Would it show up in the numbers on web usage, platform usage, etc? People who do this drop out of the sample - they don't show up in the numbers. As far as your stat gathering is concerned, they don't exist.

If you're actually doing a census of people and asking about their web usage and social habits, it'd show up. So maybe Google or Facebook has the data if they were to do say cohort analysis on Google Analytics or Chrome History or Facebook beacon logs, counting specifically the number of total unique Internet users that used to visit social media but no longer do. But such an analysis would require SVP-level privacy approval (because it joins together personal, non-anonymized data across multiple products), and why would an executive commission a study that potentially tells them that their job is in danger and their employer is making a mistake by employing them? And if they did, why would they ever publicize the results?

AFAIK, most of the major public-facing analytics platforms work by sampling their users. If their users are voluntarily choosing not to engage with the platform that their sampling runs on, they by definition cannot measure that change. They just become a biased sample that excludes specifically the population they're trying to measure.

But they still READ. So, if you 'interact' (and by that I mean do any write-like action, like commenting, posting, liking, whatever) less, that's gonna show up.