Interpreting Your Data: How To Understand Calculated Metrics

Welcome to the second post in our "Interpreting Your Data" series for affiliate advertisers, brought to you by Google Affiliate Network analysts. View all posts in the series here. 

Calculated metrics provide great insight into how specific variables are performing. A few of these commonly used metrics are conversion rate, click through rate, and average order value.
In this post, we will use examples to discuss how you can use calculated metrics to interpret your affiliate program data. 

Reviewing performance data using calculated metrics:
You’re in the Google Affiliate Network interface, reviewing performance reports and notice your conversion rate has increased. You may think -- “That’s great news!” Or, is it?

Before celebrating, you’ll need to ask a few more questions and take a deeper look to determine whether an increase in conversion rate is really a positive indicator for your overall program.

This is because conversion rate is a calculated metric that is generated by combining two other metrics -- in this case, conversions divided by clicks. Whenever you use a calculated metric, you should understand what’s happening to both of the underlying metrics to to determine an overall trend.

Example 1:

• In period 1, you have 33 conversions and 100 clicks, so your conversion rate is 33%.
• In period 2, you have 25 conversions and 50 clicks, so your conversion rate is 50%.

As you can see, the conversion rate goes up in period 2. Is that good? There are fewer conversions in period 2 and fewer clicks. You should investigate why clicks went down before you can determine if this is a positive or negative indicator.

Understanding metric composition:
With an affiliate program, you should also consider the composition of each metric. That is, the composition of all of the publishers that are driving clicks and conversions. Each individual publisher composes one part of each metric.If you see conversion rates go down, you may think that this is negative indicator. To determine the significance, you should check if the conversion rate is down for every publisher and if you added any new publishers. 

Example 2:

• Publisher X had 33 conversions from 100 clicks totaling a 33% conversion rate. 
• Then you add publisher Y, who generates 200 clicks, 20 conversions totaling 10% conversion rate.
• These two publisher combined generate 53 conversions on 300 clicks, for a conversion rate of 26.5%. 
• The combined conversion rate is down, but that’s only because you added a publisher that drives more clicks at a lower than average conversion rate. 
• You still have 33% conversion from publisher Y that drove 100 clicks.

In this example, it’s important to look more closely at the new publisher. It’s possible they have a lower conversion rate due to their business model. Adding this publisher to the program changes your overall number.

In summary, when you’re looking at the movement of a calculated metric (such as conversion rate and average order value) you shouldn’t stop there. To understand the shift, you should do a bit more investigating to find out why that metric moved.

For additional resources on analyzing your program, please see this analysis checklist help center article. 

Posted by Dan Filowitz, Manager of Affiliate Operations