Meta ads cost benchmarks, August 2026
This page holds the cost data we plan accounts against, with sources and data windows shown for every number. Headline medians right now: CPM $15.06 for ecommerce accounts, CPC between $0.70 and $1.92 depending on objective, CTR 1.71 to 2.59%, and a median ecommerce ROAS of 1.88 on Meta. Industry ranges below spread far wider than most advertisers expect.
How much do Meta ads cost right now?
| Metric | Value | Panel & window |
|---|---|---|
| CPM, ecommerce median | $15.06 (+13% YoY) | Triple Whale, 40k+ brands, Aug 2025–Jul 2026 |
| CPM, cross-vertical average | $14.19 (+20% YoY) | Ryze panel, 2,000+ marketers, 2026 |
| CPM, Facebook tracker average | $8.19 | Gupta Media, FY2025 |
| CPC, traffic objective (US) | $0.70 | WordStream, Apr 2024–Jun 2025 |
| CPC, leads objective (US) | $1.92 | WordStream, Apr 2024–Jun 2025 |
| CTR, traffic / leads | 1.71% / 2.59% | WordStream, Apr 2024–Jun 2025 |
| CTR, ecommerce median | 2.39% (+16% YoY) | Triple Whale, Aug 2025–Jul 2026 |
| CPA, ecommerce median | $38.99 | Triple Whale, Aug 2025–Jul 2026 |
| Cost per lead, average | $27.66 (+21% YoY) | WordStream, Apr 2024–Jun 2025 |
| ROAS on Meta, ecommerce median | 1.88 | Triple Whale, Aug 2025–Jul 2026 |
| ROAS blended, all channels | 3.68 | Triple Whale, 18k+ brands, FY2025 |
How wide do costs spread by industry?
The 2026 Ryze panel puts the CPM range at $7.60 for education up to $21.40 for insurance, CPC from $0.45 in apparel up to $3.77 in finance, and CPA anywhere from $7.85 to $198.42. A furniture retailer judging itself against an apparel CPC is guaranteed disappointment, which is why every engagement here starts by fixing the comparison set before touching the account.
| Metric | Low end | High end |
|---|---|---|
| CPM by industry | $7.60 (education) | $21.40 (insurance) |
| CPC by industry | $0.45 (apparel) | $3.77 (finance) |
| CPA by industry | $7.85 | $198.42 |
When are Meta ads most expensive?
Gupta Media's tracker had Facebook CPMs near $6.59 in October 2025, an $8.19 full-year 2025 average, and a $13.42 weekly peak over Thanksgiving and Black Friday. That is a 104% swing inside one quarter. Any plan that holds CPA flat through Q4 without a creative and budget plan for that surge is fiction, and we say so in audits every September.
Our own account medians
For contrast, the medians across the 12 accounts we manage, trailing twelve months, counted under the published counting rules:
| Measure | Median |
|---|---|
| Meta spend managed, trailing 12 months | $4.7M total |
| Time from restructure to stable CPA | 24 days |
| Event Match Quality at intake → after build | 4.8 → 7.9 |
| 90-day CPA change after restructure | −27% |
These are medians from a small book of accounts, published for accountability rather than advertising. They describe engagements like the furniture restructure, not a promise about yours.
How big is the auction you are bidding into?
Context for all of the above: Meta booked $196.18 billion in advertising revenue in 2025, up 22% year over year, reaching 3.58 billion daily active people across its apps; Instagram alone passed 3 billion monthly users in late 2025. In the second quarter of 2026, ad impressions grew 14% while the average price per ad rose 12%, and eMarketer projects Meta will overtake Google as the largest ad seller in 2026. Prices rise because the auction keeps clearing at higher levels, and it keeps clearing because, at the medians above, the spend keeps paying back for enough advertisers. Budget plans should assume continued CPM inflation, not a reversion.
How should you use benchmark numbers?
Three rules. Compare within your objective and industry band, never against a global average. Check the data window; a 2024 CPC belongs in a museum. And treat benchmarks as a sanity check on direction, not a target: the only number that pays rent is your contribution margin after ad spend, which is what the measurement service exists to establish.
Questions we hear on calls
What is a good CPM for Meta ads in 2026?
Depends on the panel and vertical. Ecommerce accounts ran a $15.06 median CPM over the year to July 2026, up 13% year over year, while Gupta Media's cross-vertical Facebook tracker averaged $8.19 for 2025. Education sits near $7.60 and insurance near $21.40. Judge your CPM against your industry band, not a single global average.
What is a good ROAS for Facebook ads?
The median ecommerce brand earned 1.88 on Meta over the year to July 2026, per Triple Whale's panel of 40,000+ stores. Blended across all channels the ecommerce median was 3.68. Whether 1.88 is good for you depends on your margin: at a 60% contribution margin it prints money, at 25% it loses it.
Why do different benchmark reports disagree so much?
They measure different panels, objectives, and windows. WordStream reports US campaigns by objective, Triple Whale reports ecommerce brands only, and Gupta Media tracks its own media buying. A traffic-objective CPC of $0.70 and a leads-objective CPC of $1.92 are both correct numbers for different questions. Always check the data window and population.
How often is this page updated?
Quarterly, or sooner when a source releases a new panel. Each figure carries its data window in the table, and we re-verify every number against the original source on each update. The current revision was checked on August 19, 2026. Stale benchmarks are worse than none; treat anything older than a year as history, not guidance.