The problem isn’t in a lack of testing, but in the metrics used when testing. If the A/B testing prioritizes “time on site”, then a design that shows you the u

The problem isn’t in a lack of testing, but in the metrics used when testing. If the A/B testing prioritizes “time on site”, then a design that shows you the user discussions you want to see and clearly indicates when nothing further has been posted will score lower than a site that intersperses ads into user content and fills a feed with “suggested for you” garbage. I certainly believe the new interfaces have been A/B tested, and don’t think the changes are arbitrary. But I also don’t think the changes are beneficial for users, nor do I think that was the goal.