Tangentially related to Current Discourse, but I wonder how often various sorts of software analytics actually prove (as opposed to only appear to prove) whatever it is they seek out to prove about user behavior.
It's honestly been one of the main reasons why I aggressively opt out of analytics as a rule.
It's honestly been one of the main reasons why I aggressively opt out of analytics as a rule.
Follow
@Dee As part of consulting work (paid and otherwise), I've occasionally had a look at analytics setups and what people were measuring.
Exactly zero of the systems were measuring what people *thought* they were measuring.
This is something I've long been complaining about publicly; your metrics are almost certainly worthless, and will lead you to wrong decisions.
Interview 10 random users and you'll get a much better insight into the state of things than any amount of analytics will give you.
