A good CAC payback period depends on who you sell to. Bessemer Venture Partners publishes the standard targets: "For cloud companies selling into SMB-focused accounts, you should target CAC payback <12 months; for mid-market-focused accounts, target CAC payback <18 months; and for enterprise-focused accounts, target <24 months" (Scaling to $100 Million). Measured reality runs longer. Across 198 B2B SaaS companies reporting 2025 data, "The median CAC Payback Period is 16 months" (Benchmarkit 2026, published 1 June 2026).
The formula is CAC divided by the monthly gross profit one customer generates. Gross profit, not revenue. Run your numbers in the CAC, LTV and payback calculator before comparing yourself to anyone, because a payback computed on revenue is a different metric from the benchmarks above and flatters you by roughly your cost of goods.
The formula, and the version most people get wrong
The clean statement is in Benchmarkit's glossary, 2025 B2B SaaS Performance Metrics Benchmarks: "CAC Payback Period: Sales and Marketing Expenses / (ARR from New Customers x Gross Subscription Margin) x 12".
Without an ARR schedule, the same thing in plain terms:
CAC payback in months = CAC ÷ (monthly revenue per customer × gross margin)
Worked example. You spend $10,000 and win 200 customers, so CAC is $50. Each pays $30 a month at a 70 percent gross margin, throwing off $21 of gross profit. Payback is $50 ÷ $21, or 2.4 months. Change the margin to 40 percent and the same customer pays back in 4.2 months on identical revenue. The margin does the work.
Two things the formula excludes. Bessemer's definition includes more cost than most people put in: CAC payback "usually includes sales, marketing, and customer success expenses (at least the portion that ties to renewal/upsell/cross-sell)". And it ignores churn and the time value of money. A 20 month payback on a customer who leaves at month 14 never pays back at all.
Gross margin payback versus revenue payback
This is the biggest source of fake benchmark comparisons, and the split runs cleanly along source type.
Every primary and investor source is gross margin adjusted. Benchmarkit measures the months to pay back sales and marketing expenses for new customers "on a Gross Margin adjusted basis". Bessemer: "We also measure CAC payback against gross margin-adjusted ARR given that the variable costs associated with selling a cloud software product do not accrete to profit."
Vendor glossaries frequently are not. Chargebee's definition is representative: "Calculating the CAC payback period is as simple as taking the customer acquisition cost (CAC) and dividing it by the monthly recurring revenue ( MRR )."
Both get called "CAC payback period". They are different numbers. At an 80 percent software gross margin the gap is 25 percent. At a 40 percent margin the revenue version understates your payback by more than half.
The rule: compute yours on gross margin, and check how any benchmark was computed before you feel good or bad about it. The payback calculator asks for your margin for exactly this reason.
2026 benchmarks by segment
Two kinds of number sit in this table, and mixing them up is how people chase the wrong target. The three segment rows are investor targets. The rest are measured medians.
| Segment | Benchmark | Source |
|---|---|---|
| SMB SaaS | Under 12 months (target) | Bessemer |
| Mid-market SaaS | Under 18 months (target) | Bessemer |
| Enterprise SaaS | Under 24 months (target) | Bessemer |
| All B2B SaaS | 16 months median, 10 at the 25th percentile, 24 at the 75th (measured, N=198, CY-2025) | Benchmarkit 2026 |
| SaaS at $1M to $10M ARR | 15 months average (measured, cloud portfolio) | Bessemer |
| Ecommerce / DTC | 1.7 orders, or 18 days (measured, $10.1B revenue) | AMP, 2025 DTC Mega Report |
Rows deliberately left out: ecommerce payback in months, ecommerce by AOV band or category, and SaaS payback by go-to-market motion. As of 3 September 2026 no primary dataset publishes any of them, and the sites that appear to are citing each other rather than measuring.
The spread matters more than the median. Benchmarkit: "The top quartile achieves payback in 6 months or less", while "A 4th quartile payback of 48 months is economically precarious in today's capital environment."
The trend is improving: "Median CAC Payback improved from 18 months in CY-24 to 16 months in CY-25, an 11% improvement year-over-year", with the 25th percentile dropping from 12 months to 10.
Where the measured data disagrees with the targets
Benchmarkit declines to cut this metric by SMB, mid-market and enterprise at all, and says why: CAC payback "should be evaluated in context of the company attribute most correlated to the metric's performance, which is Annual Contract Value (ACV) for this metric". Cut that way, two findings land awkwardly against the targets.
Small deals do not pay back fast. On the lowest ACV band: "Lower-ACV companies (<$5K) achieve an 11-month median payback, which is a little long for smaller ACVs." That sits just inside Bessemer's SMB target of 12, so half of low-ACV companies miss the target supposedly easiest to hit.
Enterprise runs long, but not uniformly. "Enterprise ACVs ($50K-$100K) carry the longest median at 22 months, consistent with longer cycles and higher field sales costs. The 25th percentile of 15 months, however, shows that efficient enterprise acquisition is achievable."
Two more cuts from the same sample. By growth rate: "The fastest growing companies have a median CAC Payback of 10 months, compared to 18 months for companies growing 11-20%", with slower growers at 14 months, which Benchmarkit calls "surprisingly competitive". Fast growth and efficient acquisition travel together rather than trading off. By solution type: "Vertical SaaS carries an 18-month median versus a 14-month CAC Payback Period for horizontal B2B SaaS."
So the honest version of "what segment am I in" is: what is my contract value, how fast am I growing, is my market horizontal or vertical.
Why "under 12 months" is the lazy answer
The 12 month rule is repeated everywhere, and the people publishing the data warn against it. Benchmarkit's 2025 report: "Common wisdom often says ~12 months CAC Payback Period is good - but this metric is highly correlated to ACV as you will see on the next chart."
Three reasons the single number fails.
It ignores contract terms. Twelve months matters because annual prepayment collects a year of revenue on day one. Bill annually and a payback under 12 months makes you cash positive immediately. Bill monthly and 12 months is just a number, with 11.5 months of cash still out the door.
It ignores retention. A 10 month payback with 40 percent annual churn is worse than an 18 month payback at 90 percent gross retention, because in the first a meaningful share of customers leave before month 10.
It ignores what funds it. Bessemer's targets exist because investors ask how long capital is tied up. A bootstrapped business with a strong margin can accept a longer payback than a venture backed one burning to a milestone.
Pick your number from your own constraints: cash on hand, billing terms, retention. Then check it against the segment target and the measured median. The payback calculator reports the LTV to CAC ratio alongside payback, the other half of the picture. Bessemer's rule: invest in acquisition "when CLTV / CACs are 3x+".