Ad platforms know behavioral patterns across their users. Upload a seed list, customers, high-value clients, completed bookings, and the platform models what its members share, then targets strangers matching that pattern. The seed's quality decides everything: model your best customers and it hunts for excellence, model everyone and it hunts for average.
Tracking limits shrank the data behind lookalikes, and platforms have shifted toward broader, algorithm-driven targeting where your seed guides delivery less precisely than before. Lookalikes remain useful as directional input, and the durable version is owning the seed: a clean first-party customer list, which no privacy change can take away because customers gave it to you directly.
Seed with your genuinely best relationships, keep the similarity threshold tight before widening, and remember the audience only earns its keep if the click lands somewhere built to convert. Cold lookalike traffic behaves like cold anything: the landing page and follow-up machine decide whether the statistical resemblance turns into booked work.
Platforms accept small lists but model better from hundreds. Below that, prioritize interest and engagement audiences while your customer list grows.
Constrained by geography, partially: the pattern-matching pool inside one metro is small. Local businesses often do better with intent-based search and retargeting first.
Lists are hashed before matching, and consent plus a privacy policy that discloses advertising use is the standard. If your emails were collected promising otherwise, do not.
Everything in this glossary, we build and operate for real businesses. Thirty minutes maps it to yours.