📖 Glossary

Customer Lifetime Value (CLV) — Definition, Evidence & Common Errors

Customer Lifetime Value estimates the total net profit a buyer generates over the relationship. Here's how it works, what the Ehrenberg-Bass evidence says about loyalty, and where it misleads.

·5 min read

Definition (≤25 words). Customer Lifetime Value is the total net profit a business expects to earn from a customer across the entire duration of the relationship.

TL;DR. CLV is a forward-looking estimate that turns a buyer into a balance-sheet item: if a customer is worth £600 over their life with you, that caps what you can sensibly spend to win and keep them. Sensible in principle — and frequently misused in practice, because the loyalty assumptions baked into most CLV models collide with half a century of Ehrenberg-Bass evidence that retention is hard to move and growth comes from penetration. Treat CLV as a budgeting guardrail, not a strategy.

What problem is CLV trying to solve?

The default unit of marketing accounting is the transaction: how much did this campaign sell, this month, at what cost. That framing rewards short-term, sales-led work and starves the longer view. CLV was invented to fix that by making the customer — not the campaign — the unit of value. If you can estimate what a typical buyer is worth over five or ten years, you can justify spending more to acquire them than a single transaction would cover, and you can decide whether keeping an existing buyer is cheaper than winning a new one.

That logic is sound, and it is why CLV sits at the centre of subscription businesses, telecoms, insurance, and any model where revenue is recurring. The trouble begins when the estimate is bolted onto categories where the underlying loyalty dynamics do not cooperate.

How is CLV actually calculated?

At its simplest, CLV is the expected profit a customer generates, discounted back to today:

  • Average purchase value × purchase frequency × gross margin, summed across
  • The customer's expected lifespan (driven by retention or churn rate),
  • Minus the costs of acquiring and serving that customer.

The arithmetic is not the hard part. The hard part is the inputs — particularly the retention rate and the lifespan it implies. A small change in assumed churn moves the whole number dramatically, which is why CLV figures in vendor decks tend to be heroic. A model that assumes you will keep a buyer for seven years when your category's actual retention barely clears three is not estimating value; it is manufacturing permission to overspend.

What does the evidence say about customer lifetime — and loyalty?

This is where CLV meets the evidence base it was never designed around. The Ehrenberg-Bass Institute's work on how buyers actually behave undercuts several assumptions that CLV-led strategies depend on:

  1. Retention rates are high, but strikingly similar across competing brands. This is part of the double jeopardy pattern. In most established categories, the difference in loyalty between the market leader and a small brand is modest — far smaller than the difference in their market penetration. You cannot CLV your way to a dramatically stickier customer base while a competitor with comparable service has the same churn. What separates big brands from small ones is how many buyers they have, not how loyal each one is.

  2. Growth comes from penetration, not from deepening loyalty. Byron Sharp, in How Brands Grow (2010), showed that brands grow by adding more buyers — mostly light, occasional ones — not by squeezing more from a devoted core. A CLV strategy that prioritises "maximising the value of existing customers" can quietly starve the acquisition and reach work that actually moves market share.

  3. The 80/20 rule is overstated. The popular claim that 20% of customers deliver 80% of sales does not survive contact with panel data. The heaviest fifth of buyers typically account for around half of sales, not four-fifths (the heavy buyer fallacy) — which means the "long tail" of light and lapsed buyers is far more valuable than the loyalty gospel suggests, and far harder to model neatly inside a CLV spreadsheet.

  4. Your customers are mostly other brands' customers too. The duplication of purchase law shows that competing brands share buyers in proportion to their market share — your customers routinely buy your rivals, and your rivals' customers routinely buy you. Most "loyalty" in repertoire categories is polygamous, not exclusive, which is why CLV models that assume a buyer belongs to one brand overstate how much of that buyer's lifetime spending you can ever capture.

None of this makes CLV wrong. It makes the common use of CLV — to justify ever-deeper investment in a small, loyal core — wrong, or at least badly over-rotated. The evidence points the other way: the biggest commercial opportunities sit in the large, lightly-attached majority, and the budgets that find them (broad-reach brand building) are the ones CLV thinking most often defunds.

Where people get it wrong

  • Assuming retention is infinitely improvable. CLV models often treat loyalty as a lever the brand can pull harder each year. The evidence says customer defection is remarkably stubborn and category-dependent; you can improve service and reduce churn at the margins, but you rarely double the lifespan the model assumes.
  • Confusing CLV with revenue. Many reported "lifetime value" figures are lifetime revenue, with no margin, no acquisition cost, and no discount rate applied. A customer who generates £1,000 of gross sales over five years at thin margins and high service cost may be worth less than the campaign that won them.
  • Using CLV to defend over-spending on existing customers. "It costs five times more to acquire a new customer than keep one" is one of marketing's most repeated — and least sourced — claims. Taken literally, it pushes budgets toward retention CRM and away from the penetration growth that actually drives marketing effectiveness.
  • Ignoring category context. CLV is a powerful tool in subscription and high-contract categories where churn is real and measurable. In repertoire, FMCG-style categories where buyers flit between brands routinely, a single-brand "lifetime" is a fiction, and the model's predictions degrade fast.
  • Discounting the future at the wrong rate. CLV is a present-value calculation; using too low a discount rate inflates distant, uncertain revenue into confident balance-sheet assets.

In practice — how to use CLV without being misled

  1. Use CLV as a budgeting ceiling, not a strategy. It tells you the most you can prudently spend to win and serve a customer. It does not tell you whether to chase retention or penetration — the evidence does.
  2. Base the retention input on real panel data, not aspiration. If your category's annual retention is 70%, model 70%. Do not model the 95% you hope a new loyalty programme will deliver.
  3. Always separate CLV from revenue. Strip out cost of goods, service, and acquisition before you call something "value." Apply a discount rate.
  4. Weight acquisition toward penetration. Where the Ehrenberg-Bass evidence applies, the most productive spend reaches more, lighter buyers — not deeper rewards for the few who already buy you most.
  5. Pair CLV with long-term brand metrics. If you run a CLV model, run share of search and penetration alongside it, so the board sees whether your customer-value story is matched by real market growth.
  6. Be honest about uncertainty. CLV is a forecast. Present a range, not a point estimate, especially for categories with high switching.

So what?

Customer Lifetime Value is a useful discipline forced into the wrong job. As a guardrail — "do not spend more to win a buyer than they are worth" — it is invaluable. As a strategy — "grow by extracting more from your most loyal customers" — it collides with decades of evidence that brands grow by reaching more people, more lightly. The most expensive CLV mistake is not a wrong number; it is the quiet reallocation of budget away from the broad, unglamorous reach work that actually moves market share, toward CRM theatre aimed at a core that was never going to leave anyway.

Evidence & sources

  • Sharp, B. (2010). How Brands Grow. Oxford University Press — penetration as the driver of growth; the overstated 80/20; loyalty metrics that barely differ between brands.
  • Romaniuk, J. & Sharp, B. (2016). How Brands Grow 2. Oxford University Press — mental and physical availability over relationship-deepening.
  • Ehrenberg-Bass Institute — duplication of purchase law; stationarity of buyer behaviour.
  • Binet, L. & Field, P. (2013). The Long and the Short of It. IPA — context on long-term vs short-term effects relevant to CLV time horizons.

Definition type: DefinedTerm. Reviewed: 2026-07-09. Written by Andrei Tarnovski for Boringly Effective. Fundamentals over fashion.