty Hall was the host (and later the producer) of a TV quiz show called "Let's make a deal." The show started in the early 1960s and finished as late as 1990. The show, and the host particularly, gave rise to a paradox known as the Monty Hall Problem. The essence of the problem is given with this paragraph: http://www.blogger.com/img/gl.link.gif Suppose you're on a game show, and you're given the choice of three doors: Behind one door is a car; behind the others, goats. You pick a door, say No. 1, and the host, who knows what's behind the doors, opens another door, say No. 3, which has a goat. He then says to you, "Do you want to pick door No. 2?" Is it to your advantage to switch your choice? Dan has written about this problem before and that, coupled with a recent comment from Ian where he said, "I'd rather pay 1000 quid for something that I liked than 500 for something I didn't" got me thinking. Ian went on to propose that government needed some kind of proof of concept lab, not run by the big vendors, to make sure everything would work out before too much money was spent. It turns out that in the Monty Hall problem, it's better for you to switch doors than to stay with your original choice. You won't believe me, so check the links, or play for yourself at this site, where you'll be presented with the stats from past winners and losers. So let's play out this problem in the government space. You're starting a project and there are three outcomes - two are the same, you spend a fortune and it doesn't work out (back to the two standard problems of government IT projects, (1) it costs too much and doesn't work and (2) it's all big brother); the third door is the route to success. You don't get to find out which outcome is going to be your gift upon completion right away. But let's suppose half way through the project, Monty pops up and offers you the choice of doors. Behind one is successful completion of the project for only £1000, behind the other two is failure having spent £500 (using Ian's numbers - and I know he said "something I liked", but bear with me as I slightly change that to "something that works"). You pick one door and Monty opens one of the others. It shows the project failing at a cost of £500. Do you switch doors, i.e. strategies, or do you carry on? If the odds were the same as the Monty Hall problem, you'd make a switch, because the odds of success were 66%, for the investment of only another £500. The key in all this is that Monty knows what is behind the doors. What if the numbers were bigger? What if it was £5,000 more? Or £50,000? Or £500,000? or £500,000,000? Would you still change? If you'd already spent £500,000,000 and thought that you might pull it off for another £500,000,000 (or that there was a 2/3rd probability you would), would you still go for it? In the world of government IT, the numbers are far more likely to be in the tens or hundreds of millions (and, sadly, occasionally in the billions) and, worse, there aren't just 3 doors and, still worse, no one helps out by opening a few of the doors to narrow your choices - and, we don't know what's behind the doors (other than some inevitable press headlines either way perhaps). So, back to Ian's proposal. Would a proof of concept lab materially improve the odds of delivery? Back in the original procurement for the NHS IT programme (known now as NPfIT or Connecting for Health), each of the vendors was required to construct complex proofs of concept of all of their technologies, and integrate them to other vendor results (mostly the spine) to show that their proposal would work. It was during this stage that many pulled out. Before the Tax Credits programme, several vendors were asked to come in and carry out a proof of concept, over 8 weeks if memory serves, using Java, .net and coolgen so that a decision could be made on which technology to use. Intellect, the supplier organisation, did a deal with OGC to offer up a service called "Concept Viability" that would let government test out ideas, in a multi-vendor environment (although not necessarily building them), before making a commitment to any one route. I appreciate that vendors have been involved in these and what Ian is suggesting is some kind of in-house lab, but the idea has been well-trialled. And, I assume, the tests carried out gave some degree of reassurance that the idea was sound. My experience with government IT leads me to believe that it is not the fundamentals of the technology that fall apart in most projects - the servers talk to each other, the applications do what they're supposed to do, the network connects to offices and so on. No, there are some other problems: - Often the relationship between the requirement and the solution is not understood. A proof of concept can operate with only part of the requirement clear. As the requirements evolve and time passes, decisions made early on are found to hamper later delivery - complex set ups need to be unwound and rebuilt. - The subtleties of legislation are not usually clear at the outset, the type of client base that will be supported is not known (tax credits for instance). This leads to assumptions being made that, when tested much later in the process, often turn out to be flawed. - Proofs of concept (or, perhaps, proof of concepts) rarely have to scale and operate nationally. A system designed to work in a lab doesn't have to deal with dozens or hundreds of offices scattered around the nation, it doesn't have to deal with millions of annual claims each year. Simple decisions taken early on for expedience play out poorly when scaled up massively. A good design up front coupled with some really good scenario thinking can resolve these points. A system can be designed and built that is