Experiment with more "explanatory" empty states to improve upselling
Most of our empty states are a combination of an illustration, a title and a short paragraph that describes the feature that the user doesn't have access to—some of these empty states are there simply because there's nothing to show, others are for locked features that are available in higher tiers. This issue focuses on the latter.
Proposal
Especially for the locked features (available in higher tiers), we should experiment with what I call "more explanatory" empty states. Instead of just writing a short paragraph about what the user could do if they had access to the feature we should show them. HotJar has great examples of this:
They use a combination of screenshots of features in use and text description. They have a couple of "slides" for each of the features so the users can explore the possibilities of the feature. I'm a user of HotJar and these empty states really sparked my interest so I'd say that in their case it's really working. Can we use this in GitLab?
Hypothesis:
Empty states that present locked features in more details will improve the conversion to paid/higher tiers.
Suggested steps:
- come up with a list of features we think are great upselling possibilities
- Mockup a few possible solutions (prototypes) and get some qualitative feedback.
- The solution that performed the best in the previous step proceeds to an experiment (one feature that we agree has the most potential)
- If the hypothesis is validated we should refine the approach and define a framework.
- Once we have a framework we can continue to gradually introduce these "explanatory" empty states throughout GitLab, based a priority defined by the potential for upselling. Each introduction of this "explanatory" empty state could/should be an experiment on its own.