Just because a webpage is popular doesn’t mean that it’s useful. In fact, it could be getting in the way of a more useful page.
“We’ve recently decided to remove Basic rights at work, the most visited advice page on our website,” Hannah Horton from the UK’s Citizen Advice recently wrote. “It gets 70,000 visits a month and is in the top results for lots of queries about work problems on Google.”
If an organization was judging web success based on volume of traffic they would certainly not remove their most popular page. However, Citizen’s Advice does not measure success that way. It measures success based on how successful people are at solving problems.
The page wasn’t working. Firstly, the data told them that people were not searching for this page but were rather searching for much more specific things such as ‘boss doesn’t give me breaks at work’ or ‘how much a week do I get if I go on the sick at work’. The Basic rights page is an overview. It doesn’t answer specific questions like the ones above. There are other more detailed pages that answer these kinds of questions. It had become a search dirty magnet; drawing people in the wrong direction, away from the specialized pages that were better able to answer the questions.
Web traffic statistics can often be quite misleading, particularly on mature websites. Pages can become popular for all sorts of reasons. Lots of overlap can occur. There is a very specific feature in Microsoft SharePoint called Excel Services. It used to offer training and lots of people used to go there to sign up for training thinking that they would be getting training in how to use Microsoft Excel.
Digital must be measured based on customer outcomes. Traffic, visits, time spent, page views; these are not outcomes. In the offline world, when you say you visited the dentist it implies you got something done with your teeth. Would a dentist measure success based on the number of visitors they had asking for directions to the tourist office, or who had mistaken the dentist with the butcher next door?
Another reason the Basic’s page was seen to not be working was because of the very high bounce rate, which was around 70%. According to Google, a bounce rate is “the percentage of visitors to a particular website who navigate away from the site after viewing only one page.” A bounce rate can be a good thing if the page you visit contains the answer you need, but the Basic’s page did not contain specific answers. Added to this, analysis showed that when people arrived at the page, they rarely scrolled down, instead leaving quickly.
The feedback from visitors also indicated problems:
“There appears to be no information about giving notice to your employer.”
“The paragraph doesn’t tell the client much about what action they can take over bullying.”
Suppose you were responsible for GOV.UK and you started seeing a decline in visitor statistics. For example, the Passport fees page numbers had dropped significantly. That’s bad, isn’t it? Well, no. If you search for “UK passport fees” on Google you now get the answer in the search results. So, things are actually getting better because now you are answering people’s questions even faster. Measure success based on the outcome for the customer.