flexible enough to allow late stage changes, is scalable enough to deal with national usage and capable enough to deal with the quirks of changing legislation. I'm not sure, though, why government would be any better at doing this than the vendors it employs - government long ago gave away much of its IT expertise so that it could be reinvoiced by the big players. What route in is there for new talent - and how would government compete for that talent with people who really know how to design such systems, perhaps the designers of eBay, Amazon, Google or whoever? There is, though, a need to do more up front to reduce the later risk; but there is also a need to get the later scenarios right and to test those both technically and contractually. As the range of possible scenarios goes up, the risk premium the deal attracts will, of course, go up and, in turn, the cost to government (the insurance risk) will go up. Some stress testing of contracts would make a lot of sense, notwithstanding the fact that there are is always more than one party in any contract and predicting outcomes is certainly not a science. Finally, if we knew which doors were in front of us and someone could open a door and say "down this path lies certain ruin", I'm not sure that people would make so brave a call as to halt at £500 (or £500,000,000); the vendors, likewise, would not likely take the loss at that point. The bigger the number, the harder the call and the further to fall there is. It's a big question to resolve and one that needs more than a simple blog article. It can't be solved with one idea or another, but will need much effort from many angles - vendors, government departments, interested third parties, universities (who are turning out the next generations of outsoucer employees) and interested customers of government who have a point of view. So, this is not to belittle the idea that Ian offers - I think there's merit there - it's just to lay out some of the complexities. Now, what I think could be interesting, is a place to prove legislative thinking - to understand better the impact of various policies on government, its customers and its workers. There is perhaps no finer example than Tax Credits. The Inland Revenue assumed that the beneficiaries of tax credits would be people who paid tax, i.e. their usual customer base. These people were, for the most part, nice, controlled people who paid the IR monthly through PAYE or annually through Self Assessment (with some variations to deal with e.g. self cmployed people). In reality, they got mothers with kids who would super glue their hands to the tables at tax offices, mothers who would leave their kids in the tax office and insist that they would stay there until the IR paid them money so that they could eat and, to make it all so much worse, people who had an income that changed month to month, let alone year to year. Understanding the difference there could have paved the way for a far more sensible discussion about the best way to calculate and pay the credits. There were many in the IR who thought this way at the time and who argued endlessly that there would be a better way to do it, but their protests fell on deaf ears. The system problems that occurred at launch only compounded a problem that was already going to be terrible at best, disastrous at worst. Subsequent problems with ID fraud and fraudulent claims, only make the original thinking yet more dangerous. In truth, there are some fundamental flaws that make the risks big for government projects. These are borrowed from a colleague who I suspect would choose to remain nameless: - A willingness to buy the aspiration rather than the reality - A failure to confront the gap between policy and operation (i.e. having a policy is not the same as delivering it) - A bizarre reliance on IT to solve problems which ought to be addressed in business terms (and IT's willingness to accept this challenge). This is where Mike Cross chips in with "If I had a pound for every time someone says 'it's the business, not the technology'." For as long as these remains the case, all major cross-cutting (such a lovely government word that: many people are cross in government and not nearly enough is cut, but I digress) programmes will present an extremely high level of risk. Borrowing some further words from a learned colleague, who realised all this was tough far longer ago than I did and who has spent most of his time since trying to correct it, here are the "we will nots" that every IT department should stick to: We will not: Accept the challenge of integrating, at the back end, what Policy wonks and Ministers have designed, at the front end, to be separate and different. Entertain discussion of ‘whole customer experience’ for so long as policies and products continue to be defined in terms of non-entities (i.e. things/people/organisations that government doesn't actually transact with) Put sophisticated management information tools in the hands of managers who have no aptitude for managing by reference to information. Record management information at any frequency which is shorter than the business’s decision-making cycle. For most this means annually. Accept Business Cases containing reference to benefits of greater flexibility, improved working environment for staff or better decision making. Too many broken promises, too many bruised hearts… Tolerate delay while anguished discussions take place about FoI or DPA. It’s a business issue, so sort it out before you ask us to build anything. Try to persuade you that the latest upgrade will solve all your problems. It won’t. You are on the treadmill of ever-increasing expenditure for ever-decreasing additional benefit. Get used to it. On those points, I'll close.
