Traditional analytics software like Google Analytics, Adobe Omniture, or those I’ve introduced earlier here such as Mixpanel, Heap Analytics – they are all giving you hints on what happened and what you should be looking at: e.g. the top landing pages with high bounce rate; or the top converting keywords; but it won’t tell you how exactly your users interact with your pages: whether people has scrolled all the way down to the end of the article; what are the areas that drive people attention on a page and most importantly, how does people FEEL about your pages. In this post I want to introduce a new software called hotjar, still under closed beta testing, which exists to answer these unanswered questions.
Pricing Scheme
Similar to my earlier posts I start with pricing first. Hotjar is still under beta testing, according to the site it will charge you $29 per month for unlimited usage. Right now if you invite 5 of your friends to join hotjar, you can even get a 6 months of free usage! (I did invited 5 friends now).
Installing hotjar
So again like installing Google Analytics, you will have to put the tracking code into all your pages. Simply dump that into Google Tag Manager and you’re good to go.
After you’ve installed the tracking code you can now start configuring the different user behaviour reports within hotjar. Let’s go through some of the killer features that I like one by one.
Heatmap
We are always curious about whether users actually finished reading our content or not. I’ve talked about how to use Google Analytics to track whether people has scrolled to a particular section of a page or not, but a even better visualisation would be a heatmap like the following:
So from here the “hotter” (more red colours) region, the more visitors actually did scrolled through that section. And there’re clicks heatmap or even mouse cursor movement heatmap. All these give you several insights to think about:
- How many of my visitors did actually find my content relevant (do they just left after the intro paragraph?)
- Where’s the “average fold” (the screen height) of my users’ browsers? Should I be putting important content above the average fold to make it more prominent?
- Do visitors find my content too short or too long (how many % of users did scroll to the end of the article?)
- Which is the area that received most click – so that maybe I should put my CTA over that area?
Visitor Playback
Have you ever imagined that you can sit right next to all your site’s users, watching how exactly they interact with your pages one by one? For sure you can do focus group or invite people for usability test; but how about if you can do this LIVE with your actual site’s users?
Hotjar offers a feature called “visitor playback”. It allows you to play every visitors’ visit like playing a recorded video, just like when you sit next to him to see how he uses the site:
For sure it would be a nightmare to go through every single session one by one when you have a large number of visitors. The best part is you can first filter the visits with some criteria such as duration, landing page, countries, browsers, etc. Some of the analysis that you can do:
- Filter visits with particular browsers / OS; especially if you know from other analytics package that certain browsers / OS combination show lower engagement / conversion rate
- Get the site average time spent on page, then filter visits with duration longer than the average to see what are the pages that visitors are seeing, or what are the things that visitors are doing during these visits
Feedback Polls
Most of the analytics software like Google Analytics or even Omniture Site Catalyst are reporting quantitative metrics – visits, page views, bounce rate. But those data would only show us where thing goes wrong but not what might have happened. As always a better way would be to really get qualitative data – like user comments and feedback, to truly understand what drives the user behaviour.
Hotjar allows you to create feedback polls to collect qualitative data from your site. You can choose to trigger a short survey based on several conditions:
After you’ve set up the poll, your end users will see the feedback form, and you can start collecting their data. For example, I’ve asked a question what kind of content that they’d like on my blog:
And here are the replies:
For instance here, nobody actually wants to hear me introducing new software (which is exactly what I am doing here :(), so I may prepare more GTM usage tips to meet my end users’ need.
Summary
One of the most frequently asked question from my client and my colleague is: can we use web analytics data to determine whether we should keep that button / put that icon on that location / make that CTA button larger. I think apart from conducting A/B testing, user behaviour tools like hotjar may give you some insights on how visitors interact with your site.
Hotjar is running under closed beta now and will be releasing on Mar 2015. From the meantime I highly recommend you to queue up for the hotjar trial and get it a try!