LTV vs CPA: How Much Can You Afford to Pay for a Customer?

An ad platform's cost per acquisition tells you what you paid for a counted conversion. To judge that cost, you need to know how much a new customer contributes to your business over the whole relationship. Take a hypothetical $400 acquisition cost. It can lose money on a customer who buys once at a thin margin and pay for itself many times over on one who stays for five years.

This guide compares the two numbers. It explains how to calculate lifetime value (LTV) and cost per acquisition (CPA), where the familiar 3:1 ratio comes from and the conditions attached to it, and how to turn the ratio into a CPA target you can give a media team or an agency. It also covers why the numbers drift once you start spending more.

DisclosureinBeat, which is mentioned in this article, is part of Fieldtrip, a group of marketing agencies in which Jake Jorgovan is a partner.

What LTV and CPA each measure

Lifetime value is what a customer is worth to you over the time they buy from you. The useful version counts profit, not sales. Andreessen Horowitz's startup metrics guide (opens in a new tab) defines it as "the present value of the future net profit from the customer over the duration of the relationship", and names using revenue or gross margin instead as a common mistake.

Cost per acquisition is what you pay for each conversion. In an ad platform, a conversion is whatever action you choose to count: a purchase, a sign-up, an install or a lead. Google's Target CPA documentation (opens in a new tab) describes the setting as "your desired average cost per conversion".

Customer acquisition cost (CAC) is the full cost of winning a new paying customer, across media, sales and everything else it takes. The a16z guide says it should include costs such as referral fees, credits and discounts.

The distinction matters because the ratio people quote is LTV to CAC. A platform CPA only belongs in that ratio when the conversion is a new paying customer and the media spend is the only acquisition cost. If the conversion is a lead, or a sales team closes the deal, the platform CPA understates what the customer cost.

Image

Zoom

Original comparison based on a16z’s startup metrics guide and Google Ads’ CPA documentation. A platform conversion may be a lead rather than a paying customer.a16z.com (opens in a new tab) · a16z.com (opens in a new tab) · support.google.com (opens in a new tab)

How to calculate lifetime value

The simple, undiscounted formula multiplies what a customer contributes each period by how long they stay:

LTV = average contribution per customer per period × expected lifetime in periods

Contribution is revenue minus the variable costs of serving that customer; the contribution margin is that amount as a share of revenue. For a repeat-purchase business, the same idea reads as contribution per order times the orders you expect over the relationship. For a subscription business with a steady churn rate, the expected lifetime is one divided by that rate, measured in the same period as the revenue. Venture investor David Skok gives the arithmetic in his SaaS metrics definitions (opens in a new tab): a monthly churn rate of 3% means an average lifetime of 33 months, and an annual churn rate of 20% means five years. The simple version also assumes customers pay roughly the same amount and do not expand. Where purchases are irregular or customers buy more over time, estimate contribution from what each group of customers has actually bought.

A hypothetical example: a subscription business charges $200 a month and keeps $140 of it after the variable costs of serving each customer, a 70% contribution margin. It loses 2.5% of its customers each month; assume that rate holds with no expansion. The average lifetime is 40 months. Lifetime revenue is $8,000 and lifetime contribution is $5,600. The second number is the one to compare with acquisition cost.

Three refinements make the estimate more honest:

• Count the margin that varies with the customer. The a16z guide prefers contribution margin, because an LTV counted in revenue or gross margin overstates how much you can spend to acquire a customer.
• Use observed contribution while history is short. With only a few months of data, the same guide measures value to date instead of predicting a lifespan, and compares 12-month and 24-month values once customers reach those ages.
• Calculate it by segment. A blended LTV hides the customers who leave in month three and the ones who stay for years. In Skok's guide, HubSpot's Brad Coffey describes how its segment analysis (opens in a new tab) found an LTV to CAC ratio of 1.5 when selling directly to very small businesses and 5 when selling through resellers. Within twelve months HubSpot had moved from 12 direct reps and 4 channel reps to 2 direct and 25 through the channel.

How to calculate CPA and CAC

The platform calculates CPA for you: spend divided by conversions. CAC takes more work:

CAC = acquisition spend in a period ÷ new customers acquired from that spend

Acquisition spend includes media, agency and creative fees, sales salaries and commissions where sales closes the deal, the tools used for acquisition, and incentives such as first-order discounts or referral fees. Match the spend with the customers it helped win, allowing for the time your sales cycle takes, and count an incentive once: either in CAC or as lower contribution, not both.

