Using content groupings to visualise user journeys in Google Analytics.

We are frequently asked how users navigate through our website and it feels like this should be easy enough to answer using Google Analytics. However, knowing how users get from their landing page to their exit page can be a pretty complex question!

Why is this important?

By seeing use journeys across our site we can better understand the following (to list a few):

  • How users navigate between different types of content
  • What the most common sequence is for viewing content
  • Whether certain pieces of content are frequently viewed before or after each other
  • Relationships between content
  • How easily our target audiences get to key content
  • Areas of significant drop-off

Why is visualising user journeys difficult?

Visualising user journeys can be difficult for a multitude of reasons, starting with the fact that journeys can be incredibly complex.

User behaviour can be difficult to predict and often, if there is an unusual way of getting from A to B with weird and wonderful pathways, users will find it!

Added to the complexity is the fact that Google Analytics provides navigation data on a page-by-page basis.

Navigation summary report showing the pages viewed before and after the Biology department landing page.

This means we can identify pathways from one specific page to another, but wanting to understand user journeys through the whole site, or even a whole branch, can feel almost impossible using navigation summary reports.

Google Analytics’ other option is to provide a Behaviour Flow report. However, using this to spot patterns and pathways can be just as unwieldy due to the multitude of possible user journeys.

Unfiltered Behaviour Flow diagram in Google Analytics

What did we do to try and shed light on user journeys?

Given that the Behaviour Flow report feels like it fits the bill to answer all of our user journey questions we invested time in discovering why we weren’t finding it helpful.

Our two main issues appeared to be a lack of content groupings and a lack of user personas through which we can view a simplified version of the Behaviour Flow report.

Creating user personas:

We often refer to our website users as if they were one large cohort, asking questions like ‘where do our users go from here?’

However, the reality is that a prospective undergraduate student is going to go to a different destination than a staff member, or a member of our research community or even from a prospective postgraduate student.

Viewing our different audiences journeys as one dilutes the trends we would be able to spot if we created personas and instead asked the question ‘where do our current students go from here?’ for example.

We created rough personas for potential PG and UG students, current students and staff as a starting point. These were built as Google Analytics segments so that they could be overlaid on Behaviour Flow reports. They can also be refined over time if needed.

Creating content groupings:

In order to not get bogged down in the detail of individual pages, we looked at grouping our content using the feature called ‘content grouping’ in Google Analytics. This is a process by which we can categorise groups of pages under one name and compare different groups. For example, we can group all course pages together rather than looking at Biology, Chemistry and English course pages separately.

Google analytics all pages report.
Google Analytics ‘All pages’ report
Google analytics all pages report showing content groupings.
Google Analytics ‘All pages’ report with content grouping

We needed to group content with a great enough level of detail to identify steps that users take between content, but with enough breadth that we weren’t hindered by too many steps to visualise in the Behaviour Flow report.

We created content groupings based on either the template the page uses or where on the site it is found.

Applying this content grouping to the Behaviour Flow report means that we can now see pathways between the different categories of pages defined in the content group, as opposed to between individual pages.

The end result

Viewing the Behaviour Flow diagram with both content groupings and user personas in place gives us a much clearer picture of how users interact with our content:

Behaviour Flow diagram with page content and type content grouping, and prospective postgraduate student audience segment in place.

Since these amendments we have been able to identify that most prospective students visiting our department pages, i.e. https://www.york.ac.uk/history/ go on to view course pages i.e. https://www.york.ac.uk/study/undergraduate/courses/ba-history/ before going on to view other centrally managed study content.

This may sound like a simple observation but it was very difficult to say with any certainty before adding personas and content groupings to the Behaviour Flow diagram.

Between these new and enhanced configurations we can start to gain insight into all of the important questions listed above, like what are the relationships between our most valued content? Look out for more revelations and insight to come!

Published by

Katie Shearston

Search, Analytics and Digital Advertising Specialist at the University of York. I get to dive into data analytics, SEO and our digital marketing campaigns.

Leave a Reply

Your email address will not be published.