Live More new Guides on the way! Get notified by signing up to CAST's newsletter..

Contributed by

Blood Cancer UK, is a UK-based charity dedicated to funding research into all blood cancers including leukaemia, lymphoma and myeloma, as well as offering information and support to blood cancer patients.

Use this Guide to learn how to identify potential issues with your website, investigate it with Hotjar and make a case for changing it. You could also start by using Hotjar to identify potential issues and then do usability testing to understand them better and generate solutions. Use Google Analytics data to support or challenge your insights. 

Steps to using Hotjar to measure, test and improve your website

You may already think that your website isn’t working as it should. Maybe pages and features have been added over time and content has not been regularly reviewed.

Ask your website or comms lead to find out what data they have about site performance. Speak to key stakeholders to see what they know about the content. Write down any problems you hear about. 

You can also use Google Analytics to build up a picture of how people are using your website.

Bring all the information together and decide what issues you want to prioritise and how you might measure each one. For example, if you have pages with very long content and you’ve heard that people aren’t finding information, you may look at scroll depth (how far people scroll), or whether people click on anchor links and key calls to action. 

Blood Cancer UK had realised that people had difficulty navigating their website and understanding text links.

People weren’t finding the information they were looking for. This was often because it was near the bottom of a page, or they did not recognise calls to action in anchor links and buttons.

Use a tool like Hotjar to find data that gives you insight into your issues. Look for data for pages you think people struggle with. Write a hypothesis for why you think each problem is happening.

Hotjar's heatmap feature can show problems in a visual way, which may be useful if you need to explain to stakeholders why you need to do usability testing or interviews. 

If you want to ask something specific to a lot of people, you can create a survey using Hotjar Ask and embed it on your site. You could ask “How did you find this page to navigate on a scale of 1 to 5?”. Results can help you prioritise which pages you need to focus on to improve navigation.

You can also do one-to-one interviews and usability testing to get different types of data.

Demographic data from Google Analytics can help you pinpoint who you should do usability testing with.

If you find Hotjar does not give you enough data to build a hypothesis, you could conduct insight interviews to explore:

  • people’s experience and use of the website, likes and dislikes

  • what other websites they use.

Then you can configure Hotjar to investigate these experiences further or run usability interviews (see Step 3). 

2 annotated Hotjar visualisations showing scroll depth and most clicked links on a page about getting support

Blood Cancer UK's digital team use Hotjar and Google Analytics data together to measure performance and explore issues on their website. This is because Hotjar only tracks about 7% of their traffic (due to its optional cookie settings). They have found that Google Analytics data generally validates what they observe from Hotjar. However, they would prefer it if 90% of their traffic consented to Hotjar's cookies. This would generate more robust insights on the site's least visited pages.

The team examined data from Hotjar and Google Analytics. The data suggested that people might have difficulty finding pages on the site related to their blood cancer condition. The team hypothesised that the site's information architecture (how site pages are organised, structured and labelled) was causing the problem, rather than the quality of the content on the pages themselves.

It’s important to combine data from tools like Hotjar with qualitative data from usability testing and interviews. This is because analytics tools are good at showing ‘what’ people do on a website, but not ‘why’.

If you have a clear idea of what the issues are, start usability testing specific content with 5 to 6 people. Try to get a mix of ‘warm’ (already engaged) and ‘cold’ (not yet engaged) participants.

Ask people to find something on the website, or about the organisation on Google, to see if they end up on the right page.

You can observe things like:

  • “What are people doing when they get to that page?”

  • “Are they clicking on the links at the bottom?”

  • “Are they going back to the page before?”

  • “Where are people not finding information they need?”

You can also go back to Hotjar to get a further understanding of how people are exploring these pages. For example, if you see in testing that people are not scrolling a page, you can look at that page's scroll depth data on Hotjar and compare it with your testing observations.

For both Hotjar and usability testing, make sure you compare desktop and mobile versions of the site.

Blood Cancer UK conducted usability testing with 9 people with blood cancer, asking them to find information about their condition. 6 people were ‘warm’. 3 were ‘cold’, found through paid-for social media marketing. They received £50 per session.

Blood Cancer UK found that people generally could not find the information. This clearly showed the need to improve findability through improving the site's information architecture.

You can use visual tools such as Miro or Powerpoint to:

  • bring together your findings from user research and Hotjar (and other data analysis tools)

  • present your findings to stakeholders.

Use the Hotjar's heatmaps and other visualisation tools along with your other data and insights to help you tell a story about the problems that people have with your site. For example, you could show Hotjar screenshots and heat maps, together with insights from user interviews.

2 Hotjar visualisations showing varying scroll depth on different pages. 1 page has no anchor links but the other does.

Blood Cancer UK saturated a Miro board with:

  • insights from interviews on ‘sticky notes’

  • screenshots from Hotjar and Google Analytics

  • insights from previous rounds of research.

They annotated information to provide context and pulled together themes, such as:

  • scroll depth

  • how to help people navigate through long content.

Blood Cancer UK then presents the information in a Powerpoint slideshow for non-digital stakeholders.

They have found that showing data like this play a critical role in:

  • persuading stakeholders of the impact of problems

  • showing the impact of the changes that have been made, by comparing it to previous iterations.

Stakeholders have appreciated Hotjar's data because it’s very visual:

“If I was going to choose one visual tool I would choose Hotjar because it’s very visual and the data can be so stark. It helps me quickly see the problem. It strengthens your argument, for example, 'we need to shorten this copy'. It will show what it looks like on the old page and on the new page.” - Ben Sykes, UX and Innovation Lead, Blood Cancer UK

In an ideation session, present the problem and insights then brainstorm a solution. Include stakeholders. You may need to understand some digital terms and any technological constraints.

Once you're generated a collection of ideas you may need to review it in context of your data and research to understand if it is viable and feasible.

In an ideation session, Blood Cancer UK start by presenting everything they have learned from Hotjar and research. The session is usually between 2 and 3 hours long.

They turn each problem into a “how might we” statement. For example: “how might we [help people find the right information on their condition]?” Then they ideate (mindstorm) a solution.

Sometimes they get everyone to draw or explain a low fidelity solution on paper. Then they vote on the most promising solution and have an open discussion about it.

Either change your existing content based on your findings, or create a prototype if you’ve identified a new type of page or feature.

You can use Hotjar and Google Analytics data to test your solution's impact, for example:

  • whether people are navigating as expected or intended

  • to measure other problems you discovered during your research.

Based on their data and insights, Blood Cancer UK decided to build a new homepage and to try out different content 'tiles' in different positions on the page to see what was effective. They set-up dashboards in Google Analytics Data Studio and Hotjar to monitor differences in user behaviour across positions.

In the first few months of tracking, they found that a tile at the top of the homepage that had seemed useful during testing wasn’t being clicked on. The title of the tile was “I need to speak to someone about blood cancer”, which linked to a page listing support options.

Blood Cancer UK were confident that the page listing support options was a popular page because of its high pageviews. However, they realised that people were not generally looking for that information at the top of the homepage. So they replace the tile with one linking to a page “Information about our research”. They found an increase in clicks on the tile over time.

They also made the homepage shorter, because Hotjar showed that people were not scrolling to the bottom.

Further information