Skip to content
Book a call
Glossary

Learning phase

The learning phase is the period after you launch an ad set, or make a significant edit, when Meta's delivery system is still working out how to deliver it. Meta says ad sets usually exit after about 50 results in the week after the last significant edit.

Also called
Learning stage
Category
Meta platform and measurement
Updated

How the learning phase works

Every new ad set starts in the learning phase, and so does an existing ad set after a significant edit. During this time the delivery system is exploring the best way to deliver your ads for the optimization event you chose, such as purchases. The Delivery column in Ads Manager reads Learning.

Meta says an ad set exits the learning phase once it can deliver stably, which usually happens after about 50 results in the week after its last significant edit. Shops ads have a different minimum: 17 purchases through your website and 5 through Meta.

While an ad set is learning, Meta says performance is less stable and cost per acquisition is usually higher. If the ad set is unlikely to reach about 50 optimization events in the week after its last significant edit, its status changes to Learning limited. Meta is careful to say this is not a penalty. It means the current setup cannot give the system enough data to optimize well.

What sends an ad set back into learning

Meta splits edits into two groups. Some always count as significant. Others depend on how big the change is.

The edits Meta always treats as significant are any change to targeting, any change to ad creative, any change to the optimization event, adding a new ad to the ad set, pausing the ad set for 7 days or longer, and changing the bid strategy.

The edits that may or may not count are changes to the spending limit, the bid control, cost per result goal or ROAS goal amount, and the budget. In Meta's example, raising a budget from $100 to $101 is unlikely to restart learning, while jumping from $100 to $1,000 may.

Campaign budget adds a wrinkle. Meta says ad sets do not re-enter learning when Advantage+ campaign budget moves money between them, and adding a new ad set does not reset the others. Changing the campaign budget itself, or the campaign bid strategy, can send several ad sets back into learning at once.

Why the learning phase matters for brands advertising on Meta

The learning phase sits right in the path of creative testing. Adding a new ad is a significant edit, and so is any change to an existing ad's creative. A brand that drops one new ad into its best ad set every few days can keep that ad set learning almost all the time.

Meta's advice is to wait to edit until an ad set is out of learning, skip edits you do not expect to improve performance, avoid high ad volumes and use realistic budgets. It also says that when you create many ads and ad sets, the system learns less about each one, and that combining similar ad sets combines what it learns.

Meta also says you should not try to avoid the learning phase completely, because testing new creative is how performance improves over time. We agree. The goal is to launch new creative in a way the system can actually learn from.

Common learning phase mistakes

The most common mistake is judging ads too early. Meta says results during the learning phase are not necessarily a sign of future performance. Turning an ad off on day two because its cost per purchase looks high often cuts it before the system has finished exploring.

The second is optimizing for an event the budget cannot reach. If an ad set will never get close to about 50 purchases a week, it will stay Learning limited. Meta's suggested fixes include combining ad sets, widening the audience, raising the budget or the bid or cost control, and choosing a more frequent optimization event, such as add to cart instead of purchase.

The third is splitting a modest budget across many small ad sets. Each one learns slowly, and none of them gets enough results to leave learning. The fourth is constant tinkering. Every significant edit resets learning, so a week of daily creative or targeting changes can mean a week with no stable data.

How Fifth Dimension plans around the learning phase

We treat the learning phase as part of the creative plan. Under Meta Ads Management, campaign setup and creative testing are planned together, so new concepts can go live as a group instead of one at a time. That keeps significant edits on proven ad sets to a minimum.

When the client's own media buyer runs the account, our Creative Strategy work still includes a review of results after launch. Part of that review is reading results in context: whether an ad set was still learning, and how many results it had since its last significant edit. Ads Manager can show both as columns.

For Paramedic Flash, we produced 18 creatives and 6 became consistent spenders. We do not expect every ad to win. We give each concept a fair run, then put the budget behind the ones that earn it.

Sources

  1. Meta Business Help Center: About the learning phase
  2. Meta Business Help Center: Significant edits and learning phase
  3. Meta Business Help Center: About learning limited

Questions about Learning phase

How long does the learning phase last?

There is no fixed number of days. Meta says ad sets usually exit after about 50 results in the week after the last significant edit, so the time depends on your budget and your cost per result.

Does adding a new ad reset the learning phase?

Yes. Meta lists adding a new ad to an ad set as a significant edit, which sends that ad set back into learning. Adding a new ad set to a campaign that uses Advantage+ campaign budget does not reset the other ad sets.

Is Learning limited bad for my ads?

Meta says it is not a penalty. It means the ad set is unlikely to get about 50 optimization events in the week after its last significant edit. Meta's fixes include combining ad sets, widening the audience and raising the budget.

Should I pause ads that look expensive during the learning phase?

Usually not straight away. Meta says learning phase results are less stable and are not necessarily a sign of future performance. We would rather let an ad finish learning, unless it is broken or off brand.

Put these ideas to work on your ads

Book a call and we'll look at your product, your current ads and what to test next.