How much do Facebook ads cost in 2026?

How much do Facebook ads cost? Whatever the auction decides, impression by impression. Meta describes the auction as a ranking by total value, where advertiser value is your bid multiplied by the estimated action rate, the predicted chance that this person takes the action you optimize for, plus a score for ad quality. The ad with the highest total value wins the impression, and the price you pay follows from the competition for that person at that moment.

The bill is written in CPM, the cost per thousand impressions, even when the ad set optimizes for purchases or leads. Every other cost is CPM divided by a rate: CPC is CPM divided by the clicks a thousand impressions produce, which is 1,000 times CTR, so a $10 CPM at a 1% CTR gives 10 clicks and a $1.00 click. Cost per lead is CPC divided by the conversion rate, so a $1.00 click at a 5% rate is a $20 lead. When a cost moves, one of those three terms moved, and the job is to find which.

That is also why no single Facebook ads cost exists. The same CPM is cheap for a product whose ads are clicked by 3% of viewers and expensive for one clicked by 1%. The benchmarks below are starting points for a budget, not a price you can expect.

Audience sets the floor. In Lebesgue's 2026 ecommerce data, the Facebook ads CPM for United States audiences is $16.08, against $11.81 in the United Kingdom, $9.05 in Germany and $6.95 in France. A store that sells in several countries should never read one blended CPM: a shift of budget toward the US raises the account CPM without anything in the ads changing.

Objective and placement decide which auctions you enter. Gupta Media's 2025 report, which blends awareness and reach campaigns with conversion campaigns, puts Meta's average CPM at $8.37 in September 2025; Databox's median across 2,800 companies in October 2026 is $7.26. Ecommerce stores optimizing for purchases pay more, because they compete for the people most likely to buy: Birch's January 2026 guide puts their CPMs in the low teens.

Season moves everything at once. Gupta Media reports holiday CPM increases of as much as 66% around Black Friday and Cyber Monday, and its monthly table shows December 2024 at $10.83 against January 2024 at $6.05. Birch's analysis of 2021 to 2023 US campaigns found November and December CPMs about 40% above the yearly level and January about 20% below. Creative is the lever you hold: a higher estimated action rate wins auctions at a lower bid, so the ad that lifts CTR from the 1.2% bottom quartile to the 3.2% top quartile in Databox's data cuts the price of a click by more than half at the same CPM.

• Country mix: US audiences cost more than twice what French audiences cost in Lebesgue's 2026 data
• Objective: purchase-optimized ad sets compete for fewer, more valuable people
• Calendar: November and December, then the January dip
• Creative: CTR is the term that turns CPM into CPC

The lead table gathers the 2026 benchmarks that can be opened and checked. LocaliQ's 2026 report, produced with WordStream, gives $0.60 per click and a 1.93% CTR for traffic campaigns, and $1.80 per click, a 2.70% CTR, an 8.54% conversion rate and a $27.39 cost per lead for the leads objective. Each is slightly better than its 2025 figure: CPC fell from $0.70 and $1.92, CPL from $27.66. Databox's medians are lower, $0.40 per click and a $7.26 CPM, because a median ignores the expensive tail and the group includes every objective.

Three caveats apply to every row. First, an average of accounts is not a median account: a handful of large spenders in competitive categories pulls averages up. Second, the objective changes the auction: traffic clicks are cheap because the people most likely to click are not the people most likely to buy. Third, the sample decides the country mix, and a US-heavy sample reports a US CPM.

Use the benchmarks for two things only. Before a launch, they size a budget, as the next section shows. After a launch, they tell you which term is out of range: a CPM far above the Lebesgue figure for your country, a CTR below Databox's bottom quartile, or a conversion rate that leaves CPL above what a lead is worth.

LocaliQ's 2026 lead generation table spans $13.74 per lead in real estate to $61.56 for dentists, with a 2.70% CTR and an 8.54% conversion rate on average. The spread comes from the conversion rate more than from the click: real estate converts 9.95% of clicks at $1.27 each, while dentists convert 6.07% of clicks that cost $5.70. Education converts 15.87% of clicks, the highest rate in the table, which is how a $1.72 click becomes a $26.31 lead.

