I guess you have to assume such a small percentage of people will actually be using this, so the numbers won't be too bad??
You don't "have to" assume that, but why start worrying about the source of error in your metrics now?
The cumulative wight of this and all the other problems with your data probably won't dip their utility below zero, so just go back to plugging your ears and yelling "lalalalala".
You don't "have to" assume that, but why start worrying about the source of error in your metrics now?
The cumulative wight of this and all the other problems with your data probably won't dip their utility below zero, so just go back to plugging your ears and yelling "lalalalala".