Analyze Your Testing Results – Conversion Wednesday



In my last Daily Concept, I discussed why testing was a long-term process. Today, I’ll explain why this long-term mindset is important when calculating your testing results.

I’ll start with a generic example:

Let’s say you have completed a test that compared a green Add-To-Cart button to an orange Add-To-Cart button. The orange one has increased conversion rates by 5% and revenue by 3% over three weeks. You can see the data below.

—- Sign up for our latest free webinar!.Harnessing the Profitability of E-Mail Marketing for e-Commerce Stores —

Remember that % lift is the percentage increase over the control. In this case, the metric we are looking at is revenue.

This test made $6,000 more dollars in three weeks than the control!

I’m going to walk through the results from the previous slide and explain how to read these numbers. Here are the numbers explained:

  • The metric we focus on is revenue. You can increase conversion rate by giving everything away for free, but it won’t make you more money. Revenue shows actual, final dollar gains for your controlled test experience.
  • The test didn’t just make $6,000. Remember you split traffic in half during testing (half went to the control recipe). This suggests a gain around $12,000 if all traffic had been sent to the recipe with the orange button. Which leads to…
  • This was only three weeks of testing. You need to add that 3% lift to your annual revenue to estimate the potential gain from this test. Remember that any change you make to your site as a result of testing continues to deliver month after month once the change is made.

So, lets annualize those results based on the hypothetical button-test example:

If your store makes $2 million a year online, a 3% increase in revenue would be a potential $60,000 more a year just by changing your green ATC button to an orange one! That continues to contribute the next year too!

This long-term mindset is key to understanding the value of testing:

  • Small gains in key metrics can lead to big results when compounded over the entire year.
  • You don’t only get the benefit of the test while the test is running. It continues after the change is made to your website.

So, be patient and always be thinking long-term when you are testing on your website!