Page MenuHomePhabricator

[Impact Analysis] Evaluate Impact of introducing mobile edit notices
Open, MediumPublic


This task involves the work with doing an analysis to evaluate what impact introducing the edit notices on mobile (via the EditNoticesOnMobile.js gadget and T312587): is having on:

  1. The percentage of anonymous editors who click an edit affordance and decide to create an account Per T312587#9100232, the mobile edit notices will appear after the sign up/log in view. As such, I don't think introducing mobile edit notices us likely to impact sign ups.
  2. The percentage of people, grouped by experience level, who publish edits they initiate
  3. The percentage of edits that are reverted within 48 hours of being made

Decision to be bade

This impact analysis will help us make the following decision: **Does the Editing Team think any adjustment need to be made to the EditNoticesOnMobile.js gadget and the experience T312587 introduces to ensure that is not causing any regressions in the rates at which people are:

  1. Creating accounts
  2. Abandoning and publishing the edits they start
  3. Making edits that are reverted?**

Analysis scope

Open question(s)

  • On what dates was the EditNoticesOnMobile gadget made available to each of the following segments of people? Volunteers defined these segments on Wikipedia_talk:EditNoticesOnMobile.
    • Admins (activated 23:10, 19 July 2022)
    • Extendedconfirmed accounts 2022-08-02T17:58:08‎
    • Autoconfirmed accounts 2022-08-22T18:08:38‎
    • Everyone (logged in and out) 2022-09-14T18:59:58‎

Note: In T312299, the Editing Team determined that we did not find any issues that we thought warranted blocking the deployment of the EditNoticesOnMobile.js gadget at

Event Timeline

Due to ongoing issues with the hacks to work around T313140 the second wave has been temporarily put on hold. The issues should be fixed in my userspace version, but we somehow have no testers for the userspace version and after matmarex notifying me of ENOM accidentally (and unintentionally) now fixing T313140 globally I'm no longer confident enough to rely on my own testing before asking for the MediaWiki: version to be synchronized with the userspace version.

ppelberg renamed this task from [Impact Analysis] Evaluate Impact of EditNoticesOnMobile.js gadget at to [Impact Analysis] Evaluate Impact of introducing mobile edit notices.Aug 17 2023, 7:25 PM
ppelberg updated the task description. (Show Details)
ppelberg added a project: Product-Analytics.
mpopov triaged this task as Medium priority.
mpopov moved this task from Triage to Upcoming Quarter on the Product-Analytics board.

@ppelberg I've checked and confirmed that events are being logged as expected in VisualEditorFeatureUse. See summary of checks and some initial data counts below:

The data below reflects all editing sessions where an edit notice was shown (feature: notices, action: show) from 11 October through today:

  • Events logged on both desktop and mobile web platforms following 11 October deployment:

Are these numbers for all projects and languages combined, minus those that use ENOM?

Are these numbers for all projects and languages combined, minus those that use ENOM?

@AlexisJazz Yes, that's correct