Quick answer: Build a custom audience from your pixel's Purchase event or a clean CRM export. Use 1,000 to 5,000 recent buyers as your source. Create a 1% lookalike first. Test 1%, 5%, and 10% in parallel to find your sweet spot.
---
What a Lookalike Audience from Purchasers Is
A lookalike audience finds new people who share key traits with your existing buyers. Facebook analyzes your source audience and identifies patterns across interests, behaviors, and demographics. Then it builds a fresh pool of matched users, automatically excluding anyone already in your source.
Your purchasers are the most powerful seed you can use. They already converted. That signal is hard to beat.
Why Target Purchaser Lookalikes
Finding high-intent customers who match your existing buyers
People who bought from you share something. Price sensitivity, product interest, buying habits. Facebook identifies those overlapping traits and finds more people who fit the same profile. You are not guessing at an audience. You are scaling a proven one.
Expanding beyond your current customer base
Retargeting hits a ceiling fast. Your purchaser list is finite. Lookalikes unlock a much larger pool of potential buyers without walking away from what already works.
Scaling proven customer profiles
A 1% purchaser lookalike in a major market can reach hundreds of thousands of people. Same profile. New faces. That is how you scale a product that converts.
The Two Main Sources: Pixel Events vs. Customer Lists
Using Facebook Pixel conversion data from purchases
Per the Meta Business Help Center, you can create a custom audience directly from your pixel's Purchase event. Select the Purchase event type, set a lookback window (30, 90, or 180 days), and save the audience. Facebook refreshes this audience based on purchases tracked over the past 180 days.
This approach works best when your pixel fires consistently and you have enough purchase volume.
Uploading a customer list from your CRM or database
Meta also accepts direct customer list uploads. You can include emails, phone numbers, names, or addresses. Meta hashes the data and matches it against its user database. Per the Meta Business Help Center on Customer List Custom Audiences, uploaded lists can be highly accurate when your customer data is clean and current.
Comparing source quality and accuracy
Pixel data updates automatically. Customer lists require manual uploads. If your pixel is well-configured and fires reliably on purchases, pixel-based audiences tend to stay fresher. If your pixel tracking has gaps, a clean CRM export often wins on quality.
Use whichever source reflects your real buyers most accurately. Both work.
Building Your Source Audience
Minimum source audience size (100 to 5,000 members)
Per Meta's developer documentation, a custom audience used as a lookalike seed needs at least 100 members. Meta recommends between 1,000 and 5,000 for best results. Too small and the signal is thin. Too large and precision can dilute.
Quality matters more than raw size
A list of 800 confirmed purchasers beats a list of 10,000 mixed-intent visitors every time. The signal you feed in determines the signal you get out. Focus on verified buyers, not broad traffic.
Note: Meta flags custom audiences that suggest specific health conditions or financial status. Those audiences cannot run ads. Per Meta's policy guidelines, always verify your source audience does not violate these restrictions before using it as a lookalike seed.
Recency: using recent purchasers for better targeting
Recent purchasers reflect your current product and market. Someone who bought last month is more relevant than someone from two years back. Keep your lookback window tight. Thirty to 90 days is a strong starting point.
Creating the Lookalike Audience: Step by Step
Navigating Ads Manager to create a lookalike
Open Ads Manager. Go to the Audiences section. Click "Create Audience" and choose "Lookalike Audience." Select your source custom audience. From there, choose your target locations and set your percentage. The audience will take one to six hours to fully populate.
Choosing location and percentage (1% vs. 5% vs. 10%)
Per Meta's documentation, lookalike audiences only include people from the locations you specify during setup. Choose your target country or region carefully. Percentage options range from 1% to 20% in 1% increments. Smaller percentages are more precise. Larger percentages increase reach.
Understanding "Similarity" vs. "Reach" modes
At 1%, you get the top 1% of people in your target location most similar to your source. Meta calls this the "Similarity" end of the range. At 5% and above, you shift toward "Greater Reach," which broadens the pool but loosens the match. Facebook refreshes lookalike audience membership every three days when the audience is attached to an active ad set.
Percentage Selection: Similarity vs. Reach
1% lookalike: highest precision, smallest audience
Start here. A 1% purchaser lookalike is the tightest match available. Conversion rates are often strongest at this tier. It is the right place to validate before you scale.
5% and 10% lookalikes: broader reach, slightly broader match
Once your 1% audience shows solid returns, expand. A 5% lookalike delivers more volume. A 10% lookalike goes broader still. Match quality loosens slightly, but reach increases significantly.
Testing multiple percentages in parallel
Meta recommends testing multiple tiers at the same time. Run 1%, 5%, and 10% in separate ad sets with equal budgets. Let performance data reveal which tier delivers the best results for your offer and market.
Using Your Lookalike in a Campaign
Adding the lookalike audience to your ad set
In your ad set, go to the Audiences section and select your saved lookalike. Meta automatically excludes your source audience members from seeing the ads. You will not waste budget on existing buyers.
Combining with other targeting signals
You can layer interest or behavior filters on top of your lookalike. Keep it light. Heavy layering shrinks the audience fast. A purchaser lookalike is already a strong signal. Let it work.
Measuring performance and iterating
Watch cost per purchase, return on ad spend, and frequency. If frequency climbs and results drop, your audience is saturating. Move to a higher percentage tier or refresh your source with more recent purchasers.
---
Or let Coinis do it.
From a product URL to a live Meta campaign. AI-generated creatives. On-brand copy. Direct publish to Facebook and Instagram. Real performance reporting. All in one platform.
Start free. Upgrade when you're ready.
15 AI tokens a month. No credit card.
Frequently Asked Questions
How many purchasers do I need to create a lookalike audience on Facebook?
Meta requires a minimum of 100 people in your source audience. For best results, Meta recommends between 1,000 and 5,000 verified purchasers. Prioritize recent, confirmed buyers over a larger mixed list. Quality drives match accuracy more than raw size.
Should I use pixel purchase events or a customer list as my lookalike source?
Both work well. Pixel purchase events update automatically and stay current with recent buyers. A customer list gives you precise control over data quality but requires manual uploads. If your pixel fires cleanly on purchases, start there. If tracking has gaps, a clean CRM export is more reliable.
What percentage lookalike should I use for purchasers?
Start at 1%. It targets people most similar to your purchasers and typically delivers the strongest conversion rates. Once you see positive results, test 5% and 10% in separate ad sets. Running multiple tiers in parallel is the fastest way to find the right balance of precision and scale for your market.
Does Facebook exclude my existing customers from the lookalike audience?
Yes. Meta automatically excludes people in your source audience from the lookalike it creates. Your existing buyers will not appear in the new lookalike pool, so you will not waste budget targeting people who already purchased.