I’ve recently been analysing how users navigate their way around GOV.UK and want to share the findings.
There were more than 56 million visits to GOV.UK in September 2014, a new monthly record. My task was to summarise where users landed in the site and the extent to which they browsed to other sections.
Defining the sections
The vast majority of sections on GOV.UK can be determined by the top level URL structure or from the page title. For example, mainstream browse pages all sit in the folder “/browse/” and detailed guides are labelled in their page titles. It can be reasonably assumed that most of the remaining pages belong to mainstream content (which currently don’t have an identifier in the URL or page title).
I set up custom filters in Google Analytics which aggregated all the pages within each section to allow this analysis to be done.
Navigation heatmaps
I wanted to condense the user activity in the month down to one page summaries. By cleaning and rearranging the data I produced heatmaps which showed the relative navigational relationships between the different sections of the site.
The heatmaps represent every navigation made on the site in September 2014 from one section to another. They include the instances where users didn’t navigate anywhere (known as ‘bounces’) or where the next navigation was to leave the site. I’ve also included instances where users’ next navigation was to a page within the same section.
For example, if a user landed on the homepage, then went to a browse page, then a mainstream page before exiting, the visit will be represented as:
- one navigation from the homepage to the browse pages section
- one navigation from the browse pages section to the mainstream section
- one navigation from the mainstream section to exiting the site
Note that due to the way the underlying data is filtered in Google Analytics, the numbers will not be exactly correct but a very close representation. Note also that at present it’s not possible to break down exits made due to a click on an external link from the site.
Heatmap by numbers
Click on the image to open the full version.
This heatmap shows section navigations by numbers. The sections on the left show where users navigated from, the ones at the top show where they went to. The corresponding cell shows the approximate number of times this occurred in September, with the colour portraying the relative frequency. The ‘hotter’ or darker the colour, the more times the navigation took place.
For example, there were more than 6 million times when users went from a mainstream browse page to a mainstream content page.
Heatmap by percentages
Click on the image to open the full version.
This second heatmap breaks the data down by the percentage of navigations made from each section to provide another interesting perspective of the data. The heatmap was achieved by totalling all the navigations in a row and then calculating the percentage for each section.
What the data tells us
Visualising the data in this manner reveals some intriguing insights:
- there are significant instances where users on the homepage seem to be clicking on the GOV.UK logo, perhaps thinking it links to another page
- in around 50% of all visits, users only look at a single page on GOV.UK and leave; this is not necessarily a bad thing as users could have found exactly what they needed
- there is a significantly higher tendency for users to go to a browse page from the homepage than to use site search
- 62% of navigations from a help page go to a mainstream page with only 16% leaving the site; this is encouraging as it suggests that the help pages are useful to users in directing them to content
- 28% of navigations made from a search results page are to another search page; this is due to users refining their search terms, using the filters or viewing the next page of results
Next steps
If these heatmaps generate enough interest, we’ll continue iterating them in the future as GOV.UK develops further. In the meantime, if you have any suggestions for improvement please leave a comment.