Monday, September 11, 2006
It's probably easy to get too anal about all the running data I get from my Garmin Forerunner 305, but after a tough run on Sunday I was looking for some answers. As part of the prep for the New York Marathon on November 5th (sponsor me!), I ran the Robin Hood Nottingham Half Marathon. There was no sign of Robin, and definitely no one handing out money to the poor (or even jelly babies to the runners) but there was plenty of sunshine - despite a foggy start to the day, it was 24C at the off and somewhere near 27C by the time I got round. It was easily the toughest run I've been through - partly because I wasn't feeling great, partly because of the heat (although I've run in hotter times) but mostly because there were lots of hills and they all felt steep. And so I must not be in the shape that I thought I was in. The Garmin records distance, elevation and speed. My SportTracks software graphs that better than the Garmin in-house software does (and, I think, better than MotionBased). So here are three graphs in a row, one from the Great North Run (last September, 1h 59m), one from the Liverpool Half (March this year, 1h 45m) and one from Nottingham (yesterday, 1h 46m 50s). Nottingham felt steep, but I wanted to know if it was as steep as I thought - and how did it affect my running. This first graph shows the elevation and pace for the Great North Run. I started off too fast - that's what comes of being on the white start line when they sound the klaxon. From there, I get slower throughout the race, independent of the hills (and everyone tells me that the GNR is one of the steeper races) - there's a brief blip at 14km where I caught someone who nearly fell over as I was running alongside them - and then I get a little quicker near the end as I see the 2 hr time get a little too close on my watch and pick up the pace a bit. I finished in 1h 59 something. Next up is the Liverpool Half. Here I started somewhere in the pack and so the first few hundred metres are spent dodging people. From then on I ran pretty much consistently, mostly between 4:50/km and 5:10/km. There's a big hill in the middle - the course takes you down the hill, along a little and then round and back up the same hill. That certainly took something out of me, but the hill was short and steep and it was over with quickly. Lastly, Sunday's race at Nottingham. The spread of pace here is from 4:20 to 5:20, with a high degree of volatility. The hills are more frequent, steeper and generally last longer than either Liverpool or the GNR. Half way through I was ready to drop - volatility increases, largely independent of the hills, yet I consciously try and speed up in the last 4 km, trying to get round inside my target time. I was very focused on 1h 40m at the start, and held that time for, at most, a mile. Things got worse from there and I finished in 1h 46m. In 2 weeks I have the Windsor Half Marathon which is supposed to be moderately bumpy. Between then and now I have another 30km+ run to get done, then a week off after that followed by the Nike 10k. And then it will be a month to go. So far I've run 445km in training. I'm a little ahead versus the run up to London earlier in the year but my times are looking a little slower. Of course, I ran Liverpool and Reading 4 weeks and 2 weeks before the London Marathon, whereas I'm running these 8 weeks and 6 weeks before, respectively. Lots of variables in every run. Perhaps more on this when I'm through Windsor. Hopefully the Queen will stop by to wish us all well.
