Meta ads cost calculator: how it works
Meta runs ads on Facebook and Instagram from one system, and you are normally charged in one of two ways: per 1,000 impressions (CPM) or per click (CPC). This calculator turns a budget into impressions, reach, clicks and results using those two prices.
Choose a country and the tool fills in low, expected and high CPM and CPC from third-party 2026 benchmark reports. Or switch to your own numbers and type the CPM and either the CPC or the click-through rate (CTR) from your Ads Manager. Then pick a goal: leads or purchases. Clicks are multiplied by a conversion rate to give results, and the budget divided by results gives your cost per result (CPA). Enter an order value (or the value of a lead) to see revenue and ROAS.
Formula and assumptions
Formula
Impressions = budget ÷ CPM × 1,000Clicks = budget ÷ CPC; CTR = clicks ÷ impressionsReach = impressions ÷ frequencyLeads or purchases = clicks × conversion rate; CPA = budget ÷ resultsROAS = results × value per result ÷ budget
Assumptions
- CPM and CPC come from third-party 2026 benchmark reports (see sources), not from Meta. They are averages for Facebook and Instagram ads together, in USD, and differ between sources. Low and High are the lowest and highest figure found; Expected is the median of the figures found.
- Where only one source covers a country, the label says "rough estimate" and the CPC range is scaled from that source's own CPM range. Treat it as a starting point only.
- Reach is impressions divided by frequency (default 1.5, an Earnlitix assumption: how often the same person sees the ad). Real frequency depends on audience size and budget.
- Conversion rates default to the 2026 averages published by WordStream/LocaliQ (leads, 8.54%) and Triple Whale (purchases, 1.53%). Your own page, offer and price will differ. ROAS needs an order value or lead value, which you enter; it is hidden until you do.
- Low, Expected and High use the low, median and high cost together, so they show a range of outcomes rather than a forecast. Q4 (Black Friday) costs run higher.
Worked example
$1,000 budget, United States, expected costs, selling a $73.36 product
- Expected US costs: CPM $21.74 and CPC $1.21 (the median of the figures found).
- Impressions = 1,000 ÷ 21.74 × 1,000 = about 45,998. Reach at frequency 1.5 = 45,998 ÷ 1.5 = about 30,665 people.
- Clicks = 1,000 ÷ 1.21 = about 826, a CTR of 1.80%.
- Purchases at a 1.53% conversion rate = 826 × 0.0153 = about 12.6. Cost per purchase = 1,000 ÷ 12.6 = about $79.08.
- Revenue = 12.6 × $73.36 = about $927.61, so ROAS = 927.61 ÷ 1,000 = 0.93.
Result: About 30,665 people reached and 12.6 purchases for $1,000. At these average costs and a $73.36 order the ad does not pay for itself on the first sale, so margin, repeat buyers or a better conversion rate decide whether it works.
Limitations
- Benchmarks are averages from third-party reports (Lebesgue, Adamigo, Neal Schaffer, Stackmatix, Superads), not data from Meta. They disagree, because each uses a different advertiser sample, objective and period. Your real costs depend on your audience, creative, season and competition.
- "Click" is defined differently across sources (all clicks versus link clicks), so CPC and CPA from different reports are not like-for-like. Triple Whale, for example, reports a lower median cost per purchase than this example because its brands convert better than the default.
- Where only one source covers a country, it is marked a rough estimate. Costs rise sharply in Q4 around Black Friday, and small budgets exit the learning phase slowly. Reach assumes a frequency of 1.5, which is an Earnlitix assumption.
