There is a bit of a theme to this week’s articles – how public bodies’ websites are performing. Here we see one organisation doing a fantastic job, and one organisation doing a terrible job. Then there is a report on Unistats, which is more directly relevant to us.
I have waxed poetic about Gov.uk before, and it continues to grow in influence and credibility. Last week the Guardian went behind the scenes to find out how the Government Digital Service works, and the thinking behind Gov.uk. This is highly recommended viewing.
I was interested to see the evolution of the Gov.uk homepage (3.43 to 3.50), which suggests that they attempted several different graphics-heavy designs before settling on the stripped-back, minimalist, information-focused website we see today.
It’s designed around the user, not around the needs or the desires of the government – and that’s tended to be the case in the past… Prettifying is not what it’s about. I’m thrilled that Gov.uk won this incredibly prestigious design award [the Design of the Year Award], but that wasn’t about what it looks like.
It’s about making sure that people will be using this, and they will need to get something done, or learn something. Rather than trying to make stuff look pretty for the sake of it, trying to make things easy to use.
The Office for National Statistics website is so notoriously bad that it has been featured on Radio 4’s More or Less. The ONS deputy director responsible for digital publishing, Laura Dewis, was subjected to a grilling from Tim Harford.
According to Laura Dewis, the reason the ONS website’s search function is so bad is because they listened to user feedback, which suggested that users always wanted the most recent information. What the ONS website designers failed to realise was that what users actually want is the most recent relevant information.
This is a cautionary tale. What users say they want is not always what they actually want. User feedback is only part of the mix. We should also use analytics data, which provide hard facts about what users are actually doing. Designers should also use their judgement to work out what would be the best solution, rather than slavishly following user feedback in this manner without applying any critical thinking.
Laura Dewis also revealed that the ONS focused on making their website easy for ONS staff members to update, at the expense of the user experience for the end users of the website. We should always remember that the website is for the user, not for us.
A report from HEFCE evaluating the usage of the Unistats website so far. There are a few interesting stats:
Between its September 2012 launch and (the start of?) May 2013, the Unistats website received 3.8 million pageviews from 175,000 unique visitors. In that period, Unistats had 22 pageviews per unique visitor. This suggests to me that Unistats has comparatively few users, but those users are very heavy users.
73% of visitors to the Unistats website are direct – i.e. they are typing the URL into the address bar, or using a bookmark, rather than coming from other websites or a search engine. This figure seems very high. For most websites, the majority of traffic comes from search engines.
A section about the performance of the widgets is on page 10 (page 25 of the PDF). There, concerns are raised about the visibility and usability of the widgets. It appears as though users are mistaking the widgets for adverts.
(Studies show that many users ignore anything that looks like an advert, even if it is not actually an advert. This phenomenon is known as banner blindness.)