Posted by Alan at Monday, September 11, 2006
Sunday, September 03, 2006
Tuesday, August 29, 2006
Ages ago I was asked about Value For Money and specifically whether I'd measured how we'd done in e-government and how I'd explain it to "Jim Bailey" I'm going to take this in a slightly roundabout way - partly because I'm feeling my own way through the question and partly because I don't have the time right now to condese the points into a single, short thread. I hope when I get to the end, the right route through it will be obvious to me and I can post a summary. Here's an extract from a February 2002 paper on the issues government is facing nearly 2 years into the e-government agenda. I was trying to lay down the argument for greater co-operation between departments rather than the historical silo approach. Bear in mind that, at this point, government had perhaps 800 websites in total, maybe even fewer: • Advanced technology. Requirements have developed from the initial need for “just a web site” to more complicated “portals”. New technologies will be required, including enterprise content management systems, multi-platform delivery parsers, discussion forums, e-mail exchange software and so on. Costs and risks rise per instance of deployment. Our track record for implementing any project, let alone those that involve new technology is poor. • Systems management. As we aggregate content and capability into single systems and as usage of e-government increases, each of these systems becomes a component of the Critical National Infrastructure. CNI requires rigid management, ability to absorb peak volumes (at the end of the tax year, for instance), well-rehearsed continuity of business plans coupled with disaster recovery locations and so on. • Security and accreditation. Accreditation of a single system presently runs at £100,000, whether it is connected to the GSI or to the Internet. Despite departmental experience, these costs are likely to remain, as back ends are connected the Internet, personal data is held on their portals, increasing the risk of security intrusion. • Identification and authentication. Making sure that you know who the person dealing with you is and that they have a right to see the data requested remains one of the largest challenges to delivery. Implementing digital signature signing capability can cost up to £250,000 per department (if done alone) along with additional costs of £40,000 for each new certificate provider (presently there are three but this is expected to rise to ten or more). There can be no cross-government authentication unless there is a single virtual record that different departments can trust. • Partnerships with intermediaries. There is a step cost for intermediaries to communicate electronically with government. Vendors communicating with the Gateway estimate their costs at £100,000. For many, this will be the first time that such a change has been made. Proliferating many ways of communication with government will increase these costs, reducing the number of intermediaries and therefore the number of people using the services offered. • Back end integration. Perhaps the hardest technical challenge – preparing back end (or “legacy” systems) for dealing with electronic transactions. Only a few departments are addressing this at a strategic level. Few vendors have taken what they know from one department and offered it to other departments at minimal cost (hence we have hundreds of systems that do much the same thing) – the first recent opportunity to do that was presented by the Government Gateway standardised interfaces. It is essential to engage all of the vendors in an overall programme of upgrades that will position us for the long term. The choice was simple - let a thousand flowers bloom, or concentrate on a few key programmes and drive participation around those. With the Shared Service agenda now in its first full Spring, it looks a little like too many flowers are blooming - and that the lessons learned from e-government, whilst learned, are not being lived. Offering a shared service is easy - here, come and share this, it works just the way you need it to. You take the CD, install it in your server at your data centre and everything is fine. Alternatively, you just copy your data across the network from your data centre to their data centre and off you go. The real deal is more complicated, inevitably. First, you need a few willing participants. Two is not enough - two folks can strike bilateral deals based on strong relationships, can wiggle through a few tight gaps and can, eventually, get something together. The Gateway started with three - and that brought in complexities: big departments, small departments, citizen focused, business focused, digital certificates, userid/passwords, outsourced providers, inhouse providers, technically literate customers, technically far from literate customers etc. Now we need someone in charge. That's going to be hard. Who should it be? The biggest? The smallest? The one with the most to gain or the most to lose? The one with the strongest person at the head? The one with the biggest team? The one without and IT partner, or the one with? The OGC Gateway reviewers will focus in on governce and will want to see a Senior Responsible Owner in charge with the right stakeholders engaged. And, to make it serious, everyone has to put some money into the pot. Enough to hurt if it goes belly up but not enough to cripple them. Only gambling money you can afford to lose (as the spread betting ads say) doesn't give the necessary incentive to make it work. It will be pretty weird to see government paying government money - someone with safe hands will have to guard the pot and make sure it's spent properly. To add to the spice, each of the participants should have a separate IT vendor. And this is where it will get fun. Contracts being what they are, it will be uncertain how one department offers services to another. The "indivisibility of the crown" means that one department can't sue another in the event of failure - and so a kind of musketeer principle has to prevail (all for one and one for all). That takes time, it certainly isn't natural. On top of that, the service will need, eventually at least, to be productised so that it can be offered to more people. That means documentation, service management standards, professional live service, disaster recovery and so on. Lastly, we need folks who are going to get together regularly and figure out how to get this done, who are going to sort out the real requirements from the fictious ones, who are going to prioritise good from bad and who are going to manage the various vendors and the challenges that will loom in the coming months. We want people who are going to give up large chunks of their lives to see a difficult project realised in full technicolour. We want people who can learn to trust each other and who can deal with the politics both in the room and in the rooms back at base. So three departments is the minimum, four or more is better - but increases the risk of not getting anything done because of no agreements. Money is needed, a person in charge and a way to deliver services to other departments. So how about we count how many shared services look like this?