The denominator needs the same care. The a16z guide separates blended CAC, which divides spend by all new customers, from paid CAC, which counts only customers acquired through paid marketing. Blended CAC is a legitimate figure, but it makes paid campaigns look cheaper than they are, because customers who would have found you anyway lower it; the guide says investors weigh paid CAC more heavily. Bill Gurley makes the same point in his 2012 essay on the LTV formula (opens in a new tab): organic customers "would have arrived regardless of spend."

Where the 3:1 ratio comes from

The 3:1 rule of thumb traces to Skok's SaaS metrics guidance. He introduced two guidelines in the first version of his guide and restated them in SaaS Metrics 2.0, published in 2013: LTV should be more than three times CAC, and a business should recover CAC in under 12 months. After testing them with many SaaS businesses, he reported that the best exceed 3, "sometimes as high as 7 or 8", and stressed that these are "only guidelines".

The conditions are on his definitions page. The guideline assumes LTV is calculated on revenue, with no gross margin adjustment, for a business with "a Gross Margin of 80% or higher." That condition explains a revenue-based benchmark for software; it does not make revenue a measure of what you can spend on acquisition.

Return to the hypothetical. At 3:1 on revenue ($8,000), the business could pay up to about $2,667 per customer. At 3:1 on contribution ($5,600), the limit is about $1,867. Applying 3:1 to contribution is stricter than applying it to revenue. At exactly 80% gross margin, Skok's revenue ratio of 3 corresponds to 2.4 on gross profit; a contribution ratio depends on which further variable costs you include. The extra room covers overhead and errors in the estimate. Treat 3:1 on contribution as a planning choice, and set your own target from your overhead, the uncertainty in your numbers and the cash you have.

When I described Lead Cookie's ideal buyer in 2018, it was a B2B company, typically in services or software, with a customer LTV of $10,000 or more. Customer value was one requirement; a well-defined value proposition and niche also mattered to whether the service could deliver results.

How to read the ratio once it is calculated on contribution:

• Below 1: acquisition costs more than the customer is expected to contribute over the whole relationship, before any overhead.
• Between 1 and 3: the customer is expected to repay their acquisition cost. Check whether what remains covers overhead and the uncertainty in the estimate before adding budget.
• Well above 3: you may be spending too little. Skok makes the point about payback: a business acquiring customers for less than its limit can afford to spend more on marketing or sales.

When the ratio is thin, Skok's guide puts churn first: until customers stay, more acquisition is "filling a leaky bucket". After that it lists investing in the lead sources that are profitable and stopping the poor ones until they can be fixed.

A ratio below 3 can still be a sound decision. Skok notes that "many healthy SaaS businesses don't meet the guidelines in the early days." A company may accept a thinner ratio to enter a market, or when it expects customers to buy more over time and has evidence that they do.

Payback: how long a customer takes to repay acquisition

The ratio tells you whether a customer is worth acquiring. Payback tells you how long you wait for the money, which decides how much cash growth will consume.

Skok's second guideline set payback at under 12 months. He wrote that guideline in 2011, when capital was hard to raise. His definitions page explains that access to more capital can support a payback longer than 12 months, and that in enterprise businesses that grow accounts after the first sale, payback "can be around 20 months, and the model works fine." He also shows how to work backward from a limit: a customer paying $500 a month supports up to $6,000 of acquisition cost at a 12-month payback, before any margin adjustment.

In the hypothetical, a customer contributes $140 a month. Twelve times $140 is $1,680, but that shortcut ignores customers who leave. If the first payment arrives at acquisition and 2.5% of customers leave before each later one, the expected contribution over twelve monthly payments is about $1,467 per customer acquired. That is below the $1,867 the ratio allows, so if the business sets a 12-month limit, payback sets the lower ceiling. Check the estimate against what your cohorts actually paid.

Early measures help while lifetime value is still an estimate. OTTO, the German retailer, set 7-day and 30-day return on ad spend targets for its app campaigns from August 2020 and used its own attribution model to track the estimated LTV of the new app users, according to Liftoff's case study (opens in a new tab) of the work. Return on ad spend compares early revenue with media spend, so it gives quick feedback but does not show that the full acquisition cost has been recovered. Reconcile it with contribution and total acquisition cost as the cohort ages.

Turning the ratio into a CPA you can bid

A ceiling on CAC is not yet a number a media team can use. Work back from it:

1. 1

   Choose the stricter of your two limits

   Calculate the maximum CAC from contribution LTV and your target ratio. For the payback limit, use the contribution you expect over the chosen period, allowing for churn and when payments arrive. Use the lower ceiling.

