LAL effectiveness vs interest targeting on Meta and Google.
Inputs
LAL viability
Estimated LAL CAC vs interest—
Source audience adequate?
—
Estimated LAL CAC
—
Monthly conversion delta
—
LAL similarity comparison
Similarity
Reach
Est. CAC
Best for
About this calculator
Lookalike audiences remain one of paid social\'s highest-leverage targeting tactics, but their effectiveness depends heavily on source audience quality and similarity setting. The post-iOS14 degradation (25-40% performance drop) made source quality matter even more — Meta needs strong signal to build accurate lookalikes from limited tracking data.
This calculator estimates LAL CAC versus interest targeting based on source audience size, source quality, and similarity setting. The model uses observed performance differentials: high-LTV customer source audiences typically produce 20-35% lower CAC than interest targeting; all-visitor source audiences produce 5-15% lower; site-visitor source often barely outperforms interest targeting at scale.
The strategic playbook: build a high-LTV source audience (top 20% of customers by spend) — this is the highest-quality LAL seed. Upload to Meta via CRM-list custom audience. Build 1%, 3%, and 5% LAL audiences. Run all three in separate ad sets at first; budget will gravitate toward the best performer. Refresh source audience monthly as new high-LTV customers join. Most successful programs run a portfolio of LAL similarities to balance reach and quality.
Pair with the Customer Acquisition Cost calculator (the metric LAL most affects), Tracking Gap Estimator (post-iOS14 measurement context), Whitelisting ROI calculator (LAL + whitelisted creative is high-leverage combination), and First-Party Data Value calculator (LAL\'s effectiveness depends on first-party data quality). Most successful brands review LAL performance vs interest targeting quarterly — the right mix shifts as audience saturates and creative changes.
Frequently asked questions
What's a lookalike audience?
Meta or Google takes a "source audience" you provide (purchasers, email list, website visitors) and builds a similar audience based on shared characteristics. Sizes from 1% (most similar) to 10% (broader reach). Used for prospecting — finding new customers who resemble existing ones.
What source audience works best?
Best to worst: high-LTV customers (top 20% by spend), all purchasers, all converters (incl. soft conversions), email subscribers, all-site visitors. The narrower and higher-quality the source, the better the lookalike — but you need 500-1,000+ source customers minimum for Meta to build a useful LAL.
1% vs 5% vs 10% LAL?
1%: highest similarity, smallest reach. Best for high-CAC products. 5-10%: broader reach, less similar. Better for high-volume cheaper products. Test all three; most ecommerce performs best with 1-3% LAL on hero campaigns and 5-10% LAL for scaling.
Are lookalikes still working post-iOS14?
Yes but degraded. Meta's lookalike quality dropped 25-40% after iOS14 limited tracking signal. Solution: feed Meta higher-quality source audiences (CRM uploads of LTV-tagged customers, server-side conversions via CAPI). Google's "similar audiences" also still work but are increasingly auto-managed via PMAX.
When should I stop using LAL?
When detailed targeting (interest-based) consistently outperforms LAL on the same creative. Some niches have such well-defined Meta interests that LAL provides no lift. Test by running 50/50 budget split between LAL and detailed targeting on the same campaign for 4 weeks; let data decide.