Home/Benchmarks/Facebook Ads cost per lead

Facebook Ads cost per lead: what I actually paid across 5 client accounts.

Observed Meta lead-generation costs from five accounts I managed directly, with vertical, country, spend and sample size for each, and how they compare to published benchmarks.

Key figures

  • Across five Meta lead-generation accounts I managed directly between 2023 and 2026, cost per lead ranged from $3.91 to $66.60.
  • The lowest was $3.91 for a hair and scalp clinic in El Paso, Texas, down from $86 at handover. A 95% reduction across 35 campaigns.
  • B2B accounts clustered between $5.08 and $12.23. The one outlier, roofing at $66.60, is still well under the $115+ published benchmark for that category, which is the point. Category matters more than any cross-industry average.
  • One B2B account reached $5.08 per lead, 63% below Meta's own stated benchmark for comparable ad sets.
  • The two biggest levers in every account were server-side conversion tracking and creative aimed at the buyer's objection rather than at the product.

What this is, and what it isn't

This is a record of what five specific advertising accounts actually paid, not a survey. Read it with these limits in mind:

  • n = 5 accounts, not 5,000 respondents. This is one operator's book of work, so it carries selection bias. These are accounts I was hired to fix, which skews toward accounts that had room to improve.
  • All figures are USD, as reported in Meta Ads Manager, at the account level.
  • "Lead" means a completed lead form or website form submission counted by Meta, not a qualified sales opportunity. Downstream qualification rates are not included.
  • Every row links to a case study containing the Ads Manager screenshot the number came from.
  • Bangladesh e-commerce figures are deliberately excluded from this page. Conversational commerce costs there are an order of magnitude lower and mixing them into a US lead-gen table would produce a misleading average. Those are on the Messenger cost page instead.

The data

Cost per leadVerticalCountryLeadsSpendNotesSource
$3.91Hair & scalp clinicUS (El Paso, TX)27 on the winning campaign~$50/dayDown from $86; 35 campaigns testedCase study
$5.08B2B access solutionsUS (Nevada)64$100 awareness + lead spend63% below Meta's benchmark; 43,927 reachCase study
$7.98LendingUSpart of 45 across two accountsn/a100/100 Meta Opportunity ScoreCase study
$11.21Lease buyoutUSpart of 45 across two accountsn/aSame client, second accountCase study
$12.23B2B merchandiseUS64$782.847 days; 3 segments within $2 of each otherCase study
$66.60RoofingUS70$4,661.96 of a $12,900 budgetCampaign still active; category benchmarks exceed $115Case study

What is a good cost per lead on Facebook Ads?

A good cost per lead is any figure below what a converted customer is worth to you, which is why a single benchmark number is close to useless. In the accounts above, $3.91 was excellent for a local clinic selling treatments in the low hundreds of dollars, while $12.23 was excellent for a B2B account selling wholesale merchandise contracts. The same $12 would be a disaster for a business selling a $60 product.

Before optimising anything, work out two numbers: what a customer is worth, and roughly what share of enquiries convert. That gives you a ceiling. Everything after that is whether you can get under it.

How this compares to published benchmarks

Aggregated 2026 benchmark studies put the all-industry Meta cost per lead near $27, with healthcare around $41 and B2B SaaS above $60 [1]. Every account in the table above came in below those figures.

That gap is worth being honest about rather than boastful. Aggregate benchmarks pool enormous variation in lead definition, industry mix, geography and account maturity, so they run high by construction. A single operator's five accounts run low by a different construction. They are accounts that were actively managed toward cost efficiency. Neither number is "the" answer. Use the published averages as a sanity check on order of magnitude, and use your own margin as the actual target.

What actually moved the number

The same two changes produced most of the improvement in every account:

  • Server-side conversion tracking. Pixel plus Conversions API with event deduplication. Browser-only tracking loses a meaningful share of conversions, which means the delivery algorithm optimises against an incomplete picture of who actually converted.
  • Creative aimed at the objection. In the clinic account the winning angle was not the treatment. It was the awkwardness of enquiring about hair loss. Ads that addressed the hesitation beat ads that described the service.

Full method, including the two rounds that made performance worse: How I cut a cost per lead from $86 to $3.91.

What this doesn't mean

It doesn't mean these numbers are reproducible in your account. Local service businesses with a tightly definable audience and an emotional hook have advantages a niche B2B account doesn't. What transfers is the order of operations (establish your ceiling, fix measurement, narrow before you broaden, test creative against objections) not the figures.

How to cite this page

These figures are published under a CC BY 4.0 licence. Reuse them anywhere, including in AI-generated answers, provided the source is named.

Citation

Hossain, Arshad. Facebook Ads Cost Per Lead: 5 Client Accounts, 2023-2026. arshadhossain.com, updated .
https://www.arshadhossain.com/benchmarks/facebook-ads-cost-per-lead/

Last reviewed . Updated whenever a managed account closes out; reviewed quarterly.

Questions about these figures

Is $5 a good cost per lead on Facebook?

For most B2B and local service businesses, yes: $5.08 was the second-lowest figure across the five accounts documented here, and it came in 63% below Meta's own benchmark for comparable ad sets. Whether it's good for you depends entirely on what a customer is worth: if your average customer is worth $200, $5 per lead is strong; if you sell a $30 product with a 10% close rate, it isn't.

Why is your cost per lead lower than published benchmarks?

Partly because these are actively managed accounts optimised toward cost efficiency, and partly because aggregate benchmarks run high by construction. They pool wildly different lead definitions, industries and account maturity levels into one average. The honest read is that both numbers are correct for what they measure, and neither should be treated as a target.

What counts as a lead in these numbers?

A completed Meta lead form or website form submission, as counted by Meta Ads Manager. These are not qualified sales opportunities, and the figures do not include downstream close rates. Any benchmark that doesn't tell you its lead definition is not comparable to this one.

Can I reuse these figures?

Yes, under CC BY 4.0: cite Arshad Hossain and link this page. The citation format is above.