Visitors don’t make websites, transactions do

Building customer focused, task driven council websites has been a growing movement over the last few years, and finally, with a little grease from the central government wheels*, many local authorities are turning their websites from organisational behemoths, into streamlined customer “doing” machines.

While these websites are by no means perfect, the next battle has to be measuring these websites properly, so that local authorities can make informed decisions and focus their ever dwindling resources into areas that need most attention.

Measuring visits

I loved the recent article about customer focused web development at Maidstone & Tunbridge Wells, but I admit to feeling a wave of frustration when I got to the last paragraph, where they gleefully told me in the first week of launch, there were more than 17,000 unique visitors to the site”. Visitor numbers are a very attractive metric with which to judge a site, but they are one of the most dangerous measures.

Measuring visitors is the direct result of applying the existing metaphors of contact measurement to a new technology. In a physical environment, the number of calls is an important number, not least because you physically need people to answer those calls. Online, once you have more than 10,000 visitors, the next 100,000 make very little difference.

Visit count is a number you have little control over. Your website might get 10,000 visits this week, but next week when the bin-men go on strike and half your schools close for snow, your visitor numbers will rocket. Next year, when that doesn’t happen, is it going to look bad that your numbers have gone down?

What’s important is not how many people visit your site, but what they do when they get there. 100,000 unique visitors to your website, just to find the phone number, isn’t a success.

Equally, 10,000 people trying but failing to find their bin collection date, isn’t so much a success as a massive inconvenience.

Measure success

What matters when delivering services online, is that you are delivering them successfully: that people can find, understand and use them.

Measuring the number of people who successfully achieve a task online is a start in understanding the success or otherwise of your online services. This has to be combined with the volumes on other channels. You might get 1,000 successful missed bin reports, but if the phones then get 10,000 it’s not as good as you thought.

Unsurprisingly, the Government Digital Service has some good detail on measuring transactions:

Two of the core measures are:

  • cost per transaction - when measured across the delivery of the service on all channels
  • digital take up – what percentage is online
    The aim is to make the first one come down while the second one goes up.

These are just two examples of measurements that can bring much more value to the council than site visits. During the development of the Liverpool City Council website, the question of how to measure was integral at every stage of development. As a result we could measure the rate at which users moved from the website to the other channels within the council. Crucially, we could focus in on which the areas of the site were generating contact.

Incidentally, in the first year after the launch of the new website, total visitor numbers dropped. However, the number of successful transactions on the website rose, while calls to the call centre dropped by 20% on some lines, and that is the bit that matters.

By shifting the conversation from the number of visits to a website, to how it is working alongside the rest of the council’s services, gives us numbers we can do something about. Not only that, but it helps to bring digital services into the core of service delivery, when all too often it’s still treated as an add-on, to be dealt with by the geeks in the corner.

_* Yes, central government wasn’t the first to attempt a customer-focused, task-driven website, but the example of certainly gives the idea more traction in many council offices.