Competitor Lookalike vs Custom Audience: 2-3x Better

The Core Difference: Intent and Warmth

Custom audiences and competitor lookalike audiences are not interchangeable. A custom audience is your own customer list—people who already know you, bought from you, or gave you their phone number. A lookalike audience is built by Facebook studying your best customers and then finding strangers who match their behavior, interests, and demographics across the Meta network.

The conversion gap is real: custom audiences convert 2–3x higher than lookalikes. Why? Because your existing customers have stated intent. They've already raised their hand. A lookalike is a cold guess, no matter how smart Facebook's algorithm is.

Here's a concrete example: A plumbing company in Denver uploaded 680 customers to a custom audience and ran a lead form ad. Over 4 weeks, the form saw a 16% completion rate at a $28 cost-per-lead. The same company then seeded a lookalike audience from those 680 customers. The lookalike delivered 3x more impressions and 2.5x more clicks—but only a 5% form completion rate and a $72 cost-per-lead. Same creative, same offer, same budget split. The lookalike found more people; the custom audience found better people.

This does not mean you should abandon lookalikes. It means you need both, running at the same time, with clear expectations about what each does.

When Custom Audiences Win: High Conversion, Low Volume

Custom audiences are your money maker. If you have 200+ customers, this is your primary hunting ground.

Uploading a custom audience is straightforward: export your customer list (email or phone number) from your CRM or spreadsheet, go to Audiences → Create Audience → Custom Audience, select 'Customer List,' and upload the CSV. Facebook hashes the data and matches it to user profiles. You'll typically match 20–60% of your list to active Meta users, depending on your data quality and customer age.

A matched custom audience of 500 customers (representing maybe 250 actual Meta users after matching) is small but potent. You're not paying for reach—you're buying certainty. Your cost-per-click drops, your form completion climbs, and your cost-per-lead tightens.

Real numbers: An HVAC company in Phoenix with 420 matched customers saw these results over 6 weeks:

• Budget: $1,200
• Impressions: 8,400
• Clicks: 340
• Cost-Per-Click: $3.53
• Lead forms completed: 62
• Cost-Per-Lead: $19.35

That $19 CPL is well below the regional HVAC average of $35–$50 (according to national benchmarks). Why? Because every person seeing the ad is a previous customer or warm referral, and they already trust the brand.

The downside: custom audiences are finite. Once you've saturated your existing customer base—roughly 4–6 weeks of ads to the same small group—frequency fatigue sets in. People see your ad 8+ times and stop clicking. At that point, you must either refresh creative or expand to a lookalike.

When Lookalike Audiences Win: Volume and Reach

Lookalikes are your growth engine. They find new customers who fit the profile of your best ones, even if they've never heard of you.

Building a lookalike is also simple: Audiences → Create Audience → Lookalike Audience. Pick your source (your custom audience, your website pixel, or a competitor's Facebook page fan base). Choose your country and lookalike size. Facebook offers 1% (most similar) through 10% (broadest match). A 1% lookalike of 500 customers might reach 1.2 million people; a 5% might reach 3.5 million.

Real example: An electrician in Austin seeded a 5% lookalike from 280 existing customers and ran the same $1,200 budget. Results over 6 weeks:

• Impressions: 34,200
• Clicks: 890
• Cost-Per-Click: $1.35
• Lead forms: 53
• Cost-Per-Lead: $22.64

That CPL is higher than the custom audience version (the electrician's existing customers converted at $18), but the lookalike reached 4x more people and still landed 53 leads. The volume made up for the lower conversion rate. The lookalike is cold-to-warm outreach, not hot-to-hot like custom audiences.

A common mistake: running lookalikes without first building a strong custom audience. If you seed a lookalike from 50 random website visitors or a list of competitors' followers, Facebook has weak signals to work with. Your lookalike will be scattered, expensive, and low-intent. Always seed lookalikes from your own customer list, not from thin external data.

The Math: Why Stack Both, Not Choose One

The best strategy is not either/or; it's both/and. Here's why the numbers work:

Scenario: $3,000 monthly budget for a contractor.

Option 1: All-custom (wrong). Spend $3,000 on custom audience. Land 150 leads at $20 CPL. Audience saturates after 3 weeks; CPC rises 40%, CPL climbs to $28. By week 4, you're burning money on frequency.

Option 2: All-lookalike (wrong). Spend $3,000 on lookalike. Land 100 leads at $30 CPL. You reach more people (9 million impressions vs. 15,000), but conversion is weak. Cost per lead is higher, and you never build exclusivity or warm momentum.

