Companies have always talked about loyalty and repeat customers. What has changed in recent years is how precisely they now try to put a number on those relationships.

Across investor decks, marketing dashboards and board papers, one metric has quietly moved to the front: customer lifetime value, usually shortened to CLV or LTV. It tries to answer a simple question with far-reaching consequences: how much is an average customer really worth over the whole relationship, not just at the first purchase?

That figure now shapes how much companies spend on advertising, which customers they prioritise, how they design products and whether investors view their growth as sustainable or fragile.

What customer lifetime value actually measures

There is no single official formula for CLV, but mainstream definitions are surprisingly consistent.

In marketing and finance, CLV is usually described as the present value of the future cash flows or profit a customer is expected to generate over the entire relationship with a firm. Put less technically, CLV estimates how much money a company will make from a typical customer, after costs, from the moment they first buy until the day they finally stop.

Two elements sit at the centre of almost every definition: time and economics. CLV stretches across the full lifespan of the relationship – from first purchase to final interaction – and focuses on the underlying profitability of that relationship rather than raw turnover. It is about the flow of value over time, not a single invoice or campaign response.

Two properties make CLV distinct from simpler measures such as “lifetime revenue”:

• It is forward-looking – CLV focuses on expected future value, not just what a customer has already spent. It relies on patterns such as purchase frequency, renewal behaviour, churn and product usage to predict what is likely to happen next.
• It is usually profit-based – many practitioners stress that CLV should be calculated after gross margin or net profit, not just top-line revenue. Otherwise the number can easily overstate how valuable a customer really is.

The basic idea is not new. Academic work on lifetime value appears in marketing journals as far back as the 1990s. What has changed is the environment: e-commerce, subscription models and detailed digital tracking have made it far easier to calculate CLV in practice and update it regularly, customer by customer.

How companies actually calculate CLV

Behind the three-letter acronym sits a spectrum of methods, from back-of-the-envelope rules to models embedded in data platforms. The right approach depends heavily on the business model and the quality of available data.

The simple spend-and-frequency approach

Many businesses start with a straightforward historical formula that treats CLV as an average revenue figure:

CLV = Average purchase value × Purchase frequency × Average customer lifespan

Simple CLV formula
CLV = Average purchase value × Purchase frequency × Average customer lifespan

Average purchase value is calculated as revenue divided by the number of orders over a given period. Purchase frequency is the average number of orders per customer in the same period. Average lifespan is how long, typically in years, a customer continues to buy before lapsing.

This approach is common in retail and direct-to-consumer brands because it is easy to explain and requires only basic transaction data. To get closer to economic reality, firms often replace revenue with gross margin per purchase, effectively turning the same equation into a profit-based CLV.

Retention, churn and discounted cash flow

Subscription and contract-based businesses tend to think less in terms of individual orders and more in terms of how long customers stay. Here, a different formulation is widely used, derived from discounted cash-flow logic:

CLV = Margin per period × Retention rate ÷ (1 + Discount rate − Retention rate)

Often described as the “traditional” CLV formula, it calculates the present value of a stream of profits from a customer, assuming a steady retention rate over time.

Three inputs are crucial:

• Margin per period – typically gross margin per customer per month or year, after variable costs.
• Retention rate – the proportion of customers who stay from one period to the next; the opposite of churn.
• Discount rate – a rate chosen to reflect the time value of money and the riskiness of cash flows.

Churn and retention are two sides of the same coin: if annual churn is 25 per cent, retention is 75 per cent. A simple rule of thumb links churn to expected lifetime:

Average customer lifetime (in periods) ≈ 1 ÷ churn rate

A service with 20 per cent annual churn can therefore expect an average customer lifetime of about five years, as long as conditions remain stable.

Some subscription businesses use even simpler short-hand formulas when they are less concerned with discounting, for example:

• CLV = ARPU × Gross margin × Average contract duration – where ARPU (average revenue per user) is multiplied by both margin and an expected number of periods.
• CLV = ARPU ÷ churn rate – under assumptions of constant churn and margin, often cited in SaaS playbooks.

From historical to predictive CLV

These equations are only one part of the story. As richer data has become available, companies have moved from simple historical averages toward predictive CLV: models that estimate the future value of each individual customer or cohort.

Broadly, practitioners talk about three layers:

• Historical CLV – based purely on past spend and tenure, often averaged across cohorts. It offers a baseline but can lag reality when behaviour shifts quickly.
• Probabilistic CLV – uses statistical models to estimate how likely customers are to buy again and at what level, using transaction histories. Classic models include Pareto/NBD or BG/NBD frameworks from academic literature.
• Predictive CLV – uses machine learning and broader sets of predictors, from app usage and browsing patterns to support tickets and marketing touchpoints, to forecast value at the level of individual customers.

The more advanced the approach, the more infrastructure and expertise it requires. The trade-off is flexibility: predictive CLV can adjust quickly when customer behaviour changes, rather than waiting for a full lifecycle to play out.

