Customer acquisition cost (CAC) is the total sales and marketing spend required to win one new customer, calculated by dividing those expenses by new customers acquired in the same period. SaaS companies more often use the CAC ratio, which divides sales and marketing spend by new annual recurring revenue. Benchmarkit’s 2026 report puts the median blended CAC ratio at $1.30 and the median CAC payback at 16 months.
Before any of those numbers helps you, find the asterisk.
Every CAC benchmark you’ll ever read has one, and it’s where the sample size hides. The 2026 Benchmarkit report surveyed 342 companies, and its blended CAC ratio rests on 122 of them. The 2025 edition’s blended CAC ratio rested on 43.
Forty-three. Out of a headline sample of 583.
So an article saying “according to a survey of 583 SaaS companies, blended CAC is $1.40” is describing a 43-company subsample. And hoping you don’t check.
I’m going to give you every number below with its own asterisk attached.
Quick Navigation
- What Is Customer Acquisition Cost?
- How Do You Calculate CAC?
- What Is a Good CAC in 2026?
- How to Reduce CAC (7 Strategies)
- CAC vs Customer Lifetime Value
- Industry CAC Benchmarks
- What Belongs in CAC, and What Does Not
- Which CAC Claims Do Not Hold Up?
- Frequently Asked Questions
What Is Customer Acquisition Cost?
Customer acquisition cost is the total sales and marketing spend required to win one new customer over a defined period. It answers one question: what did that logo cost to land?
The complication is that four different definitions are in common use, and they produce different answers for the same company. Pick one, write it down, and use it consistently.
- Blended CAC. All sales and marketing spend divided by all new revenue, including expansion from existing customers.
- New-logo CAC. The same spend divided by new-customer revenue only. This is the harder number and the more useful one.
- Paid CAC. Only advertising spend divided by customers from paid channels, which ignores salaries entirely.
- CAC ratio. Dollars of sales and marketing spend per dollar of new annual recurring revenue, expressed as a ratio rather than a per-customer figure.
The gap between blended and new-logo is the single biggest swing. Benchmarkit’s 2026 medians put blended at $1.30 and new-logo at $1.63, so including expansion revenue in your denominator makes CAC look roughly 20% better. CAC sits alongside fourteen others in our guide to the sales metrics worth tracking and how to compute each one.
Two more definitional forks change the answer. Whether customer success costs count, and whether the calculation is fully loaded with salaries or only counts program spend.
Benchmarkit includes customer success costs only in its expansion CAC ratio, and excludes them from new and blended. Bessemer includes customer success in CAC payback generally, at least the portion tied to renewal and upsell.
Nobody puts general and administrative or R&D costs in the CAC numerator. Bessemer’s original formula also excluded account management costs from the “farmer” organization.
Key insight
Fully loaded CAC includes salaries and headcount costs, and David Skok’s original definition is fully loaded. He also offers a partially loaded variant without headcount, for web businesses where staffing doesn’t scale with acquisition.
CAC vs CPA: What Is the Difference?
CAC counts what it costs to win a paying customer. CPA, cost per acquisition, usually counts the cost of an action short of that, such as a signup, a trial start or a demo booking.
Mixing the two understates your real cost by whatever your trial-to-paid rate happens to be. A $40 CPA on trial signups converting at 20% is a $200 CAC.
How Do You Calculate CAC?
Divide total sales and marketing spend by the number of new customers acquired in the same period. That’s the base formula, and every variant below is a refinement of it.
| Metric | Formula | Defined by |
|---|---|---|
| CAC per customer | Sales and marketing spend / new customers acquired | David Skok, forEntrepreneurs |
| Blended CAC ratio | S&M expenses / (new ARR + expansion ARR) | Benchmarkit |
| New CAC ratio | S&M expenses / new ARR | Benchmarkit |
| Expansion CAC ratio | (S&M + customer success expenses for expansion) / expansion ARR | Benchmarkit |
| CAC payback, with margin | S&M expenses / (new-customer ARR x gross subscription margin) x 12 | Benchmarkit |
| CAC payback, original | Prior-quarter S&M / (new CMRR added x gross margin %) | Bessemer, 2012 |
| CLTV:CAC | (Average ARR per account x gross margin / churn rate) / CAC per new customer | Benchmarkit |
Bessemer’s worked example is the clearest one published: $770,000 divided by $100,000 in new monthly recurring revenue at 70% gross margin gives 11 months.
