Using machine learning to predict drivers of bounce and conversions on mobile


Earlier today, Daniel An and Pat Meenan from Google shared the results of a recent research project focused on uncovering what influences the bounce and conversion rates for e-commerce sites.

Using a machine learning model developed in collaboration with SOASTA, Daniel and Pat identified that the speed and performance of a website can significantly influence the bounce rate of an e-commerce site.  To put it simply: the slower and the more complex the page, the higher the bounce rate and the lower the conversion rate. This is consistent with several research studies and shows the promise that AMP adoption in e-commerce sites holds for doing business on the mobile web.


Source: Google/SOASTA Research, 2016.


To read more about Daniel and Pat’s research check out their article on Think with Google. And if you are curious to learn more about how your website is performing, we encourage you to run the same analysis using their open-sourced code.