A celebrity lookalike may have the style, but not the talent, of the latest Netflix star. In the same way, Facebook lookalike audiences may seem similar to your existing audience, but may not be engaged shoppers. Is it time to retire this long-standing Facebook marketing strategy?

Evolving privacy rules and content guidelines may mean that lookalike audiences aren’t quite as effective as they once were. But, some brands are finding that small adjustments (and the right data) can make a big difference. Here’s a closer look at what businesses should consider when using this ad strategy. 

What is a Lookalike Audience on Facebook? 

Lookalike audiences are new audiences for your ads based on shared characteristics with people who already buy from you. They’re created based on a custom audience you’ve already made in Meta Ad Manager. The system uses demographics, behaviors and interests from your existing audiences to identify new people with similar qualities. 

The ideal source audience has between 1,000 and 5,000 people. Ideally, these people are your best customers, the superfans who are committed to shopping with you. You can then fine-tune your lookalike audience by setting how closely you would like the two groups to match. A 1% audience is the closest match, while a 10% audience might give you a broader group of potential customers.

There are some limitations to this method. Certain limitations protect children by preventing marketers from targeting them directly. Ads about credit, employment, or housing opportunities may also be ineligible for lookalike audiences. 

Despite these limitations, lookalike audiences have long been seen as an efficient targeting method to reduce ad waste and drive results. That may be changing. 

The Impact of Apple’s 2021 Data Privacy Changes

In 2021, Apple introduced a new App Tracking Transparency feature. It allowed iPhone users to choose whether apps could track them across the web. Only about 25% of users opted-in to tracking by spring 2022. The results were swift and devastating for Facebook. Facebook reported its first revenue drop and cost the company about $10 billion in revenue for 2022. 

This change closed the curtain on many transactions. Facebook could no longer see if users made purchases after viewing ads. The social media company stopped sharing detailed demographic information about customers, because that information simply wasn’t available. This made lookalike audience building more difficult. 

The Decline in Lookalike Audience Performance

Apple’s privacy changes were the headline maker, but Facebook lookalike audiences had already been shifting before that. As early as 2019, brands were noticing that the decrease in third-party data caused a dip in lookalike audience results. 

Some marketing agencies were already moving away from 1% audiences and aiming for 2-3% instead. App Tracking Transparency just shifted the landscape a little more. Basically, as the availability of tracking data has decreased, marketers have expanded their lookalike audiences to compensate for the loss. 

Volume vs. Quality: The Real Measure of Success

Make no mistake, lookalike audiences are still valuable in 2024. An audience that shares some similarities to your existing customers will often outperform a random selection of people. It’s a question of volume vs. quality. 

You can show your ad to a lot of people, and that’s great if awareness is your goal. But if you want to make sales, you’re better off showing it to people who are likely to want what you’re selling. Interested audiences are likely to convert at a higher rate. That means a bigger random audience might still deliver fewer results than a smaller lookalike audience. 

Start With a Quality Source Audience

There’s no need to give up on lookalike audiences just yet. Instead, increase the quality of your lookalike audiences by starting with a proven source audience. You don’t have to rely on Facebook to serve up your initial customer list. Instead, you can import a proven list based on who visits your website. 

Unlike many advertisers who rely on third-party data to build a pool of potential customers, Ignite by Launch Labs prioritizes first-party data for greater accuracy. Ignite’s modeled audiences filter by geographic radius and build models based on website behavior. 

The results are clear. In Beta testing, traffic from ads targeting modeled audiences outperformed standard Facebook Audiences. These visitors spent 386% more time on site.

See how you can engage, convert, and activate your audience using Ignite by Launch Labs. Set up your free demo today.