Left Sidebar: Sub-menu items as tabs
What did we learn?
The general approach seems to work quite well for discoverability. People were directly understanding very quickly how the tabs relate to the content on the page.
It also allowed them to move faster towards the Settings, as these are now both under "Settings" as well as the respective feature area. Of the 10 participants looking for "Repository Settings", 7 found it in "Repository" and 3 found it in "Settings".
What’s this issue all about?
Over the last milestones, the Editor team built out a lot of short-term improvements and quick wins around Navigation and Settings. As we also need to consider our long-term strategy, one idea was to push all sub-menu items from the left sidebar into tabs on the page itself.
Before | After |
---|---|
Who is the target user of the feature?
Everyone who uses GitLab.
What questions are you trying to answer?
Is our assumption that users tend to switch more in the left sidebar between top-level items (e.g. from Issues to Merge Requests) than sub-menu items (e.g. from Repository -> Files to Repository -> Branches) correct, and does the new design still allow users to find all current features?
Core questions
What differences in success rate, time to find, and similar metrics, are there between the current and the newly proposed design?
What hypotheses and/or assumptions do you have?
Generally, the new design makes it easier to find features because the left sidebar is easier to overview, and users most often navigate towards the first entry of a top-level item anyway.
The plan is to give the user the task to navigate to a certain piece of information within our UI, e.g.:
- Navigate towards merge request x.
- Create a new issue.
- Find branch y.
- Change the default branch for this project.
In a first study, I would love to test this out with 5-7 users in our current UI (maybe even with an actual live project instead of a prototype), and then do the same task with 5-7 users with the newly proposed UI.
@mle @katokpara I would love to hear what you two think about this plan? My immediate question: Would the baseline even be necessary if we run the study with the new design first and the observations are all positive?
What decisions will you make based on the research findings?
Whether to move forward with investigating this idea more deeply and building out the prototype for more cases.