Case Study: Optimising for Visitors Coming Via Google Shopping

This fashion brand gets high traffic from Google Shopping, but this traffic source also has a very high bounce rate. The experience Google Shopping visitors received was not personalised; they experienced the same PDP as everyone else.

Hypothesis and Psychological Technique Applied

We believed that the high bounce rate was mainly because the product page they arrived at through Google Shopping was not exactly what users were looking for. However, the site did have other products that were likely to interest them, if only they browsed through a bit more.

Therefore, we needed to encourage users to stay on the site and explore more products. We believed that by showing them other related products on the PDP they’ve arrived at, we will be able to get a foot in the users’ door and they will browse through these other products.

Initially, we ran an experiment in which we showed users a pop-up to prevent browse abandonment. However, this experiment was not successful on Mobile and Desktop, but it was on Tablet (showing a significant increase in UCR and reduction in bounce rates).

We created another Variation where we used a pre-abandonment bar on top of the page instead of a pop-up.

Experiment

For desktop, the pre-abandonment bar was shown based on exit intent. We had seen in GA that users generally left the pages around the 15-second mark, which led us to create 3 Variations for mobile and tablet: the bar was shown at 10 seconds in Variation 1, 15 seconds in Variation 2, and 20 seconds in Variation 3.

Google Shopping pre-abandonment bar experiment

Desktop Variation

Google Shopping pre-abandonment bar experiment

Mobile Variation

Mobile Variation

Results

From the observed data, we were able to see that Variation 1, which triggered at 10 seconds, showed a high probability of being better than the Control on all 3 devices, delivering higher user conversion rates and lower bounce rates.

Learnings

Sometimes, a hypothesis may be correct, but it might take one or more attempts to get the design execution right.

Our Clients