Facebook Ads and Lookalike Audiences

Forum White Hat SEO PPC Facebook Ads and Lookalike Audiences

  • This topic has 3 replies, 1 voice, and was last updated 1 week, 3 days ago by itwasntmeoritwas.
Viewing 4 posts - 1 through 4 (of 4 total)
  • Author
  • #16725

    Lookalike audiences are overrated… sometimes.

    I’ve seen so many ad accounts that are struggling to get just their first 50 sales from Facebook ads, and most of their audiences are Lookalike audiences created from data that they do not have enough of.

    I understand how Lookalike audiences can sound great in theory… you tell Facebook to use the pixel data from the 30 purchases you have on your store to turn those 30 customers into millions of similar customers.

    But the reality is… if you don’t give Facebook enough data in terms of quantity AND quality, that Lookalike audience will not perform well. Something that people often neglect is the data that Facebook has accumulated for many years. And that is interest based audiences.

    [Here are some results comparing interest ad sets to a Lookalike audience ad set](https://i.imgur.com/Ndv6gBD.png)

    This ad account was relying heavily on Lookalike audiences before I started working on it, so we left one of those ad sets on to see if it would be beaten by an interest audience. And it was 3 times.

    If you have experienced similar results when launching ads towards Lookalike audiences, then you can either

    1. Try a different type of Lookalike audience (different data source or different percentage) with the same budget
    2. Use that budget towards an interest audience instead

    The problem with Lookalike audiences is that it can cost you a lot of money to know what that sweet spot combination of data source and percentage point is. This is because of the quality and quantity required.

    Scenario 1 (High quality, low quantity): You’ve gotten 20 sales on your store of your perfect customer. Creating a purchase Lookalike might not work well without more data (or it could work, you never know unless you test it).

    Scenario 2 (High quantity, low quality): You’ve had 20,000 visitors to your website in the last 90 days. But what if only 0.2% of them are customers. That would probably not work either due to the low amount of quality buyers in that audience.

    Here’s the take home with this…

    * Don’t neglect interest audiences in favor of Lookalikes all the time
    * Don’t always assume that a purchase Lookalike audience will always outperform a traffic or add to cart Lookalike audience

    Hope that helps! I’m happy to answer any specific questions you have.


    Great write up! Makes sense and I’m pretty new to FB ads. One question, when do you know when to pause a lookalike adset? Do you have a certain spend limit based on ROAS?


    Fantastic, detailed post. Thanks!


    20 sales is a proper example? because facebook won’t even let you build LAL from 20 sales, minimum is 100

    Another question, what is ideal minimum for LALs?

    i had 1000 ATC, 500 IC, 200 Sales, none of those LALs didnt outperform cold interests.

Viewing 4 posts - 1 through 4 (of 4 total)
  • You must be logged in to reply to this topic.
Scroll to Top