Choose a decision window for an ad creative test
By Drizzle team
1 min read
An early metric can move before the campaign has delivered enough comparable traffic.
Set the rule first
Write the observation window, budget cap, and primary outcome before launching. Include the attribution definition you will read.
Watch for interruptions
Document tracking outages, stock changes, edited offers, and platform delivery resets. Each can make a neat comparison misleading.
Close the loop
Make a decision, keep testing, or mark the outcome uncertain. Do not extend a losing campaign indefinitely just to make a favored creative win.
A worked example
A campaign launches on Monday with a seven-day observation plan. On Thursday the product goes out of stock for half a day. Record that interruption before interpreting a weak purchase total. The team may need a fresh window after stock returns; silently extending the dates until a favored creative looks good would undermine the decision rule.
Decision check
Keep the numerator, denominator, date range, currency, and attribution definition with every reported result. A metric describes one part of the shopper path; compare it with visits, purchases, and order contribution before changing the creative or budget. State what the observation cannot establish.