That example comes from Bessemer’s 2012 paper “10 Laws of Cloud”, read September 10, 2026, and its figures are quarterly rather than monthly. Bessemer’s current cloud-metrics page drops the example entirely and substitutes a different one.
Leave gross margin out and your payback number gets shorter and wrong. The version without margin is Skok’s 2009 rule of thumb, and it assumes a dollar of revenue is a dollar of contribution.
What Belongs in CAC, and What Does Not
Every page defining CAC lists what to include. Almost none lists what to leave out, and that is where the number usually goes wrong.
| Include in CAC | Leave out of CAC |
|---|---|
| Paid media spend across every channel | General and administrative overhead such as finance, legal and HR |
| Salaries, commission and bonuses for sales and marketing staff | Research and development, including product engineering payroll |
| Sales and marketing software, from the CRM to the sending tools | Account management and customer success, unless you are computing expansion CAC |
| Agency and contractor fees tied to acquisition | Support and onboarding costs incurred after the contract is signed |
| Content production and creative costs | Rent, equipment and other costs untied to winning customers |
The customer success line moves the number most. Bessemer’s 2012 formula excludes account-management cost from CAC, and Benchmarkit’s expansion CAC ratio deliberately includes it, so you can’t compare those two figures.
What Is a Good CAC in 2026?
A blended CAC ratio at or below $1.30 and a CAC payback at or below 16 months put you at the median. Below $1.08 blended and under 10 months payback puts you in the top quartile.
These come from Benchmarkit’s 2026 B2B SaaS and AI-Native Performance Metrics report, which surveyed 342 companies. Each row below carries the sample that actually answered that question.
| Metric | Median | Top quartile | Sample |
|---|---|---|---|
| Blended CAC ratio | $1.30 | $1.08 | 122 of 342 |
| New-logo CAC ratio | $1.63 | $1.14 (25th percentile) | 128 of 342 |
| Expansion CAC ratio | $0.80 | Not published | 74 of 342 |
| CAC payback period | 16 months | 6 months or less | 198 of 342 |
| CLTV:CAC ratio | 4.1x | 7.8x | 146 of 342 |
| Net revenue retention | 102% | 110% (75th percentile) | 230 of 342 |
| Gross revenue retention | 84% | 91% (75th percentile) | 226 of 342 |
Two of those moved sharply year on year. New-logo CAC improved about 19%, from $2.00 in the prior report. Gross revenue retention fell four points from 88%, which Benchmarkit calls the most significant single-year deterioration in its four-year trend.
Most articles you’ll find still quote the 2025 figures. If you see new CAC at $2.00, NRR at 101% and GRR at 88%, that page is a year stale.
A Worked CAC Example, in Round Numbers
Spend $10,000 on sales and marketing in a month and win 100 new customers, and your CAC is $100. That is the version every calculator shows you.
Now run the same month with a stricter denominator. If 40 of those 100 were existing customers upgrading, your new-customer CAC is $10,000 divided by 60, or $167.
The blended figure looked 40% cheaper than the number a board would ask about. Both are correct arithmetic, and only one answers what a new logo costs.
| Version | Calculation on the same month | Result |
|---|---|---|
| Blended CAC | $10,000 of spend divided by 100 customers, new plus expansion | $100 |
| New-customer CAC | $10,000 of spend divided by the 60 net-new customers | $167 |
| Paid CAC | $6,000 of paid spend divided by the 25 customers paid channels sourced | $240 |
Three defensible numbers, one month, one set of books. Say which one you mean before anyone compares it to a benchmark.
How CAC Payback Varies by Deal Size
Deal size moves payback more than anything else you control. Benchmarkit’s 2026 spread runs from 11 months for sub-$5,000 ACV to 22 months for the $50,000 to $100,000 band.
