The "1% is always better" rule sounds logical. More similar to your source, higher precision, better results. Founders repeat it like gospel.
It's not always true — and for beginners, it's often the wrong call.
Whether Lookalike 1% or 5% wins depends on two variables most people skip before they choose: the size of your source audience and your daily budget. Get those wrong and the percentage doesn't matter. You'll burn spend either way.
How Lookalike actually works
Meta takes your source audience (a customer list, a Pixel-based Custom Audience, or video viewers), analyzes what those people have in common, then finds the Facebook users in your target country who most closely match that pattern.
- 1% = top 1% who most resemble your source ≈ 500,000 people in Thailand
- 5% = top 5% ≈ 2.5 million people in Thailand
Bigger percentage = bigger pool = less precise match. That's the entire trade-off. Neither is automatically better. Both can win or lose depending on your situation.
When Lookalike 1% wins
You're under ฿1,500/day
A pool of 500,000 people is more than enough at small budgets. Meta doesn't need to cast a wide net. It needs to find the right people. 1% gives it a smaller, higher-quality pond to fish in, which typically means lower CPM and higher conversion rate compared to 5% at the same spend level.
Your source came from actual buyers
A customer list of 1,000 people who paid money is a fundamentally different signal than 1,000 people who liked your Facebook page. With a buyer-based source, 1% Lookalike surfaces patterns that correlate with purchase intent, not just interest or curiosity.
You're starting with no prior targeting data
1% Lookalike from a customer list is your best cold-start option. It's more targeted than manual Detailed Targeting because Meta is pattern-matching against people who already opened their wallets. Not people who just share category interests.
When Lookalike 5% wins
Your source audience is under 1,000 people
Meta needs a minimum of 100 to build a Lookalike, but anything under 1,000 is shaky. The algorithm doesn't have enough signal, so it guesses. Lookalike 3–5% with a thin source performs more consistently than 1% with the same thin source.
You're scaling past ฿2,000/day and frequency is climbing
A 500,000-person pool saturates fast at higher daily budgets. When your 7-day frequency on 1% Lookalike hits 2.5–3.0, the audience is tired of seeing your ad. Don't kill the ad set — stack it. Add a 3% or 5% Lookalike as a separate ad set in the same campaign. Let them run in parallel.
Your product has mass-market appeal
If you sell something most people genuinely need (everyday personal care, food, household items), 5% gives you broader reach while still being far more targeted than pure Broad targeting. The precision trade-off is worth the pool size when your product fits a wide audience.
The trap nobody talks about
Lookalike is an audience. It isn't a guarantee.
Beginners treat Lookalike like a magic filter that surfaces ready-to-buy customers. It doesn't. It finds people who resemble your past customers. If your creative doesn't hit their specific problem, they scroll past just like everyone else.
Lookalike 1% with a weak hook loses to Lookalike 5% with a hook that nails the pain point. The audience gets you in the room. The creative closes the sale. Most beginner campaigns that "don't work" have a creative problem dressed up as a targeting problem.
Mixed source audiences kill precision
If you merge "customers who bought Product A at ฿500" and "customers who bought Product B at ฿5,000" into one source, Meta finds a blurry average pattern that doesn't accurately represent either group. Split your sources if the purchase behavior or customer profile is meaningfully different.
Quick reference
| Situation | Use |
|---|---|
| Budget < ฿1,500/day, source 1,000+ people | Lookalike 1% |
| Source audience < 1,000 people | Lookalike 3–5% |
| Scaling to ฿2,000+/day, 1% frequency rising | Stack: 1% + 3% in separate ad sets |
| Mass-market product, source from page engagement | Lookalike 3–5% |
| No customer data at all | Detailed Targeting first, build the source over time |
What to do next
Before you pick a percentage, check your source. Go to Audiences in Meta Business Suite and count how many people are in the Custom Audience you're planning to use as the source.
- Under 500 people: start with 5%, and prioritize growing that list before anything else.
- 1,000+ people: start with 1%, then monitor 7-day frequency weekly.
Open AdBlueprint and check the Targeting strategy field in your Campaign plan. The recommendation already factors in your source size and daily budget. If you're overriding it, make sure you know which condition above you're actually optimizing for. And why.