The holdout worksheet

Run a two-week holdout on one flow and find out what it really adds.

The question

“Did the email cause the order, or just get there first?”

The assignment

Hold back 10% of abandoned-checkout entrants for two weeks and compare order rates from the store's data, not the email report.

The whole write-up →

Pick the flow that attribution flatters most, usually abandoned checkout. Hold back a tenth of entrants for two weeks, keep everything else as it is, then pull orders for both groups from the store, not the email report.

01The flow
02People who entered the flow in the last two weeks
03Share held back
%
04People in the group that got the flow
05Orders from that group (store data, same window)
06People in the held-back group
07Orders from the held-back group
08Orders the email report credited to the flow

What your numbers say

A 0% holdout is about 0 people, too few for two weeks; plan on a month or two.

0 of 8 answered · kept in this browser only until you save.

Have a brand in mind?Ask Alen about your brand →