Facebook Ads Cost Benchmarks 2026: Real Numbers From Accounts I Actually Manage
Industry-published benchmarks are mostly garbage. Here are the actual ranges I see across e-commerce, SaaS, lead-gen, and app accounts in 2026 — what to expect, what flags a problem, and how to use them without lying to yourself.
I get asked this every week: "What's a normal CPM right now?" The honest answer is "it depends on more variables than I can list in one sentence." But people want a number, so here they are — with the caveats that make them useful instead of misleading.
These come from accounts I've managed or audited in the last 90 days. Real spend, real verticals, US Tier-1 unless noted otherwise.
Why "industry benchmarks" are usually wrong
The benchmark reports you find on agency blogs have three problems:
- They average across geos. A $4 CPM in India and a $40 CPM in the US average to $22 — a number that doesn't describe either market.
- They don't segment by funnel stage. A retargeting CPM of $35 vs a cold prospecting CPM of $35 mean different things. One indicates expensive warm inventory; the other indicates auction competition.
- They're old. CPMs in 2020 vs 2026 are barely comparable. Most cited benchmarks are 2-3 years stale and recycled.
Treat any benchmark — including the ones below — as a sanity check, not a target. Your own historical numbers (your account, your offer, last 90 days) are 100x more useful than someone else's averages.
CPM by geo (cold prospecting, May 2026)
| Region | Range | Typical | Notes |
|---|---|---|---|
| US Tier-1 | $18 – $40 | $24 | Spikes 30-50% in Q4 |
| UK | $12 – $26 | $17 | Cheaper Jan-March |
| Canada / Australia | $10 – $22 | $14 | AU CPMs more volatile |
| Western EU (DE, FR, IT, ES) | $8 – $18 | $11 | DE highest, ES lowest |
| Northern EU (NL, SE, NO, DK) | $10 – $22 | $14 | High-income, low volume |
| Eastern EU (PL, RO, HU) | $3 – $8 | $5 | Cheap CPMs, lower AOV |
| LATAM (BR, MX, AR) | $2 – $7 | $4 | BR highest of the three |
| SEA (PH, VN, ID, TH) | $1 – $4 | $2 | Volume-driven, low LTV |
| India | $1.50 – $5 | $2.50 | Massive volume available |
CPC and CTR (link clicks only)
CPC is just CPM divided by CTR. Knowing both helps diagnose whether a high CPC is a creative problem (low CTR) or an auction problem (high CPM).
| Vertical | CTR (link) | CPC |
|---|---|---|
| DTC e-commerce | 1.0% – 2.2% | $0.80 – $2.40 |
| B2B SaaS trial | 0.7% – 1.4% | $2 – $4.50 |
| Lead-gen (high-intent) | 0.9% – 1.8% | $1.50 – $3.50 |
| Mobile app installs | 1.4% – 3.2% | $0.30 – $1 |
| Education | 1.2% – 2.4% | $0.60 – $1.80 |
| Local services | 1.5% – 3% | $0.40 – $1.20 |
CPA by vertical (purchase or qualified lead)
The most variable metric, because conversion rate compounds on top of click cost. Take these as midpoints with ±50% normal variance.
| Vertical | Cold CPA | Blended CPA |
|---|---|---|
| DTC apparel ($60-100 AOV) | $28 – $48 | $22 – $36 |
| DTC beauty ($35-60 AOV) | $20 – $38 | $15 – $28 |
| B2B SaaS trial | $45 – $90 | $30 – $60 |
| Insurance qualified lead | $32 – $80 | $25 – $55 |
| Mortgage lead | $60 – $150 | $45 – $110 |
| App install (utility) | $2.50 – $7 | $1.80 – $5 |
| Online course ($200-500) | $60 – $140 | $45 – $95 |
How to use these without lying to yourself
Three rules:
Compare like-to-like. Your B2B SaaS account at $80 cold CPA is not a problem if your blended is $52 and your LTV is $1,200. Your DTC beauty account at $80 cold CPA on a $40 AOV is dead. Same number, totally different verdict.
Use them to spot anomalies, not set targets. If your CPM is 3x the typical for your vertical, something is wrong (account quality, audience overlap, broken pixel). If your CPA is 50% better than the range, that's usually a measurement artifact — Facebook is over-attributing.
Track your own 90-day rolling baseline. Your account's historical numbers from 90 days ago are the only benchmark that matters. Industry numbers are for sanity-checking and conversations with clients.
Seasonal multipliers
- Q4 (Oct-Dec): CPMs ×1.3-1.6 across all verticals. Plan for 30-50% efficiency degradation.
- January: CPMs drop 15-25% as advertisers pull back post-holidays. Cheap acquisition window.
- Election cycles (US): Q3 of presidential election years add ×1.1-1.2 to all US CPMs.
- Major launches (iOS, Black Friday): Tier-1 CPMs spike for 5-10 days then normalize.
- Summer (Jun-Aug): Mixed — some verticals quiet (B2B), others spike (DTC, travel).
FAQ
Why is my CPM 50% higher than your benchmark?
Usually one of: small audience size, low ad quality (look at Quality Ranking diagnostics), high frequency on the same audience, narrow geo targeting, or Q4 timing.
Are these benchmarks for cold or all traffic?
Cold prospecting unless noted. Retargeting CPMs run 30-60% higher because the audience pool is smaller and quality is more concentrated.
Does account spend level affect benchmarks?
Yes. Accounts under $500/day often see CPMs 20-40% higher because Facebook can't optimize at low volume. Accounts over $5k/day usually see better numbers within their tier.
How often should I re-baseline?
Quarterly. CPMs drift 10-20% per quarter even without algorithm changes. A target you set in January is probably wrong by April.
Bottom line
Use these to spot when something's broken, not to set goals. Your own 90-day rolling numbers are the real benchmark. Industry averages exist mostly to help you have a defensible answer when a client asks "is this normal?" — and to remind you that everyone's metrics get worse in Q4 whether they admit it or not.