Users move through different paths when using content translation, and we want to have a better understanding about how many users take each path and the results they produce.
Translators can enter the tool through different entry points (Contributions page/menu, interlanguage links, etc.), start a translation in different ways (searching for a specific article, picking a suggestion, or selecting an article they kept for later), and advance through different stages for the translation (starting a new translation, continuing an existing one, and publishing). Finally, the published content may survive, be edited further, or be deleted by the community.
Measuring and visualizing this funnel would allow understanding better which paths are used the most and which ones lead to more productive activities.
We are interested in bothImportant questions this data could help answer include:
* How much publication can be attributed to each entry point?
* Are users coming from campaigns more or less successful than other users?
* How many translations face publish failures, the overall activity for any language as well as the activity for specific wikis. Ideallyand how many eventually overcome them?
* Which type of suggestions are most effective, it would be great to focus on any particular wikinot just in getting people to start, but if a predefined set is needed we'd focus on the selected small Wikipedias (Bengali, Malayalamalso in getting things published?
* With AX, Tagalog, Javanesehow often do people translate in multiple sessions versus a single one?
* On mobile, and Mongolian).how often do users' sessions stop without passing through the confirmation dialog (suggesting that they lost their in-progress work)?
An initialThe diagram illustratingbelow shows the user workflow and possible stepproposed events to instrument is captured below:, and the flows between them:
{F31786984{F33939728, size=full}