Don’t miss our optimization tips
As optimizers we are always curious. Curious to find out why something is or isn’t working as expected to see if we can learn from it
Well, there is a tool called Hotjar which we’ve been using ever since it was still in beta phase and invites only. We have grown to love this tool and the insights we gather from it
These insights help us understand the customers /visitors of the websites we research better and helps us improve the usability, which in turn helps increase the conversions. Which is the goal of our CRO process
As a conversion optimization agency this tool gives us insights we would not want to miss. These insights help us to create improvements which we can then A/B test to validate that our solution is indeed an improvement
After working with this tool we started to wonder: Wouldn’t it be great if you could easily filter out all sessions which interacted with an A/B test to see if and how they are (inter)acting differently?
That’s when we started using this trick..
One of the tricks we have been using is to use the tagging function of hotjar when running A/B tests. Basically we add a tag to the variation(s) to be able to easily identify and filter the recordings of each variation in Hotjar. We’ve used this to easily find and analyse results of experiments but also as an additional way to triple check every experiment shortly after launch to be sure there are no bugs or unintended issues happening (You can never be to sure)
Note: Some A/B test tools, like Convert.com, Omniconvert and Optimizely now offer an integration which has an automated way of doing this. Since it’s just a short line of code you can still implement this manually as well if you choose to.
To get this to work simply add the following code on top of the code of each variation and you are good to go
hj('tagRecording', ['experiment_x_var_y']);
You can name the tag whatever you want and with how many tags you want as long as it is in this format
hj('tagRecording', ['tag1, tag2']);
IMPORTANT: If there is a chance that the code will be run before the Hotjar script has loaded, you must add this line of code just before:
window.hj=window.hj||function(){(hj.q=hj.q||[]).push(arguments)};
We’ve found it easy to use a simple naming convention for this using the same beginning every time and adding the experiment id’s and variation numbers numbers to it. This way you will be able to easily find, select and view recordings of the right experiment
This is an example of how you can filter those recordings with our naming convention in Hotjar
I am curious if you are using Hotjar also and what kinds of tricks you have found that make your work easier? Leave your comment below…