Blended CAC follows the same curve, from $0.80 at sub-$5,000 ACV to $1.71 in the $50,000 to $100,000 band. Vertical SaaS runs more expensive than horizontal at $2.22 against $1.62 on new-logo CAC.
Growth rate matters as much. Companies growing over 50% showed 10-month payback, while those growing 21% to 30% showed 22 months.
What Bessemer Says a Good Payback Period Is
Bessemer publishes targets by segment: under 12 months for SMB-focused companies, under 18 months for mid-market, and under 24 months for enterprise. Its own portfolio average in the $1M to $10M ARR band is 15 months.
Note the tension. Bessemer’s targets are prescriptive, its 15 months is descriptive of its own portfolio, and the market-wide median is 16.
The commonly repeated “12 months is good” rule is Skok’s 2009 guidance rather than Bessemer’s. The market median has never sat at 12.
How Do You Reduce Customer Acquisition Cost? 7 Strategies
Reducing CAC means improving conversion, raising deal size, or shifting mix toward cheaper channels. These seven cover all three.
1. Improve Audience Targeting Before Spending More
Sharper audience targeting lowers CAC faster than any bidding change. A list of 400 companies matching your best customers beats 40,000 names matching a job title.
2. Shift Mix Toward Organic and Referral
Organic and referral sources carry lower acquisition costs than paid. Working with SEO agencies or building the capability in house takes months to pay off, which is exactly why it lowers CAC once it does.
3. Raise Conversion Rates Instead of Traffic
Improving conversion rates reduces CAC without increasing spend at all. Landing page tools like Unbounce and session analytics from Hotjar tell you where the drop-off sits before you rebuild anything.
4. Use Retargeting on Warm Traffic
Retargeting reaches people who already showed intent, which makes it the cheapest paid inventory available to most companies. Cap frequency so you don’t spend the same budget on the same person eleven times.
5. Shorten the Sales Cycle
Every week a deal stays open costs rep time that lands in the CAC numerator. Mapping your sales funnel and tightening sales processes removes the stages that add days without adding information.
6. Automate the Repeatable Parts
Sales automation and in-product messaging through tools like Intercom handle the work that doesn’t need a person. Track your sales efficiency metrics before and after, or you won’t know whether it helped.
7. Fix Response Time
Slow replies waste acquisition spend you’ve already committed. Measuring average response time with EmailAnalytics surfaces the gap, and disciplined cost tracking keeps the fix honest.
Outbound belongs in this list too, with a caveat. Our cold email outreach lowers cost per meeting for companies with a definable account list. It does nothing for a company whose real problem is a 3% close rate.
CAC vs Customer Lifetime Value
Customer Lifetime Value (CLV) is the total gross profit a customer generates before they leave. Compared against CAC, it tells you whether acquisition is a business or a hobby.
The famous 3:1 rule is real, and it’s a rule of thumb from a blog post. David Skok published it on forEntrepreneurs on 22 December 2009, basing it on “closely watching several hundred startups that have failed.” No dataset, no sample size.
His exact wording hedges three separate ways in one sentence. He calls them “a very rough rule of thumb” and “guidelines that you might find helpful,” then writes that “it appears that LTV should be about 3 x CAC.”
He also noted in the same paragraph that public companies like Salesforce were running closer to 5x. The 3x was a floor he flagged as approximate, and it’s been repeated for seventeen years as though it were a finding.
Bessemer does prescribe the same ratio, which is worth knowing because people often claim it only prescribes payback. Bessemer’s published guidance is to invest in customer acquisition when CLTV divided by CAC sits at 3x or above.
The measured market number is higher than the rule. Benchmarkit’s 2026 CLTV:CAC median is 4.1x across 146 companies, with the top quartile at 7.8x and the bottom quartile at 1.1x.
Key insight
A 1.1x CLTV:CAC ratio, which is where the bottom quartile sits, means a customer barely covers what you spent to get them. That’s the quartile to worry about, and a quarter of companies in the sample are in it.
What Is the Average Customer Acquisition Cost by Industry?
