> Having "analytics disabled" correlate with "using that specific feature" is always a major problem for data driven decisions, worse it's a well hidden and har

> Having "analytics disabled" correlate with "using that specific feature" is always a major problem for data driven decisions, worse it's a well hidden and hard to estimate problem. You know, I’m generally a privacy conscious consumer and default to opting out of analytics, but if a company asked me to do a survey about how I use the product I absolutely would! I enjoy giving feedback, but am always going to pass on dragnet surveillance. I’ve never seen this though. Perhaps I’m naive to certain business interests or statistical techniques, but it seems like such a survey would be a necessary part of a robust decision making process based on mass-surveillance gathered analytics for a product that doesn’t earn money by directly selling that data. (if you only make money by tracking users, I understand why you might not care about users who prevent it, but even then with the power of network effects I’d expect a business would at least want a finger on the pulse) I mean, I don’t have a degree in stats, but from what I do know this is the kind of due diligence that the foundation of trust in statistics is based on. Maybe I’m just confused and someone can correct me, but I’ve come to see the selective enforcement of good statistical practices as one of the major problems in the world today, because a) I see it so much, b) it gives certainty where little is deserved, c) it prevents us from getting to truth.