Consider measuring something site-engagement-related for HTTP-bad |
|||
Issue descriptionSite engagement might be an interesting thing to track for the HTTP-bad rollout, though I'm not sure what exactly we'd want to track. Maybe, average site engagement score on sites that trigger HTTP-bad conditions? (So we can compare site engagement scores for users who saw the omnibox warning vs those who didn't)
,
Dec 1 2016
,
Dec 2 2016
We already have comparative metrics for the engagement distribution and ratio between HTTP and HTTPS sites. Are those the sorts of things you're interested in? We could possibly break up the HTTP into HTTP/HTTP-Bad or something like that.
,
Dec 2 2016
Re #3: yeah, I think splitting up HTTP into non-HTTP-bad and HTTP-bad might be helpful. Then we could compare the HTTP-bad metrics for users who are in the field trial vs those who are not to see if the omnibox warning decreases site engagement. I'm not sure it would be trivial to split up HTTP that way, though, because sites can move in and out of HTTP-bad states. Could you point me to the code where the HTTP vs HTTPS site engagement metrics are recorded?
,
Jun 7 2017
cc'ing elawrence for Phase 2 visibility, but unless he's super excited about it, I think we probably won't do this anytime soon. |
|||
►
Sign in to add a comment |
|||
Comment 1 by est...@chromium.org
, Dec 1 2016