When analyzing the conversion data, if you have been doing A/B testing you should compare Notification Activated and Notification Filtered segments between each other. If you are using multiple notifications it is recommended to drive the data into a spreadsheet program (such as Excel or Sheets) and analyze the data there to get a better view of the data.

You can find more info about setting up A/B tests in here: Creating A/B tests

More detailed documentation on the conversion data and how it is gathered can be found here: Notification data and logic

When you have gotten the data to your spreadsheet program and are looking at specific notification in relation to specific conversion, you should compare Notification Filtered to Notification Activated.

If you are unsure about if you should make decisions based on the data, you can test the statistical significance of the data. A simple way which can help you with this if you are not a data scientist is to put the data to a A/B Testing Statistically Significance Calculator like this one!

Just insert the number of users in the Notification Activated segment and conversions from the segment. Then, add the same numbers from the Notification Filtered segment and you will see how statistically significant the data is. If the certainty is low (usually below 95%), you should wait and gather more data to make conclusions. If the certainty is more than 95%, you can conclude that the notification is most likely working.

Obviously this guide is very simplified and if your data set is big, you should do a more thorough analysis for the data. Nonetheless, you can definitely get started with something as simple as this.

Did this answer your question?