How to Calculate Customer Acquisition Cost Before You Launch

Estimate acquisition cost per channel before launch: take a published cost-per-click benchmark for your category, divide it by your expected visitor-to-trial and trial-to-paid rates, then check the result with a paid test of about 1,000 clicks (roughly $2,000 on Meta). Compare it with a lifetime value you can defend.

What actually counts as acquisition cost?

Customer acquisition cost, at its simplest, is total cost of Sales & Marketing divided by the number of deals closed. A fuller version splits it into two numbers: one that counts only marketing program spend, and one that also includes the people and other expenses associated with running the sales and marketing organization, because a touchless funnel and a sales-assisted one carry very different loaded costs. Before launch you have no acquired customers to divide by, so you are not calculating CAC; you are forecasting it from two inputs: what it costs to buy attention in your category, and what fraction of that attention turns into a paying customer.

Neither input should come from a hunch. The first comes from published ad platform benchmarks. The second comes from a small paid test, because no benchmark tells you how your specific landing page or offer converts.

How do you turn a channel benchmark into a CAC estimate?

Start with cost per click for your category, then divide by two conversion rates in sequence: visitor to lead, and lead to paying customer.

The gap between industries is large enough that guessing is not a rounding error; it changes the answer by 5x. The average cost per click in Google Ads in 2025 is $5.26, but that hides real variation: Attorneys & Legal Services runs $8.58, while Arts & Entertainment sits at $1.60, Restaurants & Food at $2.05, and Travel at $2.12. Business Services, a reasonable proxy for a B2B tool, runs $5.58 on the same table.

Social platforms are a different order of magnitude for the click itself. The average cost per click in Facebook ads for traffic campaigns across all industries is $0.70, while the average cost per click in Facebook ads for leads campaigns across all industries is $1.92, higher because the platform is optimizing for a form fill rather than a click. A cheap click that never converts is not a cheap customer, which is exactly why the next step matters more than the benchmark itself.

Channel	Typical CPC	Source
Google Search Ads, Business Services	$5.58	WordStream, 2025
Google Search Ads, all industries	$5.26	WordStream, 2025
Meta traffic campaign, all industries	$0.70	WordStream, 2025
Meta leads campaign, all industries	$1.92	WordStream, 2025
Referral, organic, content	No reliable public benchmark	Measure from your own waitlist or fake-door test

That last row is not a gap to paper over with a guess. If you cannot source a number for a channel, say so and go measure it yourself instead of inventing one.

Why do you still need a small paid test?

A benchmark gives you a category average. It cannot tell you your landing page's conversion rate, your offer's clarity, or whether your targeting is actually reaching people who need the thing. Two founders paying the same $5.58 click can end up with wildly different costs per customer depending on what happens after the click.

The test does not need to be large, but it needs enough volume that a conversion rate is not just one lucky or unlucky click. This is my own working rule, not a statistical standard: treat anything under thirty conversions as too noisy to trust to the decimal point, useful for direction, not for a final go or no-go call. That rule applies to the visitor-to-trial step, the one a short paid test can actually measure within days. At a 3 percent landing page conversion rate, thirty trial signups means roughly a thousand clicks. At $5.58 a click on Google that is about $5,580, a lot to spend just to sanity-check an assumption. At $1.92 a click on a Meta leads campaign, the same thousand clicks costs about $1,920, about a third of the cost for the same sample size. That gap alone is a reason to run your first test on the cheaper channel even if you expect to scale on the more expensive one later. The trial-to-paid step needs a separate, slower measurement. At a hypothetical 25 percent conversion rate, reaching the same thirty converted customers takes about 120 trial signups. Each signup still has to run through a full trial period before it counts, so this number will not be ready on the same timeline as your ad results.

What does the full calculation look like with real numbers?

Say you are testing a hypothetical $29-a-month scheduling tool for personal trainers, sold with a free trial. You assume a 75 percent gross margin and a 5 percent monthly churn rate once customers are paying. Both are your own hypothetical inputs, not benchmarked figures, so treat them as a starting hypothesis to be corrected once you have real cohorts. You plan to test two channels, and you are making one simplifying assumption to compare them on the same table: Meta's $1.92 leads-campaign benchmark is typically measured against people who complete an on-platform instant form, not visitors who click through to an external landing page, so applying the same 3 percent landing-page conversion rate to both channels below may flatter Meta's real cost per trial.

