What the ratio compares
The ratio puts what a customer is worth over their whole relationship with you against what it cost to get them. Both halves are simple to state and only one of them is simple to calculate.
Customer acquisition cost is the easier half. Take sales and marketing spend for a period and divide it by the new customers that spend produced.
Lifetime value is where the work sits. One published formula puts it at average revenue per account, times gross margin, divided by the churn rate.
That division is the part to notice. Lifetime itself is one divided by churn, so a business losing 10 percent of customers a year has a ten-year lifetime and one losing 50 percent has two.
Everything downstream inherits that. A wrong churn rate does not make your ratio slightly wrong; it makes it wrong by whatever multiple the churn is out by.
Where the 3:1 target started
The same 3:1 target comes back page after page, and the pages that repeat it give it without saying where it began. It began in a specific place, and that place says something those pages do not.
The target comes from a SaaS metrics blog post published around 2010. It was written as a guideline for a healthy business.
Years later the same site published a piece on when startups should wait before calculating the ratio at all. Skok introduced it himself.
For Entrepreneurs, read 7 September 2026: the author of the target, correcting it
- 1He names himself as the one who introduced the goal of greater than three.
- 2He calls not telling readers when to compute LTV and CAC a significant mistake.
- 3The specific failure he names is readers computing the ratio before they had a repeatable, scalable sales process, when the acquisition cost was still going to change.
The same post says the metric should be used and not believed. It also says that when a founder is closing the deals personally, those deals and that salary do not belong in the calculation.
Who repeats it and who sources it
The same number in page after page, and what each one attaches to it.
| Source | What it gives | Where it says the number came from |
|---|---|---|
| For Entrepreneurs, the origin | Greater than 3 as a guideline, with conditions | Its own author, correcting himself |
| Harvard Business School Online, read | Three or higher, as a rule of thumb | A Harvard Business School professor |
| Wall Street Prep, read | About 3.0x, and a warning that above 5.0x is also a signal | Commonly cited, with nobody named |
| Klipfolio, read | Calls 3:1 the widely cited benchmark | Nobody |
| Cube Software, read | Calls 3:1 the ideal | Nobody |
| Geckoboard, read | 3:1 or better, 4:1 or higher as great | Nobody |
| Chargebee, read | Above 3 as healthy | Nobody |
Read the top two rows against each other. The origin frames a guideline with conditions attached, and the strongest wording here calls it an industry standard with none.
The bottom four rows name nobody. One calls 3:1 the widely cited benchmark without saying who cites it, and the other three state it flat.
Four sources, and three also warn that a high ratio is a problem. They put the line in different places. One says above 3:1 you are probably missing out; a second waits until 5:1; a third never raises it.
Why one target cannot cover the category
The reason a single number struggles here is not philosophical. It sits in the formula, and it is the denominator of the half that everybody finds harder to compute.
Lifetime is one divided by churn. Published annual churn runs from 10 percent or less for enterprise software to over 50 percent for freemium and usage-based products.
Hold revenue and margin still and vary only that, and the arithmetic does the rest.
The same acquisition cost meets two very different numerators. At $1,000 to acquire, one of these businesses is at 9:1 and the other at 1.8:1.
A billing plan list, where the exit each customer is allowed is setWhy one lifetime cannot cover a price list
The plan decides when a customer is able to leave, before any behaviour does. A monthly seat can go at the end of the month and a two year commitment cannot go at all, so a single lifetime figure is an average taken across rows that do not share a clock.
Neither of them is failing the guideline for a reason the guideline was written about. They are different businesses being held to one line.
What the ratio gets confused with
Three things get treated as this number or as substitutes for it, and each answers a question the ratio does not. Two of them look close enough to swap in by mistake.
Four numbers, four questions.
| Number | What it answers |
|---|---|
| LTV to CAC | Whether a customer returns more than they cost |
| CAC payback period | How long before that customer has repaid the cost, which is a cash question |
| Churn rate | What share of the base leaves, which sets the lifetime inside LTV |
| Gross margin | What share of revenue survives delivery, which sets the value inside LTV |
A fifth confusion is the direction. CAC to LTV is the same relationship inverted, so a 3:1 becomes 0.33 and a reader glancing at it reads a disaster.
State which way round yours is, and state the discount rate if you used one. Rates anywhere between 8 and 20 percent are in print, and the choice is subjective.
Where the term goes next
The ratio is a summary of two other numbers, so improving it means improving one of them. Which one you can move depends on facts that sit outside this page.
The denominator is acquisition cost, and that is a channel question before it is an efficiency question.
The numerator is lifetime value, and churn sits underneath it as the divisor that sets the lifetime.
Churn is the input with the fivefold spread, so find your own rate before you accept anybody else's lifetime.
Then check the condition the origin attached. If your growth process is not yet repeatable and scalable, the acquisition cost you divide by is a number about your current experiments, and the ratio will move as those settle.
Calculate it by cohort where you can. The single average hides exactly the difference the arithmetic above makes visible, which is also the trouble with pipeline velocity and every other composite.
One more check before you carry the ratio anywhere. Suppose your lifetime value came from a publisher who subtracts acquisition cost first: dividing it by your acquisition cost counts that cost twice. Which SaaS metrics contain which has a page of its own.
Suppose the acquisition half is the one you want moved: that is our B2B PPC service.
Sources
- For Entrepreneurs Why early-stage startups should wait to calculate LTV:CAC, by Jared Sleeper, with an introduction by David Skok
- Wall Street Prep LTV/CAC ratio, with the lifetime and LTV formulas and a cohort caveat
- Harvard Business School Online LTV/CAC ratio: what it is and how to calculate it, by Darwin Janes, 6 March 2025
- Klipfolio LTV:CAC ratio: what it is and how to calculate it
- Cube Software LTV/CAC ratio: your secret weapon to measure sales and marketing ROI
- Geckoboard LTV:CAC ratio, KPI examples
- Chargebee LTV CAC ratio: how to define, optimize and calculate
- Google The AI Overview for this term, giving the formula and the bands
Questions people ask
What is a good LTV to CAC ratio?
The number everyone quotes is 3:1, and it began as a guideline for a healthy SaaS business, not as a standard. Its author later wrote that he had made a significant mistake by not telling readers when computing the ratio makes sense.
Before adopting the target, check that you have a repeatable and scalable growth process. Without one, the acquisition cost in the denominator is going to change as you build it.
How do I calculate LTV CAC?
Lifetime value divided by acquisition cost. Acquisition cost is sales and marketing spend for a period divided by the new customers it produced.
Lifetime value is the harder half. One published formula puts it at average revenue per account times gross margin, divided by the churn rate, because lifetime itself is one divided by churn.
Why is 3x LTV CAC good?
Neither publisher gives a reason, and the origin publication frames the number as a guideline for a healthy SaaS business without deriving it.
So the target is a judgement somebody made and other people adopted. Its author later wrote that the missing part was never the number, but the conditions under which computing it makes sense at all.
Does the ratio work for every business?
Churn moves it more than anything you do. Lifetime is one divided by churn, so churn is the divisor inside lifetime value. Published churn runs from 10 percent a year for enterprise software to over 50 percent for freemium products, so your row decides your lifetime.
On identical revenue and margin, that is a fivefold difference in lifetime value. A single target across all of it is comparing businesses whose numerators differ by five times.