Monday, May 30, 2005

Unfreezing the interface

Broadband is now officially bigger than dial-up. Who'd have thought it could happen so fast? When I got my first broadband connection in the UK in July or August 2001, I remember BT having to dig up the road outside my apartment because there wasn't a good enough line into the house. It took 6 weeks of to and fro to make it happen. In 2001, everyone was saying that for broadband to make it big, there would have to be useful services offered over it - that speed alone was not enough of a sales pitch. I used to do conference presentations and joke that the single biggest thing I used it for was to stay up to date with software patches that were, by then, already getting too big for dial-up. Itunes, Limewire and BitTorrent are, I'm sure, some of the big drivers of usage today. 16 million people can't be wrong. First it was the geeks that led usage before the chasm was crossed and Joe Public took up the charge. In a conversation with Marc Andreessen, over dinner at Zuma, a couple of years ago I asked him about the "back" and "forward" buttons and when he thought we might move on from there. I figured that they were already passe, being around 10 years old, and that they were a mark of a very linear web experience, yet, now, sites were starting to evolve far more complicated structures. He said something that has stuck with me since, "Interfaces freeze early." If it works and enough people get used to it, changing it dramatically is a hard thing to do. BillG said something similar once - he said that only incremental changes to the Windows UI could be made as anything too big would make it too great a leap for users who would be confused. The interface to government has been frozen for a long time. It sort of works and a lot of people use it. That's one of the main reasons that most online transactions with government look just the same as the offline transactions. Attempts to create different, better front ends have largely failed - if you need evidence, compare the usage of the Inland Revenue's own Self Assessment form (which is just the same as the paper one) with Which Tax Calc (which tried an innovative approach of asking questions about your finances and completed the form in the background); yes, Tax Calc cost money (I think perhaps £15 or £20) and yes, the IR form was free, but is that the only reason or was "good enough" coupled with familiarity with the form enough to keep people using the old interface, albeit with a shiny, electronic coat of paint? I had thought that competition bred the chance to create change in the interface. Based on the Tax Calc example above, perhaps I'm wrong. If Tax Calc had offered its product for free - and perhaps gone the Google way with targetted ads (easily done with ads about finance and so on based on a profile of the tax payer created by the product), would it have been more successful? I'm not sure. I'd like to think it would have been, but I wonder, again, whether the interface in this case is, in fact, the Inland Revenue, and that's what has frozen early. Thinking of freezing, when the Somerset House ice rink was being constructed for the first time - Winter 2000 I think - I suggested to Nick Montagu, former Chairman of the IR, that we should open it a day early exclusively for the IR staff to have some fun. He held his head in his hands and shook it wearily, counselling me that he could "see the headlines now ... Inland Revenue on thin ice again." And so, it opened on time for the general public alone. All the news about Gordon Brown setting up a new "benefit" for new house-buyers - where the government would own a share of the house, as would the building society lending the money - got me thinking. I have no idea if this equity-sharing thing will work and don't really want to get into the politics (e.g. MIRAS all over again?) or finances of it (e.g. who bears the brunt?), so bear with me whilst I ignore all that and, instead, use it as an example of how we might unfreeze the interface. First the assumptions:
  • The Chancellor has a declared desire to cut red tape. He wants to merge 29 regulators into 7, for instance. He also wants to take a "risk-based approach to regulation", something I thought was already done but perhaps he means a better stratified approach. As part of this approach, he wants to cut form-filling by 25%.
  • Those applying for a home purchase are already known to government. They're over 16 so have an NI number, they're probably employed so have a relationship through PAYE, they may be receiving tax credits, they might be claiming child benefit for one or more children, they're living somewhere now so pay council tax etc. They may also be known the financial institution where they're getting a loan, perhaps because they have a bank account or a credit card (indeed, to make that more certain we could insist that you could only get a loan from a bank/building society where you've had a relationship for more than x years)
  • It is realistically possible, today, to create an entirely paperless transaction that will be legally acceptable and be accessible to the entire country, whether through broadband in their home (whether they live in rented property, with their parents or wherever) or through a service offered by an intermediary (the CAB, the building society doing the loan etc)
Combining all of these points, we need to design a service that
  • Is Internet only or, better still, entirely transparent within existing processes
  • Reduces to near zero the number of "forms" to be filled in, i.e. does not require anything new to be filled in
  • Relies heavily on existing relationship data, with government and/or with the lender
  • Is self-maintained e.g. when the house is sold, the update to the processes happens in the background without any need for intervention