For stores, the lead is a purchase and the rate is far lower. Lebesgue's 2026 category figures run from 0.61% for home decor to 1.54% for food and drink, and cost per purchase across industries from $20.47 to $39.03, with beauty at $34.29 behind a $12.46 CPM, the highest CPM of the five categories in the table. A 1% purchase rate at a $1.00 click is a $100 purchase cost, which only a high order value or a measured repeat rate can carry.

A lead benchmark says nothing about lead quality. Instant forms produce cheap leads that a sales team may not reach; a lead that fills a form on your site costs more and converts further. Compare cost per qualified lead, with your CRM's definition, before deciding that a channel is cheap.

Meta's learning phase sets the floor. Its help page on learning limited says an ad set is marked learning limited when it is unlikely to receive around 50 optimization events in the week after its last significant edit. So a test budget is the cost of 50 results in seven days: weekly budget equals target cost per result times 50, and daily budget is that divided by 7. At LocaliQ's 2026 cost per lead of $27.39, that is about $1,370 a week, or about $196 a day for one ad set.

Birch's 2026 budgeting guide reaches the same place from the other end: for a first test it suggests a daily budget of 1.5 to 2 times the target CPA, and for validation a daily budget of 7 to 10 times the target CPA, which is 7 to 10 conversions a day or about 50 a week. Below the test level you learn whether the ad gets clicks; only at the validation level do you learn what a result costs.

If the business cannot fund 50 purchases a week, optimize one step up the funnel, toward add to cart or initiate checkout, where 50 events cost less, and read purchases alongside. The Facebook Ads cost calculator on this site works backward from a revenue goal, order value, conversion rate and target ROAS to the budget, orders and clicks a plan needs, so the test budget can be checked against the goal before any money moves.

• Weekly test budget = target cost per result × 50
• Daily budget = weekly budget ÷ 7
• Fewer than 50 results a week: optimize for an earlier event
• One ad set per test, so the 50 events are not split

A benchmark says what others pay. Whether you can pay it is a question for your margin. The most you can pay for an order before the sale loses money is the contribution per order: order value minus product cost, shipping, payment fees and expected returns. An $80 order that keeps 45% after those costs can carry a $36 cost per purchase, whatever LocaliQ or Lebesgue report.

The same arithmetic in ROAS terms: break-even ROAS is 1 divided by contribution margin, so 2.50x at a 40% margin and 2.00x at 50%. Databox's October 2026 median purchase ROAS of 3.16x is a profit for the store at 40% and a thin one at 30%, where break-even is 3.33x. The break-even ROAS calculator on this site returns both the break-even ROAS and the maximum CPA from one order's figures; the ROAS calculator turns a campaign's revenue and spend into ROAS, CPA and revenue per conversion.

Set the target cost per result from that ceiling, below it by the profit you want to keep, and give that target to the ad set as its cost per result goal only once the account has the volume to hold it. Until then, use the benchmark to decide whether the test is even plausible: if the industry CPL is twice what a lead is worth to you, the fix is the offer or the funnel, not the bid.

A weekly read takes four numbers in a fixed order: CPM, CTR, CPC, cost per result, for the last 7 days against the previous 7 and against your own trailing median. CPM up with CTR flat is the auction: season, a new competitor, a shift in country mix. CPM flat with CTR down is the creative: fatigue or a weaker new ad. Both flat with cost per result up is the landing page, the offer or the attribution window. The metrics calculator on this site produces the four from spend, impressions, clicks and conversions if a report gives only the totals.

Compare with the published benchmark once a quarter, not every week: it changes yearly and your account changes weekly. The weekly comparison that matters is with your own account at the same time last year, because the seasonal pattern Birch and Gupta Media describe repeats, and a 40% November CPM rise is not a problem to fix.

In Adrails, Analysis sets your Meta accounts side by side for the chosen period, with ad spend, revenue, ROAS and results in the cards and what Shopify recorded for the same days on its own row, so CPM is read against the revenue it produced rather than against a number from another business. Workflows check campaign, ad set and ad metrics every hour and send an alert to Slack, Discord or email when a value you set is crossed, so a CPM or CPA jump reaches you before the weekly read.