In the experiment's metrics when the data is pulled from BigQuery. Does it show success for the same ID fro before the ID was exposed to the experiment? Say if the test starts Dec 1 2023, the ID purchase on Dec 2. 2023 but only sees the experiment on Dec 12. I am pulling in data from Dec 1. 2023 to Dec 15. 2023 from Bigquery. Would the ID be counted as a purchase for the variant the ID was exposed to?
f
fresh-football-47124
12/15/2023, 9:25 AM
You can set an attribution window
fresh-football-47124
12/15/2023, 9:26 AM
by default, we look for the metric event after the exposure event
fresh-football-47124
12/15/2023, 9:26 AM
so in that case, we would only count the purchase if it happened after the 12th
fresh-football-47124
12/15/2023, 9:27 AM
(you can set a negative look-back window if you really want to)