Option 3: Stack both (right). Split: $1,500 to custom audience, $1,500 to lookalike. Custom lands 75 leads at $20 each. Lookalike lands 50 leads at $30 each. Total: 125 leads at $24 average CPL. You've captured your warm audience AND expanded to cold outreach, all while staying under the fatigue threshold with each segment.

Why does stacking win? Because lookalikes improve over time. Once you add the 50 lookalike leads back into your customer list and re-upload, the next month's lookalike is smarter. You've given Facebook more data about what a good-fit customer looks like. Your second-month lookalike might convert at 7% instead of 5%. Your third month at 8%.

Seeding Lookalikes: Your Own Customers vs. Competitors

There are two ways to build a lookalike: from your own customer base, or from a competitor's Facebook page fan base.

Seed from your own customers (best method): Upload a custom audience of your top 300+ customers (or all customers if you have fewer than 1,000). Then create a lookalike from that audience. Facebook learns from real buyers and finds more real buyers. This is the highest-intent lookalike you can build. Cost-per-lead typically runs 30–50% higher than custom, not double or triple.

Seed from competitor's followers (secondary method): Go to your competitor's Facebook page, note their username, then create a lookalike using their fan base. This works—Facebook has millions of data points on their followers—but the match is weaker. You're guessing at who a competitor's followers are, not knowing who your buyers are. This method is useful if you're brand-new and have no customer list yet. See our deeper guide on competitor audience targeting for the full setup.

Many contractors make the mistake of seeding lookalikes from a thin list: 30 website visitors, 15 email newsletter subscribers, or a handful of recent leads. The algorithm has nothing to learn. You'll get a scattered, expensive lookalike. Build a lookalike only after you have 200+ real customers in your list.

Frequency Fatigue and Creative Decay: Why You Must Rotate

Once your custom audience reaches 4–6 weeks of continuous ads, fatigue sets in. You'll see:

• Cost-per-click rises 20–40%
• Click-through rate drops from 3–5% to 1–2%
• Form completion rate falls
• Cost-per-lead nearly doubles

The same 250 people have seen your ad 8–12 times. Diminishing returns. Your solution is not to pause; it's to rotate creative or expand your audience.

Refresh your ad copy and image every 2–3 weeks if you're targeting a small custom audience. A plumber running ads to 200 existing customers should have 2–3 versions of the ad running at the same time. Alternate them weekly. Or expand your lookalike from 1% to 2–3% to find new people before existing customers burn out.

This is where the stack strategy shines: while your custom audience rests or rotates, your lookalike keeps delivering. And the data from both feeds back into the next iteration of your custom audience. It's a flywheel.

Cost-Per-Lead Ranges: Realistic Benchmarks by Audience Type

Here's what you should expect, by trade and audience type. All figures assume a well-built lead form, geo-targeting to a metro area, and tested copy.

These ranges are realistic. If your custom audience CPL is $20 and your lookalike is $35, you're in the normal zone. If lookalikes are 4–5x more expensive, your lookalike is poorly built (too broad, seeded from weak data, or targeting the wrong geo area).

For deeper benchmarks by metro area and seasonal variation, see our guide on cost-per-lead by trade and region.

When This Does NOT Work: Honest Limits

Custom audiences and lookalikes are not a silver bullet. They fail in these scenarios:

• You have fewer than 100 customers: Your custom audience is too small to match to enough Meta users. You'll reach maybe 30–50 people. Lookalikes seeded from that tiny group will be even weaker. If you're brand-new, focus on retargeting website visitors and form abandoners instead. Build your customer base first, then upload it.
• Your customer list data is dirty: If 40% of your emails are typos, old, or unverifiable, Facebook can't match them. Duplicate or malformed phone numbers cause the same problem. Before uploading, deduplicate and validate your list. A 40% match rate instead of 60% cripples your custom audience.
• Your lookalike is seeded from a competitor's fans, not your customers: Competitor lookalikes convert 3–5x worse than customer-seeded ones. They work for brand awareness or rough reach, not for leads. If you have no customer list, they're a placeholder. But don't expect the conversion of a real custom audience.
• Your audience is geo-constrained and already saturated: A roofing company in a town of 40,000 people has a finite pool of customers and lookalikes. After 8–12 weeks of ads, your custom audience is tapped. Your lookalike finds few new people because the metro area is small and you've already reached most of them. In micro-local markets, expand your radius or accept audience saturation and rotate seasonal campaigns.
• Your offer or call-to-action is weak: A perfect audience means nothing if your headline says