Today, Parse is adding a new feature to its platform, Parse Push Experiments. The new tool now lets you easily create A/B tests for push notifications so you can study the analytics data and see what content or timing is most effective.
To try it, you just need to open up Parse’s push Web console and then allocate some users into a test group. You can then send each group a different message.
Afterwards, you’ll see the push console give you real time metrics about the two test groups, including a statistical confidence value on any differences in performance.
Finally, you can elect to send the better performing version of the notification to all users once you’re done testing.
In addition to A/B testing content, you can also test when it’s best to send notifications, using parameters such a location and/or time. You could, for instance, test sending a push notification for a coupon when a user is close to a restaurant.
Push is available to any developers on Parse for free – you need simply log in to the web console and select the “Use A/B Testing” from the push composer page.
A/B testing is common for things like email marketing, but less so for push notifications. It’s a significant move by Parse – more users are becoming accustomed to the immediacy of having information pushed to them rather than having to look for it or refresh emails.
But push notifications also run a greater risk of being more annoying than informative. The new tools could make it easier for marketing teams to provide effective and engaging notifications.