How Long Should I Wait Before Judging a Meta Campaign?

Wait for the clock that matches what you last changed. Meta documents separate ones: the learning phase (usually about 50 results in the week after a significant edit), a full week for reading everyday swings, at least 7 days after you adjust a cost per result goal, and a recommended minimum of 7 days for an A/B test. Each one has its own trigger, so start by finding yours.

What starts the learning-phase clock?

Meta says an ad set enters the learning phase when you create it or make a significant edit. It usually exits after about 50 results in the week after its last significant edit. The Delivery column reads Learning until then, and Learning limited if the ad set is not getting enough results to exit.

Meta lists these as significant edits:

• any change to targeting, ad creative or optimization event
• adding a new ad to the ad set
• pausing the ad set for 7 days or longer
• changing bid strategy

A change to a budget, spending limit or goal amount may or may not be significant, depending on its size. Meta recommends waiting to edit an ad set until it is out of the learning phase.

To see where you stand, add the last significant edit column in Ads Manager and check your results since that date.

When is a single day too short to read?

Meta’s page on fluctuations says to analyze performance over at least a full week, to see whether you are getting a relatively steady amount of results. It says daily budgets are an average over the week and may run up to 75% above budget on some days. If you break results down by hour, Meta says you may see swings that average out over time.

Meta’s guidance is to look at the campaign (with Advantage+ campaign budget) or the ad set (with ad set budgets) across a full week, and to be careful with hourly breakdowns.

How long after you change a cost goal?

For a cost per result goal, Meta says day-to-day costs naturally fluctuate and to evaluate with weekly averages rather than daily results. It says to wait at least 7 days after any adjustment before re-evaluating.

Meta’s best-practice page adds two figures. It suggests a measurement window that captures at least 50 to 100 conversions. It also says to run cost per result goal campaigns for at least 3 weeks before you scale. Meta’s page on cost per result goal also says your actual weekly average cost per result may vary from your set goal.

Note the date of each goal change and read the weekly average only after 7 days have passed.

How long should an A/B test run?

Meta’s page on A/B testing best practices recommends a minimum of 7 days for the most reliable results. It says tests shorter than 7 days may produce inconclusive results. If your typical customer takes more than 7 days to convert after seeing an ad, Meta’s example is a longer test, such as 10 days. The entry linked below covers test length in full.

Which clock applies to your campaign?

Match your last action to the clock:

• New ad set, or a significant edit: the learning-phase clock, about 50 results in the following week.
• Changed a cost per result goal: weekly averages, and at least 7 days after the change.
• Comparing two versions: an A/B test, with Meta recommending at least 7 days.
• Reading a noisy day: a full week.

If you did several of these, each clock starts at its own change. Reading after the latest start date is one way to apply them; that is an inference from these pages, not Meta’s wording.

FAQ

Does the learning phase last a set number of days?

Meta ties it to results inside a week: ad sets usually exit after about 50 results in the week after the last significant edit. The time therefore depends on how quickly your ad set gets results. If it is not getting enough, the Delivery column reads Learning limited.

Why does Meta say to use weekly averages?

Meta says day-to-day costs naturally fluctuate and that cost per result is an average, so it is best to avoid restrictive time breakdowns when comparing it to your goal. Its example is higher costs during one hour that average out over the rest of the day. For a cost per result goal, Meta says to evaluate with weekly averages.

Does pausing an ad set restart the clock?

Meta lists pausing an ad set for 7 days or longer as a significant edit, and says the ad set re-enters the learning phase once you unpause it. The last significant edit column in Ads Manager shows when you last made a change that could reset learning.