E-Commerce marketers run two kinds of campaigns with Facebook ads: performance based and brand awareness campaigns.  

Now most brand awareness campaigns are still following good old mass marketing techniques while performance-based campaigns are predominantly data-driven and use micro-segmentation in order to provide customers with personalized experiences.

While executing a Facebook campaign, ad sets go into learning or learning limited. Digital marketers always try to work with limited budgets to avoid the learning phase and achieve the maximum benefit out of the Facebook algorithm.

The way you structure your Facebook ad campaigns can make or break your results. To execute successful performance campaigns with Facebook ads, it is important to have a proper foundation or a campaign structure.  

If you take the time to plan out your campaign and set up the right ad sets, you'll be able to reach your target audience, get them to convert, and see a positive ROI on your investment. So, let’s talk about how you tackle the learning phase.  

What is the Facebook ads learning limit?

When you launch a new campaign, the ad sets within that campaign enter the learning phase, which is specific to each ad set.  

The Facebook ad learning phase begins when you create a new ad set or significantly edit an existing one. During this phase of trial and error, the Facebook algorithm tests different ways by optimizing your ad for creativity, audience targeting, place and time to yield maximum conversions.  The learning phase ends when Facebook has enough insight to predict your ad’s future performance and delivery.  

To exit the learning phase, your ad set must generate 50 conversions within 7 days after launch or a significant edit to the target audience.  

When your ad set doesn't generate enough optimization/conversion events during the first week (i.e the 50 required to leave the learning phase), it enters learning limited. This means that Facebook predicts your ad won’t perform well in the future and leads to worse results for your ad sets, as Facebook's algorithm can't use all its resources to optimize your ads.  

Myths about Facebook learning limited ad sets

Here are a few myths surrounding Facebook learning limited ad sets that can keep you from getting the most out of your advertising dollars (and time):  

Myth #1: Copying an ad set will copy the learning  

Marketers often believe that if they reuse an old ad set and duplicate it for newer campaigns, they’ll avoid going through the learning phase all over again. However, this isn’t true. Copying an ad set doesn’t copy the learning and the duplicate ad sets take longer to complete their learning phase.  

Myth #2: Any edit to the ad set will result in the ad set going to learning  

A lot of advertisers are confronted with this fear when they have to make a change in their ad set. However, this is not true. Not all changes result in re-entering the learning phase. Ad sets go into learning again only when significant edits are made.  

The edits that trigger the learning phase are:

  • Any Change to Targeting
  • Any Change to Ad Creative
  • Any Change to Optimization
  • Adding a New Ad to Your Ad Set
  • Pausing Your Ad Set for 7 Days or Longer
  • Changing Bid Strategy      

Myth #3: Any ad set goes to learning limited after 7 days

Not all ad sets enter learning limited after 7 days. If the Facebook algorithm identifies that the ad set has failed to get 50 conversions within a 7-day period, it will go to learning limited.  

Some of the factors that cause Learning Limited are low budget, small audience and high auction overlap (i.e using the exact same audience or ad in multiple ad sets).

How to solve the learning limited problem “during” the first 7 days with CBO campaigns?

The first 7 days of your campaign are crucial. If you don't get enough people to like your page in the first week, it's unlikely that you'll recover from it.

Fortunately, you can solve this using CBO campaigns. Here are a few practices you’ll need to follow:

  1. Keep a maximum of 4 ad sets per campaign and 3 ads per ad set.
  2. Make sure that your ad set audience sizes are not too narrowed. If you have narrow audiences, combine them into one ad set. (Ad sets with more than 1 million audience size are a healthy audience size for top-funnel. If you have limited visitors, target all the visitors and fanpage/Instagram engagers in the same ad set)
  3. Set the campaign daily minimum to 50 (Your ad set needs 50 transactions per week) x CPA (Calculate the budget) x 1.5 (Initially the CPA is high) x 3 (Budget for at least 3 ad sets) / 7 (Convert weekly budget to daily budget)
  4. After 2 days of the launch, pause any ad set if it gets 0 transaction
  5. After the 4th day, pause any ad sets with RoAS less than 50% of the campaign RoAS
  6. After the 4th day, If ad sets are performing below the run rate of 50 per week, then:
    - Pause ad sets that are not going to achieve 50 events or increase your ad budget.
    - Change bid caps or cost controls such as minimum ROAS or bid cap to achieve more conversions.
    - Change optimization event to ATC  

How to solve the learning limited problem “after” the first 7 days with CBO campaigns?

If your campaigns have run for 7 days or longer and you've noticed that your ads are still not showing, it's likely that you're experiencing the "learning limited" problem.  

Once your ads have had a chance to learn what works best, you can take action to increase your conversion rate even more:  

  • If at least one ad set has completed the learning, then pause all other ad sets or increase your ad budget by 20% to allow other ad sets to have enough budget.
  • Pause unprofitable ad sets and launch new ad sets with new audiences to replace those ones that could not make it through the learning phase.  


When running campaigns on Facebook, spend time optimizing your ad sets.

The more time you spend developing them, the better your results will be. This is because Facebook AI optimizes your ads to ensure they are reaching the right audience at the right time, based on what it learns about your campaign.

In conclusion, we have seen that ad sets with learning completed perform better than learning limited ad sets as it utilizes the benefit of Facebook AI algorithm. Always make sure that campaigns are optimized to achieve the learning completed status for your ad sets.

Use Alavi and Scale Without Fear

Developed for small and medium businesses, Alavi is a martech application that performs predicative analytics on your customers’ behavioural data. By combining AI, machine learning and automation, it helps digital marketers better understand their audiences, improve their targeting, and scale confidently profitably.

Alavi has a proven record of helping brands in a variety industries grow their revenue online. It has a very short time-to-value and is easy to setup and use.

Get a Demo Now

To contact us,
Please email