So, you’ve got a strong hypothesis, the design has been approved and the development of the test variations are looking great. But then you launch the test, and after a couple of weeks, you notice that your metrics aren’t showing the right amount of visitors. It’s confusing, right? You might even start challenging your research-backed hypothesis, only to later find out that the metrics weren’t correctly implemented.
Now, you’ve lost two weeks’ worth of testing that you’ll never get back. Not ideal, obviously. Well, this is a major mistake that can be easily avoided, using just four simple steps.
Step1: Set aside time for metrics setup
When you’re estimating how long it’ll take to build the test, make sure that you set aside specific time for implementing the right metrics.
Step 2: Assess and define the metrics you want to track
Ensure that you’ve got the right metrics as part of your test plan or brief that you want to track. Define them clearly;, regardless of whether it’s a simple click goal, a pageview goal or a goal that’ll only be tracked in certain conditions (for example, ‘only click ATB buttons when the user selects a quantity’).
Step 3: Check whether the metrics are firing correctly
When you’re QAing your tests, make sure you interact with the pages to see if the metrics are firing as they should. There’re a few ways of doing this. Some A/B testing tools provide a preview option that also shows the metrics that’re being tracked. You can also check that the right metrics are being fired within the console/network tab of your browser. Ideally, using a separate solutions engineer to make this check to find issues with any setup will be the best way to ensure there’s no problems.
Step 4: Check if the data is coming through on the reports
Finally, launch the experiment internally (via QA Cookie or by targeting specific IP addresses) and interact with page elements that should fire the metrics. Check within your report to see if those metrics are correctly showing on the results page as expected. To us, this is the most important step to ensure that all your metrics are working and tracking properly for your tests.
You might be thinking that all this will add time to your overall testing process, but when you think about the consequences of failing to setup the right metrics, the extra time to get things right is much better than how much testing time you’d lose if you don’t!