Data Interpretation – Step 4 of Web Analytics Process

Reading time: 5 minutes

Last week, we have seen with the article “Data Analysis – Step 3 of Web Analytics Process”, how to analyse the standard metrics provided by your web analytics tool provider, and how to display the data that will help you visualize your KPIs, and measure your achievement. Today, we’re moving to the 4th and last step of the Web Analytics process, the Data Interpretation. We’ll first see a few tips and best practices on how to execute Data Interpretation correctly, and then focus on how to test our interpretation with AB Testing.

How to « Data Interpretation »

One of the first thing that is important when trying to interpret data, is to focus on the data related to our KPIs. Your website generates a lot of data, related or not to each other, and you can easily lose your main objective. Stay focused on what really matters to you.

Second, OBJECTIVITY. Don’t over interpret as you want it to be. It is easy to think that some data translate what you « thought » but, be objective. You want to know what’s really happening, you don’t want to try to confirm by any mean what you thought.

Third, don’t be satisfied with the « What is happening? », look for the “Why is it happening?”. For example, a high bounce rate means that a lot of people leave your website without having done anything. This is the “what” is happening. The best to do is to get why do people leave your website. Is the loading time to high? Is your content really related to what your audience is looking for? Those are the real questions to ask. So now the matter is, how to answer those questions? One of the best practice is to analyze multiple metrics together. One metric alone can mean something, but coupled with another one, it can help us reduce the number of interpretation possible.

Last, make sure you don’t put all your eggs in the same basket. There are many dimensions, such as the source from where your audience is from, the geolocation, the time… And you must know that your metrics can differ depending on the dimension value applied. For example, your bounce rate of people that landed on your website through Google Adwords can be lower than the bounce rate of the people that landed on your website organically.

The Role of AB Testing

Once you’re done with the Data Interpretation, you now have to turn this interpretation into actions. Most of the people are directly optimizing their websites, thinking that it will work straight away. Unfortunately, the interpretation remains an interpretation, so it is never 100% sure that it has been well interpreted. The best way to test your interpretation is to do it via AB testing. AB testing allows you to compare your original web page performance, vs. the same web page with modifications (issued from your interpretation). You can then see, in real time, which page performs the best, and then decide which version to use.

The use of the AB testing is now a great advantage since you can compare multiple pages together at the same time (ABCD..Testing), and also run another testing on other pages of your website. There are many tools that provide this service, and they are usually pretty easy to use (usually, a single JavaScript tag is needed on your website)
Well, during the last 4 weeks, we have studied in detail the each of the 4 steps of the Web Analytics process. We focused on the importance of each of these steps, and gave you best practices and tips in order to facilitate your global analysis. If you follow each of these steps correctly, you will be able to get the most of your data, and drastically increase your ROI.




Stay updated! Join our newsletter for webanalysis tips