Step	Google Search Ads (hypothetical)	Meta Leads Campaign (hypothetical)
Cost per click	$5.58	$1.92
Visitor-to-trial conversion (assumed)	3%	3%
Cost per trial signup	$186.00	$64.00
Trial-to-paid conversion (assumed)	25%	25%
Estimated CAC	$744.00	$256.00
Monthly gross profit ($29 × 75%)	$21.75	$21.75
Payback period (CAC ÷ monthly gross profit)	~34 months	~11.8 months
Estimated LTV ($29 × 75% ÷ 5% churn)	$435.00	$435.00
LTV : CAC	0.58 : 1	1.70 : 1

In this hypothetical, neither channel clears the 3:1 ratio, but the gap between them is the whole point of doing the arithmetic before launch. On Google, this hypothetical business would lose money on every customer at these assumptions, and it would take about 34 months to recoup the acquisition cost. On Meta the ratio is still short of 3:1, but the payback period, about 11.8 months, clears Skok's 12-month bar. Meta passes the payback test even though it fails the 3:1 ratio, which is exactly why payback deserves as much weight as the ratio itself. The fix is not to round the churn assumption more optimistically until the math works. It is to either raise the trial-to-paid conversion with a better onboarding flow, raise the price, or find a cheaper channel, before committing a real budget.

What LTV:CAC ratio do investors actually expect?

The number everyone reaches for is 3:1. Almost everyone in SaaS has heard that we should target an LTV to CAC ratio of at least three, popularized by the content published by David Skok on SaaS metrics. Skok introduced the ratio as his own rule of thumb in a February 2010 post on his site, forEntrepreneurs, writing plainly that "my rule of thumb is that LTV must be at least 3x greater than CAC." The logic holds up in principle: a customer needs to return meaningfully more than their acquisition cost, because that gap is what funds product development, support and the next round of acquisition spend, not just the marketing line item.

When does the 3:1 rule not apply?

It does not apply cleanly to you if you are pre-launch, and it may not even apply once you have some customers. Skok's own site flags the limitation directly: while in theory a well calculated LTV:CAC ratio provides an answer, startups must also watch a related metric, months to recover CAC, because if it takes too long to recover marketing spend, a startup may require unworkable amounts of capital.

The venture writer Tomasz Tunguz makes the sharper version of the argument: a company one, two or even three years into sales can't yet accurately forecast customer lifetimes. Put bluntly, LTV built on a guessed lifetime is a guess wearing a ratio's clothing. His alternative is to lean on payback period instead, because within 14 to 18 months most startups will have collected concrete data about the payback periods for customer acquisition, instead of 3, 5 or 7 years for LTV observations. Before you have a single cohort, you have neither number honestly, so use the ratio as a filter for whether an idea is worth testing further, not as a scorecard to report.

Payback targets also vary by who you sell to. Skok's own rule of thumb is that startups need to recover their cost of customer acquisition in less than 12 months, a bar set for capital-scarce startups rather than for the banks and wireless carriers that can tolerate payback periods of a year or more. A self-serve product selling to small businesses is closer to that startup case: there is no large contract behind it to absorb a slow payback. An enterprise sale with a multi-year contract can tolerate a longer payback because the eventual contract value is so much larger. Comparing revenue LTV directly against CAC also overstates the ratio, because revenue includes the cost of goods sold that never reaches the bottom line; multiplying by gross margin first, as the worked example above does, corrects for it.

How do you decide which channel to build around?

1. Pull the published CPC or CPL for your category from a platform benchmark report, not a generic blended CAC figure that mixes every industry together.
2. Write down your visitor-to-lead and lead-to-paid conversion assumptions separately, and label them as assumptions in whatever document you show a co-founder or investor.
3. Multiply through to get an estimated CAC per channel, the way the table above does.
4. Run a paid test on the cheapest plausible channel first. Aim for at least thirty trial signups, roughly a thousand clicks at a 3 percent landing-page rate, before you trust the visitor-to-trial percentage. The trial-to-paid percentage needs a longer, separate test. At a 25 percent assumed rate you need about 120 signups to reach the same thirty converted customers, and you have to wait out a full trial period for each one.
5. Recompute CAC with your real numbers, then compare it against a defensible LTV, price times gross margin divided by expected churn, rather than against 3:1 in the abstract, until you have enough cohort data to trust either side of the ratio.
6. If the ratio or payback period fails, change the funnel or the channel before you change the story you tell yourself about the numbers.

The exercise is not about hitting 3:1 on a spreadsheet before you have sold anything. It is about knowing, in dollars, what a customer will cost you to find, so that when you do launch, you are choosing a channel on evidence instead of on which platform's ad manager was easiest to set up.