Hacker Newsnew | past | comments | ask | show | jobs | submitlogin

But how do they measure those things, fuzzy or not? And more importantly, can they actually invalidate their own hypotheses?


On your B2B contact us form, ask 'how did you hear about us', or do it as part of the onboarding.

When people leave, ask them where they're going, and where they heard about them.

Etc. Are you going to get acurate and complete data? No to both. But you might get actionable data. Will you get real-time feedback? Nope.


Some things are worth doing even if you can't scientifically prove that they are.


It's not because you can't measure it that it's not scientific (and vice versa).


But you don't know which ones are. Which means you're wasting money on those that aren't.


Even if you think you know, do you really?

Attribution is a notoriously tricky issue and it's only getting harder with GDPR and browsers making changes to cookies.




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: