Skip to content
Faisal Mahmud TusherDigital marketing, SEO & analytics
with Faisal Tusher

Growth Marketing › Growth Economics

LTV and CAC

How lifetime value and customer acquisition cost are defined and calculated, the ratio and payback that matter, and blended versus channel CAC.
Guideintermediate11 min readUpdated

Why these two numbers

Most marketing measures describe a channel or a campaign. Lifetime value (LTV) and customer acquisition cost (CAC) describe a customer, and that is why a growth team organises its economics around them. A channel can look excellent on cost per order and still be destroying value if the customers it brings never order again; a channel can look expensive on cost per order and be the best investment the business has if its customers stay for years. Only a customer-level view can tell the difference.

The two numbers also connect marketing to the rest of the company. Finance recognises margin and cash; product recognises retention; sales recognises expansion. Lifetime value is where all three meet, and acquisition cost is the marketing line in the same equation. When the growth economics hub describes the unit economics table every growth team keeps, these are its two anchor columns, and the KPI tree framework places them as the economic layer under the north star.

The catch is that both are defined by choices, and the choices are often made to produce a comfortable answer. The rest of this guide is about making them deliberately.

Defining LTV

A lifetime value figure is the product of several decisions, each of which changes the number. Make them explicitly and write them down next to the figure.

ChoiceOptionsRecommendation
Value basisRevenue, gross margin, contribution marginContribution margin: revenue less cost of goods, fulfilment, payment, variable service and any ongoing incentive costs
HorizonObserved lifetime, fixed window (a year, two years), projected to churnA fixed window for decisions, with the projected figure shown alongside and labelled as a projection
Which customersAll acquired, paying only, by cohort, by channel or segmentPaying customers by cohort and channel; the average across everyone hides the variation that matters
Retention inputObserved cohort curve, assumed churn rate, survival modelThe observed curve from the cohort table wherever cohorts are old enough; an assumed rate only for new businesses, labelled as such
Expansion and referralsIncluded, excludedExpansion revenue included where it is measurable; referred customers counted in their own right, with the referral credited to the loop, not to the referrer's value
DiscountingNone, cost of capitalDiscount where the horizon exceeds a year or capital is constrained
Costs of retentionIgnored, deductedDeduct the costs of keeping the customer (service, loyalty rewards, retention offers) from the margin

The formula that results is simple to state and only as good as its inputs.

LTV = contribution margin per customer per period, summed over the periods a typical customer of that cohort is retained(currency)

The retention input is the one most often faked. A lifetime value "projected to churn" using a churn rate taken from the whole base, where old loyal customers dilute the churn of new ones, can be several times the figure that an honest cohort curve supports. The cohort analysis tutorial builds the curve; the retention and payback guide explains how to read it into a lifetime figure, including the case where the curve flattens and a projected lifetime becomes defensible.

Defining CAC: blended vs channel

Customer acquisition cost is acquisition spend divided by customers acquired, and both halves of the fraction hide choices.

What counts as spend. The narrow definition is media cost. The honest definition adds the people who run the channels, the agencies and tools, the creative, the incentives (discounts, free trials, referral credits) and the share of brand spend that the business decides to allocate to acquisition. Finance will insist on the honest definition; marketing dashboards show the narrow one; the gap between them is where most CAC arguments live. Agree the definition once and label the figure with it.

What counts as acquired. A paying customer, not a sign-up, a lead or a trial. Counting trials makes CAC look low and lifetime value look low in the same stroke, and the ratio can appear fine while the business loses money on every trial that never converts. Whether a customer is new should come from the order system or CRM, not from analytics' count of new users, as the guide to new and returning customers explains.

Blended or channel. Blended CAC divides all acquisition spend by all new customers, including those who arrived organically, through referral or through a loop. It is the right figure for one question: is the business as a whole acquiring customers at a cost it can bear? It is the wrong figure for every budget decision, because it lets organic customers subsidise expensive channels. Channel CAC divides a channel's spend by the customers attributed to it, and that is the figure for deciding where the next pound goes. The two drift apart as a business grows: a healthy loop drives blended CAC down while paid channel CAC rises with saturation, and a team that watches only the blend will keep spending into a channel that stopped paying back.

Marginal CAC. The most useful and least reported figure: what the last tranche of spend cost per customer. Channels saturate, so the average understates what the next increase will cost. Where the data allows, read CAC by spend level, not just by channel.

Attribution sits under all of this. Channel CAC depends on the attribution rule, and the rule should be the business's own, applied consistently, rather than each platform's self-report; the budget allocation guide covers how a consistent rule feeds the allocation decision.

How customer acquisition cost differs from the cost per acquisition each platform reports, and which belongs in which report, is covered in the CPA versus CAC comparison.

Margin, not revenue

The single most common error in lifetime value is computing it on revenue. A customer who buys goods with thin margins, expensive fulfilment and a loyalty discount can show a large revenue lifetime value and a contribution lifetime value close to nothing. The ratio against acquisition cost then looks healthy while the business funds its growth from investors' money.

Contribution margin is revenue less every cost that varies with the order or the customer: cost of goods, packaging and delivery, payment processing, returns, variable customer service, and ongoing incentives such as loyalty rewards or subscription discounts. Fixed costs are excluded, which is why contribution is the right basis for the marginal decision "should we acquire one more customer like this": that customer brings their contribution and none of the fixed cost.

Average order value (AOV) is part of the arithmetic, but it is revenue, and raising it by discounting or by free-delivery thresholds can lower contribution per order even as it raises revenue. The break-even calculator turns margin and acquisition cost into the order count a customer must reach before they are profitable, which is a useful first check before the full lifetime value is built. For paid channels, return on ad spend (see the ROAS glossary entry) is a revenue measure; the margin-based version, which the growth team should insist on, is return on investment (ROI) on contribution, and the ROAS calculator is the place to see how the two differ for a given margin.

Ratio and payback

Two summaries matter, and they answer different questions.

The ratio of lifetime value to acquisition cost says whether customers are worth acquiring at all. A ratio near one means the business recovers its acquisition cost and nothing more; a ratio below one means growth destroys value; a ratio comfortably above one means there is room to spend more or to accept a higher CAC for faster growth. Folklore offers a single threshold for a healthy ratio; it is a rule of thumb that depends on margin structure, capital cost and how reliable the lifetime projection is, and it should never be quoted without the definitions behind the two numbers.

The payback period is the number of periods until a customer's cumulative contribution equals their acquisition cost. It answers a cash question the ratio cannot: how long is the business out of pocket? A subscription with an excellent ratio but a payback of years needs capital to grow; a retailer with a modest ratio and payback inside the first order can grow from its own cash flow.

Payback period = the first period at which cumulative contribution margin per customer reaches acquisition cost, read from the cohort's cumulative margin curve(periods)

Read together, the two figures classify every channel and segment: good ratio and fast payback (scale it); good ratio and slow payback (scale it as cash allows); poor ratio and fast payback (fine for cash, but there is no lifetime behind it, usually a one-off purchase segment); poor ratio and slow payback (stop). The calculator below takes acquisition cost, contribution margin per period and a retention assumption and returns the ratio and payback, so that the classification can be tested on real inputs before the cohort table is built.

LTV and CAC calculator

Pitfalls

The ways these figures go wrong are predictable enough to list. Lifetime value computed on revenue; retention taken from the whole base rather than the cohort; trials counted as customers; incentives left out of acquisition cost; organic customers blended into paid channel CAC; and lifetime projected beyond any horizon the business has observed. Each inflates the ratio, and several together can turn a loss-making channel into a dashboard hero. The example below runs the same channel through both sets of definitions.

Illustrative example — not a real result

The same channel, two sets of definitions

A subscription coffee company reports a paid social LTV:CAC ratio of 4.8 and asks whether to double spend. The first version of the figures: average revenue per subscriber per month £24, projected lifetime 14 months from the whole-base churn rate of 7% a month, giving an LTV of £336; media spend £68,000 divided by 970 new subscribers gives a CAC of £70. Rebuilt with the definitions above: contribution per subscriber per month is £9.60 after coffee, packaging, delivery and payment costs; the paid social cohort's own curve shows 55% still subscribed at month three and 31% at month six, implying an expected lifetime of just over five months within a twelve-month window; LTV on contribution is £52. Acquisition cost adds the first-month half-price offer (£12 per subscriber), the agency fee (£9,000) and the share of the creative team's time (£6,000), giving £94,640 over the same 970 subscribers: a CAC of £98. The honest ratio is 0.53, and payback never arrives within the window. The channel is not doubled; it is narrowed to the audiences whose cohort curve matches the company's email-acquired subscribers, which pay back in four months.

Using LTV and CAC in decisions

Once the definitions are honest, the two numbers earn their keep in four recurring decisions.

Budget allocation. Channels are ranked by channel CAC against the lifetime value of the customers they bring, not by cost per order, and the budget moves towards the channels with the best ratio and acceptable payback. Marginal CAC decides how far each channel can be pushed before the ratio collapses. The budget allocation guide covers the method.

Target setting. A target CAC for a channel is derived from lifetime value and the ratio the business can accept, not from last year's CAC. Automated bidding targets in paid platforms should be set from this figure, which is one of the few places where growth economics directly instruct a channel. The same logic sets what the business can pay for an email subscriber who is not yet a customer, and the guide to what an email subscriber is worth builds that figure from sign-up cohorts and a holdout.

Growth speed. Payback period and cash on hand together set how fast the business can grow without raising money. The retention and payback guide works through that constraint.

Where to invest beyond acquisition. If lifetime value is capped by early churn, activation and onboarding are the best investments; if by order frequency, lifecycle marketing is; if by margin, pricing and fulfilment are. The decomposition of lifetime value into retention, frequency and margin tells the team which lever is loose.

A final discipline: refresh both figures monthly from the cohort table, by channel, and keep the definitions in the same document as the numbers. A lifetime value with no stated definition is an opinion.

To place these economics inside a lifecycle dashboard, AARRR for ecommerce sales prices the first order as activation and counts revenue per customer over a fixed horizon.

To turn a payback window into a working acquisition ceiling, the break-even calculator guide for subscription businesses enters revenue per subscriber after churn instead of a single order.

When the acquisition cost series climbs month after month, the diagnosis of customer acquisition cost rising splits it into traffic price, conversion and close rate, and checks the definition and the channel mix before blaming the market.

Changes to the sign-up flow move CAC as much as conversion, and the card-required trial test shows how to judge asking for payment details at sign-up on paying customers per visitor rather than per trial.

Common mistakes

  • Revenue LTV. The commonest and most expensive error; contribution margin or nothing.
  • Whole-base churn. Old loyal customers dilute the churn of the cohort being judged; use the cohort curve.
  • Trials as customers. Both numbers shrink and the ratio flatters.
  • Media-only CAC. People, agencies, tools, creative and incentives are acquisition costs.
  • Blended CAC for channel decisions. Organic customers end up subsidising the channel that stopped paying back.
  • Projecting beyond the data. A lifetime longer than the oldest cohort the business has observed is a guess; label it.
  • Quoting a ratio without its definitions. Two teams with different margin definitions will produce different ratios from the same customers.
  • Never reading marginal CAC. The average hides the saturation that the next budget increase will hit.

Frequently asked questions

Should LTV be calculated on revenue or margin?

On contribution margin: revenue less cost of goods, fulfilment, payment, returns, variable service and ongoing incentives. A revenue figure answers no decision a growth team faces, because the question is always whether one more customer of this kind is worth the cost of acquiring them, and only margin can be compared with cost. Show the revenue figure if stakeholders expect it, but label it and never compare it with CAC.

What is the difference between blended and channel CAC?

Blended CAC divides all acquisition spend by all new customers, including organic and referred ones; channel CAC divides one channel's spend by the customers attributed to it. The blend answers whether the business as a whole acquires at an affordable cost. Channel CAC answers where the next budget increase should go, and marginal CAC, the cost of the most recent tranche of spend, answers how far a channel can be pushed.

How long a horizon should LTV use?

No longer than the business has observed. A fixed window such as a year, read from the cohort table, is honest and comparable across channels; a projection to churn is useful when the cohort curve has clearly flattened, and misleading when it has not. Show the windowed figure as the decision number and the projection alongside it, labelled as a projection with its retention assumption.

Terms used on this page

Tools to try

Frameworks

Where this page is used

Part of these learning paths

Also related

Related