Forum Discussion
1 Reply
- DavidOStrategist II
It was an onboarding experiment we ran in December. We uncovered some data that suggested our users were more likely to export their first video, a very high-value action for us if they completed at least 5 actions in the video editor. Coincidence or causality? Only one way to find out.
We created a canvas with an action path that measured a user's first action in our editor over a 6-minute window, the time we knew it took for 80% of our users to complete their first action after registration. The action they took was ranked and from here they were sent down separate paths to a series of second action paths that checked for any other key actions they took over a 6-hour window. We have found a 6-hour delay before sending an onboarding email performed better for us from previous experiments.
From here each user received a personalised email built from content blocks and catalog items that suggested two other editor actions they may have yet to try. There were 13 personalised variations in this experiment but by using catalogs and action paths, it could easily be expanded to hundreds. During the 4-week experiment window, we saw a 10.8% change over our no messaging control which was 5.1% more than our previous onboarding flow, with 100% confidence.
It was our first big journey into super personalised messaging combining action paths, content cards and catalogs, so I can wholly endorse those. It was a great success and we will look at using this concept more in the future.
Related Content
- 4 months ago
- 4 months ago
- 11 months ago
- 7 months ago