Custom and lookalike audiences in Meta Ads are two of the highest quality targeting tools you can use to reach the right people, whether they already know your brand or look exactly like those who do. If you run paid campaigns and want better conversions without wasting budget, understanding how these audiences work changes the game.

¿What is a custom audience in Meta Ads?

A custom audience is a group of people who already took a specific action with your brand. This data is not public; it only exists because you set up tracking through your pixel, Conversions API, or other owned sources.

¿What is a custom audience? It is a private list of users who interacted with your brand, like people who bought from you in the last 15 days. Only you have access to it because it comes from your own tracking.

Custom audiences pull from two big buckets of origins. Your own origins are tied to your pixel or Conversions API and include website activity, app activity, product catalog, customer list, and offline activity. Meta origins are tracked automatically by Meta without extra setup, like video views, lead forms, instant experiences, shopping activity on Instagram, events, your Facebook page, and Facebook posts [02:30].

¿How do you upload a customer list as a custom audience?

In the ad set, go to the audience section, click on custom audience inclusions, and select create. Choose customer list as the origin and upload your CSV file. Meta will ask if your file includes a column for customer value, and this is where things get powerful.

If you add a column like lifetime value (LTV), Meta can identify which users on your list are worth more, so it learns who to prioritize when looking for similar people. It is not mandatory, but it gives Meta richer signals.

Name your audience clearly and include the date range you used, for example the last 15 days. Then map the columns: emails, phones, country, and name. Meta processes the file row by row, in this case 144 rows, and your custom audience is ready [05:40].

¿How does a lookalike audience work?

A lookalike audience targets new people who do not know you yet, but it is built on top of a custom audience. You share a high quality custom audience with Meta, and Meta projects millions of users with the same characteristics as your seed group.

You need a seed audience of at least 100 people. From there, Meta can build new audiences that range from one million up to ten million users.

¿How big should a seed audience be for a lookalike? At least 100 people from the country where you want to advertise. If your seed lacks users from that country, Meta will likely reject the lookalike.

¿What does the lookalike percentage mean?

The percentage controls how close the new audience is to your seed. A 1% lookalike is the most similar to your origin, while higher percentages widen the reach but lower the match quality.

• 1% gives you around one million people very similar to your seed.
• 4% pushes the reach up to roughly 7.2 million, but with less precision.
• Higher percentages mean broader reach with looser matching criteria.

The recommendation is to stay between 1% and 3% and unify the ranges, for example from 0% to 3% in a single audience. You can also create three separate lookalikes (0–1%, 1–2%, 2–3%) and test them against each other, but only if you want to compare performance. For most campaigns, a single 1% lookalike works as your baseline target [09:50].

¿When should you use custom vs lookalike audiences?

Each audience type solves a different problem in your funnel, so picking the right one depends on whether you want to retarget, exclude, or prospect.

¿What are the main uses of a custom audience?

Custom audiences shine in three scenarios. First, when you need to communicate something specific to a narrow group, like sending a promo only to people who added to cart but did not buy. Second, when you want to exclude users from a cold campaign, for example removing past buyers from a prospecting effort. Third, when you need a high quality origin to build a lookalike audience.

¿When does a lookalike audience make sense?

Use lookalikes when you are ready to scale, test new angles, and increase ad spend. They also fit perfectly for quality prospecting, when you want to reach cold users but still maintain a strong match to your best customers. If your seed origin is weak, do not build a lookalike yet, because the origin is everything.

Best practices for custom and lookalike audiences

A few habits will keep your audiences clean and your campaigns efficient.

• Do not mix origins inside a single custom audience. If it is a buyers list, keep only buyers, not checkout or cart events.
• Match the time window to your funnel. Use short windows (3 to 7 days) for high intent retargeting and longer windows (30, 60, 90, even 180 days) for general exclusions.
• Use clear naming so you can identify each audience at a glance.
• Always exclude the final conversion in cold campaigns. In a sales campaign, exclude past buyers; in a lead campaign, exclude people who already submitted their info.
• For lookalikes, start with low percentages, ideally 1%, and only expand if performance drops.
• Test multiple origins against each other in separate campaigns: buyers, add to cart, long page visitors, and lookalikes built with LTV vs without it.

Remember that the time range you choose changes everything. Talking to someone who abandoned a cart seven days ago is not the same as reaching out to someone who did it a month ago. The intent decays fast, and your audiences should reflect that.

¿Have you tested a lookalike built on a lifetime value seed against a regular buyers seed? Share which one converted better in your account.