Why is it important to use a lookalike audience in advertising?
The Facebook ad account learns at the account level and over time starts serving ads to the same audience again and again. We break down how this mechanism works and why a lookalike audience is critical for a steady flow of new customers.
How an ad account learns
Under Facebook rules, you are allowed to use one ad account for one business. The ad account itself has a unique capability (artificial intelligence) to remember the account's performance and the audience that responded to the ads, and to learn from it. As a result, learning happens not only at the ad set level but also at the ad account level.
To better understand why using a lookalike audience matters, let's look at an example. Imagine you own a flower shop. You launched an ad and selected your audience by age, gender and interests (it doesn't matter whether through the Boost button or the ad account). From the moment you launched your first ad campaign, the learning process begins at the ad set level and at the ad account level as a whole.
The ad set learning phase
From the moment the ad passes review, the ad set learning phase begins.
What does that mean? Suppose you got 20 unique target actions from the ad. The ad set studies what these people have in common: their interests, gender, which placement these unique actions came from, and users' behavioural patterns. But the learning process is not finished yet. By default, the platform needs 50 unique actions to complete the learning phase.
Sometimes this number is lower, sometimes higher — it depends on how similar the unique users are to one another. For example, if 50 unique actions occurred but the platform did not find anything in common between them, learning will continue. After the learning phase, the ad set's performance usually stabilizes, it is trained, and you get stable results. BUT!
Why frequency rises and the flow of customers drops
If you look into the platform's mechanism and the statistics, you can notice a slowly rising frequency. The thing is, your ad starts being shown again to the audience that interacted with your ad set, and it is also shown to your lookalike audience. When you launch a new ad campaign, the ad will already be served primarily to warm and lookalike audiences.
As a result, sooner or later the ad account's performance starts to decline and the flow of new customers drops — because you are working with the same audience. That is exactly why one of the most important rules is to independently build lookalike and warm audiences, as well as to develop a communication strategy with loyal users.
Don't forget about excluding audiences
In addition, you should not forget about excluding audiences. Without taking this into account, your ad sets will overlap. As a result, the health of the ad account will be compromised. When audiences overlap, you compete against yourself and, as a consequence, one of the ad sets gets too few impressions.
Want to apply this to your business? See the related service.
Analytics and CRO