When a search suggestion is selected, measure the search suggestion position in the suggestions block (1st, 2nd, etc.) and suggestion string length.

We want to make the same measurements for the search bar.

This is probably implemented using an enumerated histogram (position) and an exponential histogram (string length).

(In reply to Georg Fritzsche [:gfritzsche] [at workweek] from comment #2) > This is probably implemented using an enumerated histogram (position) and an > exponential histogram (string length). Assuming you don't want to correlate the two, in other words know what is the average length of all the second search results that are selected by users, then two separate histograms are fine. Can we add more context about what the data from this probe will be used for? Note that when using an IME just typing one single Unicode character may be a complex operation, so we might have to do some locale-specific analysis.

I don't think it will be important to correlate the length of the search results with their position at this point. In broad terms, this probe will be used to evaluate search suggestion performance, together with the edit distance measured in bug 1334630. This may not be something we want to track in general, but it will be useful in comparing suggestion performance between search engines, or say between branches of a test that affects search suggestions. We can discuss this more during the workweek.

Moving to p3 because no activity for at least 1 year(s). See https://github.com/mozilla/bug-handling/blob/master/policy/triage-bugzilla.md#how-do-you-triage for more information