Posted by Alan at Tuesday, August 29, 2006
Sunday, August 27, 2006
Saturday, August 26, 2006
Every where I go now all the talk is of Shared Services. Where once the mantra was "2005 / 100% online" (or, for a few, "joined up / citizen focused") it is now all about sharing. From the CIO's very own website: There are over 1300 public sector organisations in the UK and there is considerable scope to consolidate how we can share services across organisational boundaries so that our common processes are provided faster, better and cheaper. Google lists 582,000 mentions of the string "'shared services' government uk". That's quite a few less than the string "'e-government' government UK" which has 9,990,000. But, nonetheless, a healthy total given the relative newness of what we might call government 2.0 were a similar name/version not already in vogue for community-based sites around the globe. Indeed, I'd probably go for government 3.0. 1.0 was all about silos, castles and fortifications (was ever thus, at least since Cromwell's day), 2.0 was all about the promise of the web to break down the barriers and join up the front end whilst buying time for the back ends to be integrated and 3.0 is the promise of true sharing, whilst being cognizant that 2.0 didn't quite come off. A well placed source in government told me only the other day that he feared that there might already be 1,300 shared service projects in UK Government, i.e. one for every governmental organisation. That number is eerily familiar as is the thinking behind it. In the days of 2005 / 100% online, the start was one website per department (the end, if today can be referred to as the end, was sadly different with well over 4,000 websites). It doesn't seem unlikely that every one of the 1,300 government organisations is embarking on a shared services project, secure in the thinking that "of course their services is shareable, any body who wants to can come and use it". You can see the business cases now. Have service to share, all offers welcome. Sadly, with very few exceptions, not one of them will be shared with anyone else. Long ago, I used to put this picture up at conferences as an intro to shared services. The image below is from a deck presented to government's "e-champions" in October 2001. I'd ask the audience how many of them ran an expenses application. Every hand would go up. Then I'd ask if they thought that they could consolidate all of those into a single application - after all, expenses processing has to be one of the simplest applications out there. All the hands would go up again. I'd then say "so how about we do it then - we can start tomorrow." There's be a bit of shuffling, a few hems and hahs and a lot of people looking at their feet. Of course, I'd say, that would ne naive. Some of you like having the name of the person claiming at the top left, others at the top right, some of you sum the numbers differently, you'd all have different cost codes, some of you would self-certify and others would require some workflow, some would write cheques direct and others would do it through the payroll. It was never going to work. So I'd propose a way forward. Rather than consider a shared service as a project now, why didn't we define a standard for how expenses programmes would work in the future. Then, as everyone refreshed their applications and built new ones, they'd be built to the new way. We'd work with SAP and Peoplesoft and so on to have the new standard built into their applications, using the might of government and, pretty soon, a standard expenses form would be a dropdown choice in the configuration menu. People could converge when they were ready to. Had we started that in October 2001, I wonder where'd we be now. But then, it's not as easy as that. Joining up is hard to do. Three related factors get in the way: Power, Control, Trust. No one in government wants to reduce their power base by giving up control of delivery or operation and handing it to someone else who they don't trust to do it right. How do I know this? The first three government shared services: the Government Gateway, DotP (host of directgov) and the Knowledge Network (designed to make the sourcing and compilation of ministerial briefs simpler) are still, five and a half years after launch the only shared services out there. I was involved in the building of two and the operation of the third. In June 2001, this picture attempted to show the output of the business plan for the e-Delivery team, inside the Cabinet Office Government organisations, as is typical, have latched like barnacles to the bottom of an old trawler, to the promise of new funding for shared services - hey, I did the same in 2000/1, when the badge needed had a different name. Today, everything is being badged as shareable. Pretty soon, we will see napster.gov (formerly used by the Inland Revenue to let you share tax returns that weren't yours) rebranded as a clearing house for sharing services. The number of "tracks" available for sharing will soar, the actual number of downloads will hover near zero. There's one person in government who has the potential to change this - and that is Sir David Varney. Whilst all has been quiet since his appointment as a new Permanent Secretary at the Treasury, I hope that he's working on how government can build a convergence plan, starting with one or two focused but high impact services, that will make the case for doing this right.
