is there any rule of thumb as to how many visitors...
# ask-questions
r
is there any rule of thumb as to how many visitors or how long a test should run before i can start drawing conclusions?
f
depends on the metric you're trying to move, not the total audience
the rule of thumb is that you want to have at least 150 to 200 conversion events per variant - so if you're trying to improve sign up rate, you'll want 300 to 400 total signups split between a two way test
if your conversion rate is 10% from visitors to signups (for example), then you'll need 3000 to 4000 total visitors
but the smaller the effect of your change, the more traffic (or power) you'll need to detect it