No credible market-wide average CAC by industry exists. The B2B tables in circulation trace to one SEO agency’s client book, and the ecommerce figures to a 2021 Shopify dataset.
| Publisher | What the figures are | Sample stated | What it leaves out (read October 7, 2026) |
|---|---|---|---|
| First Page Sage, B2B edition (updated January 26, 2026) | Its own clients, January 2022 to August 2025; 29 industries and 22 SaaS verticals; B2B SaaS at $239 combined | None | Email, events, direct mail and outdoor ads; the combined figure is weighted 75% organic |
| Shopify (page dated July 29, 2024) | 2021 data for ecommerce brands with fewer than four employees; five industries, including $21 for arts and entertainment and $377 for electronics | None | Every brand with four or more employees, and every year since 2021 |
| Benchmarkit 2026 | CAC ratios per $1 of new ARR, from a survey of 342 B2B SaaS companies | Per metric, such as 122 for blended CAC | Dollar CAC per customer, and every industry outside SaaS |
Most “average CAC by industry” posts copy the First Page Sage table and drop its date range and weighting. Check a table’s source page for a sample size before you benchmark against it.
Use ratios instead. A CAC ratio and a payback period compare across companies in a way a dollar figure never does, because the dollar figure moves with contract size.
Which Benchmark Publishers Still Publish CAC
| Publisher | Publishes CAC? | Sample |
|---|---|---|
| Benchmarkit 2026 | Yes, with per-metric samples | 342 companies |
| SaaS Capital 2026 | No. Growth and NRR only | Over 1,000 companies |
| High Alpha (formerly the OpenView report) | Partly. A payback heatmap, no numeric median on the public page | 800+ respondents |
| ChartMogul | No. Its last public report is 2023 and contains no CAC metric | 2,100+ businesses |
| Maxio 2026 | No. Billings-based growth only | 2,000+ companies |
| KeyBanc and Sapphire 2025 | Gated. The public release discloses no sample size | Not disclosed |
The OpenView benchmarks did not stop. High Alpha took over the report, and the 2025 edition is the ninth annual.
Which CAC Claims Do Not Hold Up?
“It costs 5 to 25 times more to acquire a customer than to retain one”
This is folklore, and I traced exactly where the chain breaks. The popularizing source is a Harvard Business Review digital article from October 2014 by Amy Gallo, a contributing editor rather than a researcher.
Her opening sentence is the tell. She writes that “depending on which study you believe,” acquiring a new customer costs five to 25 times more than retaining one. She names no study.
The article’s only hyperlink supports a different claim entirely. It points at a Bain brief by Fred Reichheld for the separate statement that raising retention 5% raises profits 25% to 95%.
I fetched that Bain PDF. It contains no 5x claim and no 5-to-25x claim. Its only relevant line says a 5% increase in customer retention produces more than a 25% increase in profit in financial services, with no sample size and no methodology.
The 1990 Reichheld and Sasser HBR article usually credited alongside it is paywalled, and its public abstract contains no acquisition-cost multiple. Two claims got conflated over twelve years, and the surviving one is unsourceable.
What Peer-Reviewed Research Says About Acquisition Against Retention Spend
Two academic papers speak directly to this, and no page ranking on the first page for “customer acquisition cost” cites either one.
Peter Pfeifer of the Darden School examined the 5X claim in “The optimal ratio of acquisition and retention costs”, Journal of Targeting, 2005. His finding is sharper than saying the number is unsourced.
Pfeifer’s abstract states that if the 5X ratio refers to average costs, then a 5X ratio does not necessarily imply the firm should spend more on retention. At the optimum, marginal acquisition cost should equal marginal retention cost.
So even if the folklore figure were real, the conclusion everyone draws from it would still be wrong. Spend gets optimised at the margin rather than at the average.
The largest empirical measurement of acquisition cost I could find is Min, Zhang, Kim and Srivastava, Journal of Marketing Research, 2016. It models firm-level wireless telecom data across 41 countries from 1999 to 2007.
Its useful finding for a marketer: acquisition cost per customer responds more to market position and competition than retention cost does. Market-share leaders hold a real acquisition cost advantage and no retention cost advantage.
No government series and no standards body publishes a CAC benchmark. I checked BLS, Census, the SEC full-text search and the FTC, and CAC exists as a statistical series in none of them.
