Working with

Facebook's Learning Phase

The best way to use Facebook’s machine learning is to reduce your ad set volume. When we publish too many ad sets at once, Facebook delivers each ad set less often. Fewer ad sets will exit Facebook’s learning stage, and more your budget is spent. But how do you narrow your options intelligently?

What is Facebook Learning Phase?

Each time an ad is shown, Facebook ads delivery system learns more about the best people and places to show the ad. The more an ad is shown, the better the delivery system becomes at optimizing the ad’s performance. The learning phase is the period when the delivery system still has a lot to learn. During the learning phase, the delivery system is exploring the best way to deliver your ad set - so performance is less stable and cost-per-action (CPA) is usually worse. The learning phase occurs when you create a new ad or ad set or make a significant edit to an existing one.

How to best adjust to the Learning Phase?

This all goes back to ad set optimization, turning on and off ad sets is not the way to get optimal results. When you have an ad set that is not converting well, it is best to wait until your ad set has completed the Learning Phase (around 50 conversions) to begin adjusting your targeting or adding new creative.

Facebook suggests:

  • Avoid unnecessary edits that cause ad sets to re-enter the learning phase. Edits that meaningfully change how your ad set might perform in the future can cause an ad set to re-enter the learning phase. Only edit your ads or ad set when you have reason to believe that doing so should improve performance.
  • Avoid high ad volumes. When you create many ads and ad sets, the delivery system learns less about each ad and ad set than when you create fewer ads and ad sets. By combining similar ad sets, you also combine learnings.
  • Use realistic budgets. If you set a very small or inflated budget, the delivery system has an inaccurate indicator of the people for whom the delivery system should optimize. Set a budget large enough to get at least 50 total conversions and avoid frequent budget changes (which can cause an ad set to re-enter the learning phase).
  • How do I navigate the Facebook Learning Phase Guidelines?

    Unfortunately, Facebook doesn't necessarily provide the appropriate tools to help users through their suggested guideline. This is why we created LIFTR; to give more visibility to the Facebook advertising process. As a marketer, you're constantly tweaking your ads for optimal results and asking questions like:

  • How am I supposed to know what an unnecessary or significant edit is?
  • How am I supposed to decrease your ad volume if I don't know what will work?
  • What is the right ad set volume to have?
  • What is a realistic budget is for each ad set?
  • The list goes on and on...

    LIFTR to the rescue

    With LIFTR you can test your best ideas against your existing Facebook audiences before you launch. By connecting your Facebook account, you are able to run tests against a LIFTR model that takes in your previous campaign data and uses it to adapt its engagement understanding. Using LIFTR's engagement scoring will highlight which edits are necessary, by showing relative scores for all creative versions.

    LIFTR also provides detailed information around ad and ad set performance, informing you with insights such as how much of your budget should be dedicated to each customer segment. These insights help decrease your ad and ad set volume by going further than just suggesting that you combine similar ad sets. Facebook wants high engagement ads. They want people to engage with your ads. It is core to their business and user experience. In order to get the most out of your ads, you should only publish ads LIFTR identifies as the most engaging.

    Facebook has developed a great advertising system, but there is still a lack of visibility for users to who want to know what actions they need to take to improve their ad performance. This lack of visibility forces a lot of guessing, testing and adjusting all while depleting your budget. LIFTR provides additional visibility and functionality to make data driven decisions.