[Engineering] Experiment: Test SAST entry points
Experiment Summary
As an initial experiment in support of #322 (closed), we'd like to test out 2 entry points to determine which is more effective at driving traffic before we build on the rest of the walkthrough.
To do this, we'll place 2 CTAs on the projects page, one as a banner and one as a tooltip, and measure the clicks that they receive, which will lead the user to documentation to start.
The UX issue is #374 (closed)
Design
Assignment will be at the namespace level and will be sticky.
Rollout strategy
In order to track the success of this experiment, we will enroll new signups (namespaces) into either the control or experiment groups, tracked in the experiments_users table with a unique experiment_id.
We will enroll 34% of traffic into the control, 33% into candidate 1, 33% into candidate 2.
Inclusions and exclusions
n/a
Segmentation
n/a
Tracking Details
In the control and candidate experiment cohorts, we'll want to add events for:
- Which cohort they are in
- clicks on the banner or bubble
- Successfully setting up SAST (https://docs.gitlab.com/ee/user/application_security/sast/#configuration)
-
json schema used in
gitlab-experiment
tracking. - see taxonomy for a guide.
activity | category | action | label | context | property | value |
---|---|---|---|---|---|---|
json schema |