“Blended CAC across the industry is X”
Check the asterisk before you repeat any figure like this. Benchmarkit’s 2025 blended CAC ratio rested on 43 companies out of a headline 583, a 7% response rate on that metric.
The 2026 edition is better at 122 out of 342, and two thirds of respondents still don’t report it. Benchmarkit itself warns that year-on-year comparisons could reflect selection bias among participants.
One more inconsistency worth knowing. The 2025 report states its participant count as 583 in the key findings and 563 in every participant-profile chart, and never reconciles the two.
“You can compare your CAC payback to a public company’s”
You can’t, and Benchmarkit says so directly. CAC payback at public companies measures net new implied ARR, which includes churn, downsells and expansion, so it isn’t the same calculation.
Frequently Asked Questions
What is a good customer acquisition cost?
Stated as a percentage, a good CAC is about $1.30 of sales and marketing spend per $1.00 of new annual recurring revenue, or 130%, per Benchmarkit’s 2026 report. Top quartile runs near 108%.
For B2B SaaS in 2026, a blended CAC ratio at or below $1.30 and a CAC payback period at or below 16 months sit at the median, per Benchmarkit’s report covering 342 companies. Top-quartile performance is $1.08 blended and payback of 6 months or less. Deal size drives most of the variation, with sub-$5,000 ACV companies paying back in 11 months and $50,000 to $100,000 ACV companies taking 22.
What is the average customer acquisition cost?
No credible market-wide average exists in dollar terms. The widely circulated industry tables come from a single marketing agency’s client analytics accounts, and that agency’s own limitations section states its data is weighted 75% organic to reflect its own services and excludes several major channels. Use the CAC ratio and payback period instead, since both are comparable across companies with different contract sizes.
How do you calculate CAC?
Divide total sales and marketing spend by the number of new customers acquired in the same period. For SaaS, use the CAC ratio instead, dividing sales and marketing expenses by new annual recurring revenue. Decide up front whether you’re measuring blended CAC, which includes expansion revenue in the denominator, or new-logo CAC, which does not, because the two differ by roughly 25%.
What is the CAC payback period?
CAC payback is the number of months it takes for gross profit from a new customer to cover what you spent acquiring them. Bessemer’s formula divides prior-quarter sales and marketing spend by new committed monthly recurring revenue multiplied by gross margin, and Benchmarkit’s 2026 median is 16 months across 198 companies. Leaving gross margin out produces a shorter and incorrect answer.
Is the 3:1 LTV to CAC rule real?
It originates in a David Skok blog post published 22 December 2009 on forEntrepreneurs, which called it “a very rough rule of thumb” with no dataset or sample size behind it. Skok noted in the same paragraph that public companies were running nearer 5x. Bessemer does independently recommend investing in acquisition at 3x or above, and Benchmarkit’s measured 2026 median is 4.1x across 146 companies.
Does it cost five times more to acquire a customer than to retain one?
There is no primary source for that claim. It traces to an October 2014 Harvard Business Review digital article whose author writes “depending on which study you believe” and cites nothing. The Bain brief she links supports a different statement, that a 5% increase in retention produces more than a 25% increase in profit in financial services, and contains no acquisition-cost multiple at all.
Should customer success costs be included in CAC?
It depends whose definition you use, and you should state which. Benchmarkit includes customer success costs only in its expansion CAC ratio and excludes them from new and blended CAC. Bessemer includes customer success in CAC payback generally, at least the portion tied to renewal, upsell and cross-sell, while excluding account management costs from the farmer organization.
Pick one definition, state it on every report, and check the sample size behind every benchmark you compare yourself to.
Jayson is a long-time columnist for Forbes, Entrepreneur, BusinessInsider, Inc.com, and various other major media publications, where he has authored over 1,000 articles since 2012, covering technology, marketing, and entrepreneurship. He keynoted the 2013 MarketingProfs University, and won the “Entrepreneur Blogger of the Year” award in 2015 from the Oxford Center for Entrepreneurs. In 2010, he founded a marketing agency that appeared on the Inc. 5000 before selling it in January of 2019, and he is now the CEO of EmailAnalytics and OutreachBloom.
