Clearer analytics/more detailed reports that still respect user privacy #6090
Description
What's the problem this feature will solve?
In order to make better decisions about user experience, deprecating features, rearranging routes and pages, etc., we could use more data about how our users use PyPI. Example questions we want to answer: how many times per year does the median or mode user visit the site? Do logged-in users use search more than anonymous users do, and is it the first thing they usually want to do when they hit the front page? We have about a 38.6% bounce percentage from the front page; is that high for this kind of website? According to @nlhkabu we'd want to dig into that data by looking at:
- Where they come from (e.g. referrer, search, social, etc.)
- What device or browser are they using (are we getting a higher bounce rate on different devices?)
- Where they are located (maybe it's a language issue?)
This kind of analysis does not currently seem possible; we don't think we have (for instance) configured analytics to track whether users are logged in or not.
Describe the solution you'd like
Maybe there is someone in the community with a lot of analytics experience who can help us improve our Analytics implementation, and create custom reports? Or, if that doesn't pan out, we could get funding to hire a consulting firm to help? And along the way we would pay attention to our privacy policy to ensure that we continue to respect users' privacy.