in Danish
A client came to us with a problem. He and his colleagues were concerned that the page load speed on their website was unacceptably high and that customers as a result ended up buying elsewhere. They needed to investigate this further and wanted to know the exact critical threshold as to when load speeds had a diminishing effect on conversion rates and revenue. This would help them determine whether focusing on load speed optimization would be a good business case and what it would require of them in terms of investments and resources.
Solution
In order to validate their hypothesis and pinpoint the root cause, we took a closer look at their Adobe Analytics implementation. Since they had already implemented the ‘performanceTiming’ plugin, we knew that we could use Python to analyze the data as every single page load on the website was timed. With Python we would be able to estimate the average page load speed throughout their users’ visits and even identify whether the intended journey was being disrupted by pages loading slowly.
Result
Based on our analysis, the client was able to confirm their hypothesis – high page load speeds resulted in a severe loss of revenue. The pattern we discovered was that load speeds started to have a severe impact on conversion rates after a certain threshold. Thus, in terms of reducing load speeds we could identify the optimal trade-off between investments in load speed optimization and potential revenue uplift to maximize profit.
Faced with a similar problem?
Are you worried that you are in a similar situation as our client or simply interested in knowing how to make predictions about your visitors? If so, you need to model the data outside your analytics tool. Luckily, our analytics and optimization expert Troels Moltsen has created a guide that shows you exactly how this is done.
If you are interested, you can download his new white paper “How to Extract Hit Level Data from Adobe Analytics, Dump them in Amazon S3 and Use them for ML and AI-analysis with Python” by filling in the form on the right. It contains everything you need to know, and it’s free of charge.
in Danish