2. 2

   Subtract the costs that are not media

   Take out sales cost, agency and creative fees, acquisition tools and any other costs that are not media, each allocated per new customer. Include a first-order incentive here only if it has not already reduced contribution. What remains is the media budget per new customer.

3. 3

   Convert to the event the platform counts

   If the platform optimizes for leads, multiply by the share of counted leads that become paying customers. In the hypothetical, about $1,467 minus $480 of sales cost leaves about $987 of media per customer. If the platform counts qualified leads and one in five becomes a paying customer, that supports a target of about $197 per lead.

4. 4

   Give the platform the target and watch the average

   Google says some conversions will cost more than the target and some less, and that setting a target too low "may cause you to forgo clicks that could result in conversions." It recommends judging performance over the last 30 days with at least 30 conversions.

5. 5

   Review by channel and segment

   Recalculate when churn, prices or costs change, and compare channels and customer segments separately, as HubSpot did.

Image

Zoom

The article’s hypothetical: a $1,467 CAC ceiling leaves $987 of media spend per customer after sales cost, or about $197 per qualified lead at a 20% close rate.

Platforms can also use customer value directly. For purchase campaigns that bid on conversion value, Google's customer lifecycle goals (opens in a new tab) can bid higher for new customers than for existing ones, and in Search and Performance Max higher again for new customers who resemble the most valuable existing ones. Those settings need a purchase conversion goal and value-based bidding; with a CPA target, the acquisition option bids only for new customers. Either way, the platform knows what a customer is worth only if the business tells it.

Why the numbers drift once you scale

The ratio is calculated on the customers you have acquired so far. Gurley's essay argues that the next ones usually cost more and are often worth less. Spending more buys more of a limited audience, so acquisition cost rises. Raising prices to lift LTV tends to raise churn. And customers bought through marketing usually underperform those who arrived on their own: "Purchased Customers Underperform Organic on Almost Every Metric." His conclusion is to use LTV as a tool for comparing marketing programs and channels, and to have finance monitor it, since the team asking for the budget has a stake in the answer.

Costs do not always rise with spend. TubeScience's Fabletics case (opens in a new tab) describes a brand that distributes budgets using LTV data and kept refreshing its video ads. On Facebook in the first quarter of 2020, TubeScience reports, Fabletics raised its marketing budget 47% while subscribers rose 67% and CAC fell 11%.

Channels also wear out. In 2019 I wrote about the churn built into Lead Cookie, the LinkedIn prospecting service I ran then. Clients rotated their budgets as channels became saturated, and "even ads reach a point of saturation or viewers begin to experience ad fatigue." My conclusion at the time was that we had to extend customer lifetime value with more than one service. That mechanism belonged to a business selling a channel-specific marketing service. In other businesses, the lesson is narrower: revise LTV when retention changes, instead of carrying last year's churn rate forward.

Discounts hide a related cost: they can lower CPA while lowering the margin each customer brings. inBeat, the creator and performance creative agency in the Fieldtrip group, describes its work with Hurom (opens in a new tab), a premium juicer brand whose CPA was rising while its sales depended on discounts. The new ads gave buyers a reason to pay full price, built on health benefits instead of sales. inBeat reports that return on ad spend rose 300%, cost per acquisition fell 65% and month-over-month profit margin grew 33%. Judge CPA and margin together.

Questions about LTV and CPA

What is a good LTV to CPA ratio?

The common guideline is LTV more than three times CAC, from Skok's SaaS guidance. It assumes revenue-based LTV and a gross margin of at least 80%. Calculate LTV on contribution, choose your own target ratio, check payback beside it, and compare the ratio across your own channels and segments.

Is CPA the same as CAC?

Only when the conversion counted is a new paying customer and media is the only acquisition cost. Otherwise CPA is one input to CAC, along with sales cost, fees, incentives and your conversion rate from lead to customer.

What is a good CPA?

A good CPA is one below the limit your own LTV and payback set. Industry averages describe other companies' prices, margins and retention, so they cannot tell you what you can afford.

How do you estimate LTV without much history?

Measure the contribution each group of customers has brought to date, and compare groups at the same age. Report 12-month and 24-month values once customers reach those ages. Until then, keep acquisition targets within the contribution and payback your history supports.

What are the risks of cutting CPA without looking at LTV?

You can buy cheaper customers who leave sooner or spend less, and discounts that lower CPA also lower margin. Track retention and margin by the channel and offer that acquired each customer.