How a single hover effect reduced our bounce rate by 8.4%
Bounce rate is one of our most difficult metrics to move at Rome2rio and as such has often taken a back seat in our decision-making around which parts of the site to pursue experiments on. The following experiment came about as a bit of an accident after a really passionate team meeting about the future of our landing pages.
Bounce Rate is the percentage of single-page sessions (i.e. sessions in which the person left your site from the entrance page without interacting with the page).
The meeting
Last month our content team organised a meeting to discuss how we were going to group bus providers that provided similar services on Rome2rio. We ended up veering off topic into a discussion about some more fundamental feature changes we could make to the product. We were talking about the hover over the different trip options when somebody mentioned that we could implement a hover effect which changes the route shown on the map instead of just the background colour of the item being hovered over. It was a really easy change so after the meeting I got back to my desk and just implemented it, not thinking about it too much.
It took us from this:
…to this:
Spot the difference? The only change we made was that we now drew the active route on the map when the user hovered over it. After our weekly release, we noticed a sudden drop in bounce rate — we weren’t 100% sure we could attribute it to the hover so we rolled back the change and ran it as an experiment.
The experiment
We ran the experiment on 622,000 desktop users for a week. That would see just over 300,000 users being shown each variant.
Running the experiment confirmed our suspicions. Our bounce rate decreased by 8.4% for the experiment group. The users who received the new hover effect were now 8.4% more likely to stay on the page and browse to a different part of the website, which inevitably had a flow on effect to other revenue streams.
Interestingly this didn’t immediately translate to an increase in conversions on our flight bookings suggesting that the visitors who stayed were more likely to be in an anticipation or destination phase of their journey rather than the preparation stage. However it’s not all about booking conversions — the fact that more users are motivated to explore our product is a really great win for us and hopefully, this translates to an increase in repeat users in the future.