This wouldn't reduce "red tape" but at least it wouldn't result in an increase - and it reduces the ratio of "forms:services". The idea would be to create a model here that could gradually be backtracked into existing processes so that there would be a reduction in bureaucracy a piece at a time, with over-lapping and duplicate forms being eliminated (and ceremonially burnt in a huge bonfire come November 5th) Any new benefit, tax credit or other government service could be introduced only if it was a natural extension from an existing service that required no additional bureaucracy to qualify for it, or to prove entitlement. If it wasn't an extension, then something else would have to be deleted to allow this one to surface - or, better still, 5 things would be eliminated to introduce 1 new one, with some measure applied to ensure that the 5 things deleted weren't all 1 page simple processes to allow a 500 page mammoth entry. The lending bank, as part of its "Know Your Customer" process has already identified who you are, you have had a relationship with them for a while, your salary is paid into the account so they know where you get your money (and how much you can afford to borrow). The bank also knows how to talk to government and exchange data with it and the various departments involved in the process. The trick then is to link the "bank-known" data with the "government-known" data, i.e. to match "Joe Smith, Account number 010101010" with "Joe Smith, NI number 1010101010". So, in this case, the IR would provide the tracing services, cross-links to, say, the local council for other data and, data protection rules notwithstanding, we have the potential to set up a service where:
  • You have no need to prove your identity to government because the bank has done that and taken the trouble to link your "real world id" to your "government world id" - the two are not in any way the same, believe me. With that initial link made, leaping to new services online would be simply a matter of tagging extra identifiers to your main ID. I'm not talking "ID cards" here, just ID as in identifiers.
  • If we really had to, we could issue a digital certificate, perhaps on a USB dongle or as one of those little calculator-style things with constantly changing numbers.
  • The bank sets up the payment of the government's part of the mortgage either as a direct payment to them or to you as some kind of tax credit or benefit, to the bank account that you hold with them
  • If you sell the house, the bank will be involved in the sale and so can cancel the transaction. The bank can even cross-sell you insurance and god knows what at the same time, defraying some of the costs perhaps
I think this could work. More importantly, it starts with the idea that no new government service should be created that is based on a "form" or a "paper process." Only if we take that first step can we unfreeze the interface. Five years ago, I was involved in lengthy dialogues with a certain, large government department. They were launching a new service. It involved oodles of new IT, deployed at a huge cost. Risks and tensions were high given government's IT track record. The pitch that I and several others made was that the service should be introduced on the assumption that, by the time it was launched, it would be viable for it to be a fully online service. We were laughed at - this was mid-2000, when dotcom was in full flow (if perhaps in the first stage of its downturn) and everything was possible, or at least so we all thought. I sponsored (and paid for personally) a big dinner where we brought together all of the main protagonists at Teatro restaurant (now, sadly, a night-club come bar that has declined enormously - but I suspect its owner has far more significant things on her mind now). We agreed a ground-breaking deal where an Internet front end would be created for this service that would be available months before the paper and back end mainframe process, reducing the risk of the mainframe IT, reducing the bow wave of applications that would result from the paper process and hopefully, creating a transition to an Internet-only model. We asked for many other things to happen along with this, including changes to the back end architecture and parts of the process to make future changes towards Internet as the default much simpler. But, in the end, the most expensive dinner I'd ever paid for (at the time, it's since been surpassed), resulted in a monumental deal. It was, perhaps, the first step towards reducing red tape, reducing the need to fill in forms and unfreezing the interface. Five years on, it's time to restart such initiatives and lead a drive to decouple the link between "government service" and "form".

1 comment:

  1. Was amazed to find someone blogging about mainly eGovt stuff. =p Actually I came accross this site while looking for inspiration for a m-govt tagline. I'm still clueless =(
    Govt websites are not quite 'frozen', it just takes somewhat monumental effort to change them, be it to improve the interface, integrate several services for greater customer convenience, put more bits of the service online, include more payment modes, etc. Old sticks in the mud are... old sticks in the mud. It might sadly be easier to wait for the mud to swallow the old sticks than to pull em out.
    In any case, e-filing of income tax is possible in my country. In fact, the utilisation rate is about 60%. However, it seriously doesn't seem like the best application to be put on a mobile phone... then again, the Swedes are doing it.
    Well, govts arnd the world are hyping on 'customer-centric' services, so the best way to un-freeze might be to ask your customers what they really want from govt e-services...

    ReplyDelete