Posted by Alan at Saturday, August 26, 2006
Saturday, August 19, 2006
The scoreline for this occasional series is reading government 2, e-government 1. Sadly for the future of online services, today's topic is not going to improve things for the e-world. I'm not particularly picking out services that are under-performing, but selecting things as they come up in the news or just because some data comes my way that piques my interest. If there are other services I should look at that can redress the balance, leave a comment. What got me thinking today was the latest monthly report (June 2006) from the e-Delivery folks (who are perhaps now the PMDU folks - reliable sources tell me that they have not only fused organisationally but physically too, strengthening the link they had with the CIO, John Suffolk, working for Ian Watmore). It shows the Government Gateway cruising past 12,000,000 transactions since launch. Sure it's only a fraction of the annual total let alone the total that government has handled on paper since then, but it's growing. Except, and this is the kicker, it's not growing as it might have done. For the last few months, the number of transactions from those applying for, or updating, tax credits has held steady at 1,971,889. My memory is vague but I'm pretty sure this service crossed a million transactions in its first year. Indeed, over 447,000 people have enrolled such that they could update their claims electronically every year. Those people have all been disenfranchised. Since November 2005, this service has been closed after apparently wide-scale fraud was discovered and after some £130 million had been falsely claimed. A lot of money, but not as much as has been written off through error or because of the policy of basing claims on prior year earnings with the resultant need to retroactively reclaim money when earnings change in year. Over £2 billion has perhaps been written off because of these problems. These numbers are highly variable. The BBC quotes £2.2bn written off, of which around 15% was the result of fraud. The claim reclamation process was always controversial: you give people money, they spend it; few put aside any "in case the taxman wants it back" - how would you know how much with such involved rules? I advocated making early government services the giver of money to the public (see give and take.gov and this one), rather than the taker so it's time I thought through the implications of this. My thinking originally was that tax payments, whilst the first transaction that almost every government (globally) put on line would attract a niche group of "early adopters" (Geeks?), giving people the chance to actually get money faster from government than usual should result in a rush of traffic. Far more people take money from government through transactions that they initiate than send it - PAYE, one of the larger transactions (27 million employed people), is not self-initiated but handled through less than 2 million employers. Tax credits proved that the rush would come - although perhaps not for the right reasons (as the BBC says, "Low Risk, High Reward, Easy Hours") - with the online service made available months before the paper service, coupled with an advertising campaign that directed people to the website. Interestingly, if you try and access the tax credits site from the home page of HMRC, using the left hand navigation bar, it gives a 404 page. The service availability page tells me that tax credits is "temporarily closed" . 9 months is more than temporary. The rest of the site covers what to do now that you can't do things online: Renewing your claim is more straightforward than you think 1. Wait for your Annual Review Pack to arrive in the post. It will contain everything you need. 2. In the meantime you should get together the documents you'll need to check the details about your award and to work out your income. These include your last payslip or P60 for the tax year 2005-06, any receipts and records for childcare costs, and, if you are self-employed, your Self Assessment tax return. 3. When you receive your Pack read the instructions on your Annual Review form and check that the information we have about your personal circumstances throughout your award are correct and complete. 4. When we write to you, we'll tell you if you need to fill out any forms and when you'll need to return them. If you do, the next step is to give us the information we need and wait for your details to be processed - we'll aim to do this within 30 working days. That would not appear, to me at least, not nearly as straightforward as doing it online. Still, I never qualified for tax credits, although I was one of the first few to try the online calculator so can't be absolutely sure that the online process was easier. I'm rambling. The real point I wanted to get to was to ask "is this the death knell for online receipt of benefits (the give of giveandtake.gov) or is this something else?" The folks at HMRC have been pretty coy about what actually went on, but the folks at the BBC have a handle on a few things: The fraud was made much easier by a devastating security breach at the DWP: the theft of payroll data from the 2003-4 financial year which put names, addresses, dates of birth and national insurance numbers of at least 13,000 DWP employees in criminal hands, most likely thanks to action by insiders. Staff in centres ... have now found that bank accounts have been opened in their names and money siphoned out, potentially destroying