One of the hardest parts of advertising is finding new people likely to become customers. Lookalike audiences help solve this. By giving an ad platform a source audience, such as your best customers, the platform finds other people with similar characteristics and behaviours. It's a way to scale what already works.
How lookalike audiences work
On Meta, you create a lookalike audience by choosing:
- A source audience: customer lists, website visitors, people who engaged with your content or app users.
- A location: where you want to find similar people.
- An audience size: usually a percentage of the population in that location. Smaller percentages are more similar to the source; larger ones offer more reach.
Meta then identifies people who share characteristics with your source audience.
Lookalikes in a changing advertising landscape
Meta increasingly encourages broad targeting and Advantage+ audience options, where its algorithm decides who to show ads to, using your inputs as suggestions. Lookalike audiences can still be used, often as signals to guide this automation rather than strict boundaries. Results vary by account, so test lookalikes against broad targeting.
Choosing a good source audience
Lookalikes are only as good as their source. Strong sources include:
- Customers with high value: your best clients by revenue or repeat purchases.
- Qualified leads: leads that your sales team confirmed as good.
- Purchasers: people who bought online.
- High-intent website visitors: visitors to pricing or booking pages.
- Engaged video viewers: people who watched most of your videos.
Weaker sources include all website visitors or all page likes, which may include many low-intent people.
Source audience size
Meta recommends a source audience of reasonable size, often at least several hundred people, for better results. Quality matters more than size, though. A smaller list of great customers may produce better lookalikes than a large list of random contacts.
Uploading customer lists
You can upload customer lists to Meta as custom audiences. Data such as email and phone is hashed before matching. Make sure you have the right to use customer data this way, and follow privacy rules and Meta's terms.
Lookalike size
- 1–2%: closest match; smaller reach.
- 3–5%: balance of similarity and reach.
- 6–10%: broader reach; less similar.
In smaller markets like the UAE, even a 1% audience may be relatively small. Test different sizes.
Using lookalikes on other platforms
- Google Ads: Google has moved away from traditional similar audiences towards automated optimised targeting and audience signals, especially in Performance Max and Demand Gen.
- TikTok: offers lookalike audiences based on custom audiences.
- LinkedIn: offers audience expansion and predictive audiences.
- Snapchat: offers lookalike audiences.
Features change frequently, so check current options on each platform.
Testing lookalikes
Run tests comparing:
- Lookalike audiences of different sources.
- Different lookalike sizes.
- Lookalike vs broad or Advantage+ targeting.
Use the same creative and budget to keep the comparison fair. Measure cost per lead, lead quality and cost per customer.
Keep sources fresh
Update customer lists and source audiences regularly. As your customer base grows and changes, your lookalikes should reflect your current best customers.
Exclusions
Exclude existing customers, recent converters or current leads from prospecting campaigns, so you're not paying to reach people who already know you.
Common mistakes
- Using poor-quality source audiences.
- Expecting lookalikes to fix weak creative.
- Never refreshing source data.
- Not comparing with broad targeting.
- Ignoring privacy and consent requirements.
Combine lookalikes with strong creative
Even the best audience won't convert if the ad doesn't resonate. Pair lookalike audiences with creative that speaks directly to the needs and motivations of your best customers. Use insights from your source audience, such as what they bought, why they chose you and what they value, to shape your messaging. Lookalikes find the right people; creative persuades them.
Multiple sources for multiple services
If you offer several services, create separate source audiences for each, such as customers who bought renovation services and customers who bought maintenance plans. This produces lookalikes that match each service's ideal customer more closely. Refresh lists regularly to keep them relevant.
Final thoughts
Lookalike audiences help you find new people who resemble your best customers. Choose high-quality sources, test sizes and approaches, refresh data and combine lookalikes with strong creative. As platforms lean more on automation, treat lookalikes as useful signals and let testing guide your decisions.
Buzz Bee Studio builds audience strategies for Meta, TikTok and Google campaigns. Get in touch to reach more customers like your best ones.



