[yjftsjthsd-h]

Abstract: It's still spying on users.

Practical: Supposedly-aggregated stats have a history of actually being perfectly possible to analyze back into individually identifiable information. Also, it's conveniently the same tech stack in a way that makes it easier to make an actual slippery slope.

[tempestn]

The practical argument I can understand. From the abstract argument though, it sounds like you'd be opposed even if the anonymization could be guaranteed, which I don't understand. Why is it "spying" to try to understand in aggregate how users are using your website? How are you supposed to eg. identify usability problems without this information? And what is the harm to users? (Again, in the abstract case where we leave aside any possibility of individual users being identified.)

[micromacrofoot]

because no one has shown an anonymization guarantee yet