CLV, CAC and the LTV:CAC ratio

On its own, CLV is a description. It becomes a decision tool when compared to what it costs to acquire customers in the first place.

Customer acquisition cost, or CAC, is usually defined as the total sales and marketing spend over a period divided by the number of new customers acquired in that period. It includes media spend, salaries and often associated software and agency fees.

The ratio between lifetime value and acquisition cost has become one of the headline metrics in software and other recurring-revenue sectors:

LTV:CAC = Customer lifetime value ÷ Customer acquisition cost

Guides from business schools and SaaS-focused platforms frequently cite a benchmark of around 3:1 – for every £1 or $1 spent acquiring a customer, the company aims for roughly £3 or $3 in lifetime value. (see Harvard Business School Online; Stripe; Lucid)

• Around 1:1 – the business is roughly breaking even on new customers; growth at this level is hard to sustain without outside capital.
• Roughly 2:1 to 3:1 – unit economics are generally seen as healthy, with room to cover overheads and invest in product and service.
• Comfortably above 3:1 – suggests strong economics; once the ratio climbs beyond about 5:1, it may signal under-investment in marketing and sales, with potential growth left untapped.

Investors and operators often pair the ratio with a second measure: payback period, the time it takes for the gross margin from a typical customer to repay the acquisition cost. Shorter payback periods generally lower risk and expand strategic options.

How CLV changes decisions inside a business

Once companies have even a rough estimate of CLV, it tends to seep into a wide range of decisions.

In marketing, CLV is used to compare channels and campaigns. If customers from one advertising channel have twice the lifetime value of those from another, teams can in principle afford to bid more aggressively on the high-value channel, even if the initial cost per acquisition looks higher. CLV per segment also influences how much discounting or promotional spend makes sense for different audiences.

In product teams, CLV by cohort or behaviour pattern can highlight which features correlate with more durable relationships. When customers who use a particular feature or bundle have significantly higher lifetime value, this can shape product roadmaps, onboarding flows and in-app nudges, even if the feature itself does not drive much direct revenue.

Sales organisations use CLV to prioritise segments and accounts.

Higher-value segments may receive more tailored outreach, richer service or different contract structures than low-value, high-churn groups. Commercial leaders may also design compensation schemes that reward not just closing deals, but attracting customers who stay and expand.

Finance and leadership teams rely on CLV as part of unit economics. Taken across the entire customer base, CLV connects to the idea of “customer equity” – the sum of the lifetime value of all customers – which in turn links to valuations and strategic choices. Work by academics and practitioners in Harvard Business Review has argued that disclosing such customer metrics can give investors a clearer view of a company’s underlying health.

The limits and risks of relying on CLV

Despite its appeal, CLV is not a magic number. It rests on assumptions and data that can be fragile, and missteps can be costly.

• Model risk – CLV formulas rely on assumptions about churn, margin and discount rates. If those assumptions are wrong, the output can be badly mis-stated. A formula that treated a cohort as unprofitable may, in reality, have written off customers who would have become valuable later.
• Data quality – reliable CLV estimates depend on accurate records of orders, cancellations, discounts and returns, all linked to the right customer identifiers. In practice, many firms still struggle with fragmented systems and incomplete histories.
• Overconfidence in averages – a single average CLV can hide huge differences between segments. High-value customers may subsidise low-value or even loss-making groups. Without segmentation, management risks designing for an imaginary “average” customer who does not really exist.
• Ethics and fairness – if CLV feeds into who receives better prices, faster service or more lenient terms, the models behind it can become a quiet source of inequality. Data used to predict value may reflect historical biases; CLV-driven targeting can entrench them unless checked.
• Changing environments – CLV is always a product of assumptions about the future. Shifts in the economy, regulation, competition or technology can rapidly shorten or lengthen typical customer lifetimes. Over-reliance on yesterday’s CLV can lock decisions to a world that no longer exists.
• Correlation versus causation – the fact that customers with high CLV tend to use a certain feature or channel does not prove that feature causes the value. Treating correlations as causal can lead teams to invest heavily in the wrong levers.

Harvard Business Review has gone so far as to highlight “the flaw in customer lifetime value” when used rigidly, arguing that the traditional calculation can lead companies to undervalue certain segments that competitors later turn into profitable customers. (Harvard Business Review)

A single metric, many futures

Customer lifetime value is, at root, an attempt to collapse a complex relationship into a single number.

Done wrongly, it can mislead.

Done thoughtfully, it can pull companies away from short-term extraction towards longer-term relationships, where the goal is not simply to make a sale, but to keep earning the right to another one.

Used alongside measures of acquisition cost, payback and risk, CLV offers a way to judge whether growth is creating genuine economic value or simply buying revenue at too high a price. It does not predict the future with certainty, but it does force a harder question than “What did we earn this quarter?”

The more revealing question, and the one CLV tries to answer, is this: if a company keeps acquiring customers in the way it does today, what is each new relationship likely to be worth in the years ahead – and is that enough to justify the effort?