What is Lookalike audience?
A lookalike audience is a targeting segment a platform builds by finding users who resemble a source list you supply — past purchasers, high-value customers, or people who completed a specific action. Quality depends almost entirely on the source.
A list of a few thousand genuine repeat buyers produces a far better lookalike than a large list of newsletter signups, because the model can only learn the patterns present in what it is given. Platforms let you set how closely the audience matches, typically expressed as a percentage of a country's population: tighter percentages resemble the source more and reach fewer people, broader ones scale further with weaker resemblance. The practical trend since 2021 has been away from finely tuned lookalikes and toward broad targeting with strong creative, because signal loss degraded source-list quality and platform algorithms became better at finding buyers when given room to explore. Lookalikes remain useful as a seed for a cold campaign and for excluding existing customers from prospecting, but they are no longer the primary lever they once were.
How it relates to AI UGC
As lookalike precision has degraded with signal loss, the compensating lever has moved to creative: broad targeting with segment-specific creative now outperforms narrow targeting with generic creative on most accounts. That shifts the work from audience configuration to producing enough distinct creative angles to speak to each segment, which is the volume problem ppl.studio exists to solve.
Key statistics
- Lookalike audiences on Meta typically deliver 20–50% lower CPA than broad interest targeting.
- Pairing lookalike audiences with UGC-style creative can improve conversion rates by 2–3x vs. stock imagery.