Background
As part of the simple summary experiment, we created an explicit opt-in modal that forces users to explicitly accept the terms of service before being enrolled in an experiment. We predict that this workflow will lead to a fall-off in experiment participants, either due to users ignoring the modal or clicking 'decline'.
In order to offset this drop, we should re-run the QuickSurvey we used to initially enroll users in the browser extension.
User story
- As a casual Wikipedia user, I'm interested in being enrolled in experimental features for Wikipedia.
Requirements
- Preview current experiment enabled statistics (in statsd) to determine if more participants are needed
- Determine how many more participants are required for this experiment
- Previously, the extension had about 10k pageviews per day. When enabling this QuickSurvey, we should monitor the experiment Grafana graphs for an increase in pageviews. Once the pageview level reaches about 10k per day, we can disable the quickSurvey.
- Rerun the QuickSurvey: https://gerrit.wikimedia.org/r/c/operations/mediawiki-config/+/1071890
- Given the simple summaries experiment is only running on english Wikipedia, it makes sense to only target the quickSurvey to that wiki.
NOTES: The quickSurvey sampling rate will have to be slightly modified. Currently, if multiple surveys are running with the same sampling rate, the user will only ever see one of those surveys. With that in mind, we should examine any currently running surveys and choose a sampling value that is unique.
Acceptance criteria
- Determine if re-running the QuickSurvey is necessary
- Rerun the quickSurvey for a limited time period or until the number of participants increase
- Disable the quickSurvey after the number of participants has increased, based on the 'enabled' experiment metrics.
This task was created by Version 1.2.0 of the Web team task template using phabulous