The question of whether there should be a local version of the Government Digital Service rumbles on. I wrote about it a couple of months ago, and lots of other people have too, most of them far more expert on the question than I am. That amounts to a lot of well-informed and passionate commentary, but much of it is quite abstract. This post comes at the issue from the opposite direction: if solving problems such as those described here is the question, is a local GDS the answer?
That thought was promoted by going through a new digital service offered by my local authority. I was delighted to find that it existed (it replaces queueing up in an office half way across the borough). It seemed to work (though at the time I wasn’t entirely sure since fulfilment remains firmly undigital and took a long time to happen). Some of it felt quite liberating (not needing to provide the same evidence of identity for the umpteenth time as for the first). And some of it seemed designed to snatch defeat from the jaws of victory. This starts as a story about a website, but it turns out to be much less about that than it might first appear.
So first a few reflection on being a user of the service. And then a few more general thoughts on what that might point to for doing digital better.
I want to buy some parking permits. Not for my own car, because I don’t have one, but so that occasional visitors can park in residents’ spaces.
At first blush, the new modern looking service might have come out of the GDS stable. But on closer inspection – or actually using it – it’s riddled with small details (and one or two very big ones) which make clear that it’s actually a very different kind of beast.
Let’s begin at the beginning. “Do it online” the landing page exhorts, followed mysteriously by “(external site)”. A couple of screens later, you get to choose what kind of permit you want, in a two-dimensional menu, the like of which I have never seen before. Clicking on “Visitors vouchers” has the unexpected effect of making the page scroll down, to prompt either logging on or continuing without an account. The chosen option is in fact highlighted – but on my reasonably large screen, the scrolling goes far enough to make that invisible, introducing another sense of uncertainty about what might actually be going on. Moving on, they want to establish that I am entitled to buy permits in the first place. This is genuinely a big improvement: instead of scrabbling around to find old gas bills, they just do a credit reference search.
Then it gets really strange. The next screen allows me to select the number of parking vouchers I want. Or does it? It’s the same two-dimensional tiled menu as before, but this time clicking on any of the options has no apparent effect whatsoever. There is a basket on the right of the screen which remains resolutely empty with an amount due of £0.00 as I try to choose ever larger quantities of permits. Nothing happens. In the end, since there is nothing else to do, I click on ‘Continue to terms and conditions’. Two things happen on the screen which follows. The first is that the basket now shows the number of vouchers and the price. But since I was randomly clicking around trying to get some kind – any kind – of reaction on the previous screen, the number in the basket is much larger than the number I actually want to buy.
The second thing which becomes apparent is that there is no way back. None. Not then, not at any subsequent point in the process. Having made an invisible choice for an invisible basket, the only way of correcting that choice is to abandon the process and start all over again. But if you try that, there is a further sting in the tail. I kept going with the wrong amount as far as I possibly could, in the expectation that at some point there just had to be a a way of updating the quantity. I was wrong. But it then turned out I was worse than wrong. These permits are rationed – 50 per household per year. This new system keeps a running total of how many I have bought and thus of how many I am still entitled to buy. And it has deducted from my total the permits I didn’t buy, as well as those I subsequently did.
But eventually I did get to the end, ready to buy the permits I wanted, though with my confidence a little shaken by the effort it had taken to get there. The payment process did not offer quite the reassurance I had been hoping for – “Redirecting to Payoffshore……” it proclaimed, flourishing a name which could have been invented by a Nigerian spammer on a particularly unimaginative day.
But that’s finally the end of the process. Or rather it’s just the beginning. I have asked for – and paid for – some permits. But I don’t yet have the permits and there is some back office work to be done. Now as it happens I know roughly how long that takes, because in the old system I used to queue up in an office and watch somebody do the work on the spot. Five minutes is plenty. Ten minutes would be generous. Fifteen minutes would be absurd. Apparently, though, it may take them up to ten working days to find the five minutes needed.
And remarkably it turned out that they needed every one of those ten working days. Counting actual days and time in the post, which is what matters to me, it took 17 days for those permits to get to me. It’s rather a good thing I wasn’t in a hurry.
I have told that story not because the details are important but because it’s a really good illustration of lessons at three different levels. Only one of them is ostensibly about the web service, which is a good reminder that ‘fixing the website’ is rarely a good way of framing the problem.
At the first level, though, even if fixing the website isn’t sufficient, it’s certainly necessary. User confidence is a product of complex cues and responses to them which are not always conscious. No one of the problems I have described – or of the several more I could have described – breaks the process and any one of them is surmountable. But cumulatively they do a lot to undermine confidence. I have no idea what the completion rate is, but I am willing to bet it could be a great deal higher.
This is also a good illustration of the principle that if you are going to use non-standard interaction design, you need to be really confident that it works, which includes making absolutely sure that it offers the right affordance.
One level up from that, it’s pretty clear that polishing the website will not address the problem that it is a veneer over a broken process. A process time of five minutes should not go with an elapsed time of 17 days, particularly when the old, slow inefficient manual process has an elapsed time of about 90 minutes, including travelling and waiting. It looks as though somebody has designed a web transaction. It doesn’t look as though anybody has designed a service.
But we shouldn’t stop there. I am not really buying small cardboard rectangles, which need to be physically delivered. I am buying the right to park in certain circumstances, which doesn’t. Tax discs are in the same category and it seems auspicious that this post is being published on the day they are being abolished. Long before there was any prospect of that happening, but when it was first possible to renew the tax disc online, William Heath was pithy and precise:
Hurrah for the car tax disc renewal process. It used to be both inconvenient and pointless. Now it’s just pointless.
It’s always a bad sign when a piece of information held by government is converted into physical form in order to allow the same or another part of government to convert it back to digital. Why have a shiny new service to issue parking permits when it would be so much better to have a shiny new service to not issue parking permits?
The moral of the moral
This has been a pretty limited story. It is one person’s experience of one service offered by one local authority. But that service is newly redesigned and newly automated, and I am pretty sure that it’s not yet as good as it could be or as it needs to be. More specifically, I doubt that it would meet the GDS digital by default service standard. That’s not to say that this is a job for GDS, still less one that only GDS could do. But as local authority services go, this is at the simple end of the spectrum, and it looks as though something better is needed to get things right.
Is a local GDS the solution? I made slightly heavy weather of the general answer to that question in my earlier post. Coming at it afresh from this slightly different perspective feels easier in some ways – but only in some. It is much harder to do this stuff well than to write blog posts criticising the doing of it. Concentrating and amplifying that expertise has much to commend it. But the other thing which GDS critically has is a licence to operate. It has renewed and extended that licence over time, but it did not create it – the initial spark has to come from elsewhere.
On balance, the story told here makes me more inclined to the idea of a local GDS. It would have to find ways of being local. It would have to find ways of not being GDS. It would have to be as concerned about back office processes as about visible interactions. But if it could be all those things, there is no shortage of great work waiting to be done.