their credit ratings. and The new method [of fraud] is an identity theft attack on existing claimants, where a bogus telephone call changes the claimant's registered address. Then, once enough time has elapsed to assuage suspicion, the fraudster changes the bank details as well. Payments, it seems, do not have to go to an account in the name of the claimant - and some accounts are being used to aggregate such fraudulent returns, something which is beginning to be flagged by banks' and building societies' anti-fraud systems. and, perhaps the clincher: The first the claimant knows is that the payments suddenly stop. But if they try to resolve the issue, they then fail the security checks thanks to the fraudulently altered personal information. So we're not really talking about an online fraud here - this isn't about passwords being guessed or accounts being hacked into (take note HSBC). This is about old-fashioned social engineering, stealing data on paper or insiders making off with payroll data. The internet is then used as a quick and easy way to bulk upload the claims into the tax credits system - I can see an offshore "farm" of people now, busily tapping away and getting a few pounds for every claim they put through. That means, presumably, the fraud is still going on now, except that same farm is busy writing out, in neat copperplate, dozens or hundreds of multi-page paper applications. The real issue here is that authentication checks are inadequate across the whole of our paper-based transaction system, whether that is banks, government or loyalty cards. In the modern age, "tell me your mother's maiden name" doesn't cut it - it hasn't cut it for a decade, yet so many services (online and offline still use it). That latter point is what attracted us to digital certificates in late 2000 when the Gateway was first built. If only we and the industry had been able to figure the technology out to make them reliable and useable. But, digital certificates, even if they work well, only secure the online part of the deal - the paper and telephone channels still rely on signatures, the occasional face to face visit and a few chunks of ID that appear to be readily available and certainly available for less than a full year of tax credits claims. Nope, to make this really work, we're going to have to bring together some far more rigorous tests involving, at the very least, dynamic data. Already, HMRC are using this for the online VAT service where you have to provide a wealth of information including how much VAT you paid last time, what your last payment date was and so on. I haven't seen a service aimed at individuals do this yet, but I'd like to know of some. Five years ago, we proposed a link with Experian (the credit checking company which was, at the time, part of GUS but I think is in the process of demerger now) to help with dynamic checks. Indeed, a pilot was run for offline applications at the Passport Office for a short period - although I believe staff didn't like the "probability" assessment that Experian used and so rejected it in the end. Not long after that, we proposed a kind of "Green Shield Stamps" model where you'd get various stamps in a virtual book that indicated the depth of relationship government had with you - if your bank vouched for you and you banked online that would be worth "X" points, if you had paid Self Assessment from that same bank account for the last 3 years that would be worth "2X", when you got to, say, "5X" we'd allow you to claim money from government, to that same bank account. We thought this had a lot of merit, but it didn't get past go and didn't enable us or anyone else (least of all the fraudsters) to collect £200. It had its flaws for sure, not least is that it wouldn't have been any good for the first transaction anyone made - but maybe that's the point? Dynamic data coupled with cross-service enrollment is surely the right way for government to deal with this fraud. For a fraudster to perpetuate over several months a fraud that involves multiple services, making payments to government and using accounts that have been verified by the major banks and their Know Your Customer policies is certainly possible, but the effort is far harder and therefore the payoff in terms of risk/reward not so good. Fraud is much like terrorism though, you squeeze down in one area and another route is found. With the online channel shut, the fraudsters are doing it the old way, that's proving harder now with more focus on it, so they're checking out other means. Has the absence of an online tax credits service put back the journey to ubiquity of e-government? It probably did in November/December when it hit the news. But since then, things have been quiet and people have perhaps forgotten the service even exists - or can't find it when they click on the link anyway. When the service comes back, perhaps quite soon (how temporary can they mean?), there'll be a renwed bout of press interest and, not unlike Self Assessment's Napsterisation (when you could apparently see the tax returns of other people), the press will drive interest and there might be a new rush of usage - that, in turn, could drive cross-usage of other services (after all, it's after 2005 now, 100% is apparently online). It worked for the PRO all that time ago after all. But, I'm calling this as e-government None, government 1, taking the overall tally to 3-1 to the corridors of power in the old fashioned edifices of Whitehall.
Posted by Alan at Saturday, August 19, 2006