Customer lifetime value: the number that justifies what you spend to win someone

Customer lifetime value is the total profit you expect from a customer across the whole relationship. It answers the question every acquisition decision depends on: how much is it rational to spend to win one? Without it, an advertising budget is a guess, and a discount is either an investment or a loss with no way to tell which.

It is also the number most often inflated, in a way that is rarely deliberate. Three habits do most of the damage: calculating on revenue instead of margin, assuming a lifetime longer than any customer has actually stayed, and ignoring that money arriving in four years is worth less than money arriving now.

The calculation, in ascending order of honesty

1. Simplest: average purchase value multiplied by purchases per year multiplied by expected years. Quick, and overstates almost every time.
2. Better: use gross margin per purchase rather than revenue, so the number reflects what you keep. This is the single most important correction, and it comes straight out of gross margin.
3. Better still: derive the lifetime from your actual retention. If you lose a fifth of customers each year, the average lifetime is five years — the reciprocal of the churn rate, covered in churn rate.
4. Honest: cap the horizon at three to five years even when retention implies longer. Nobody can defend a projection about a customer relationship a decade out.
5. Optional: discount future cash to present value. Worth doing for a subscription business, over-engineering for a shop.

Calculate lifetime value on margin, never on revenue. A customer spending 10 000 a year at a 20% margin is worth 2 000 a year to you, not 10 000. Businesses that quote the revenue figure end up spending acquisition budgets against money that belongs to their suppliers, and the error is invisible until the year-end accounts show volume up and profit down.

Segment it, or it misleads

• By acquisition channel. Customers won through referral routinely outlast customers won through discounting, and the averaged figure hides it.
• By first purchase. What somebody bought first often predicts how long they stay better than anything else you know about them.
• By cohort — customers who joined in the same period, tracked forward. This is the only way to see whether lifetime value is improving or you are simply averaging over an older, better group.
• By size. In most businesses a small number of customers account for a large share of the value, and treating them as average is how they leave without anyone noticing.

What to do with it

The number is worth calculating only if it changes a decision, and it changes four. What you can afford to spend on acquisition, which is the pairing with customer acquisition cost. Which channels deserve budget, since a channel bringing cheap customers who leave quickly is worse than an expensive one bringing customers who stay. How much a retention improvement is worth, which is usually more than anyone expects. And which customers deserve unusual effort — the answer is not always the loudest.

Frequency can be bought deliberately. A loyalty program trades a point of margin for repeat visits, and it works when the reward feels reachable rather than aspirational.

Where it comes from

Ettex CRM holds the customer records, purchase history and source that any segmented calculation needs, with the sales themselves in Ettex Invoices and the margin figures from Ettex Books. The arithmetic belongs in Ettex Sheets, where the horizon and retention assumptions stay visible instead of being buried in a dashboard.

To be direct: there is no predictive model here, no cohort analysis engine and no automatic lifetime value estimate. That is partly a limitation and partly deliberate — a projected number generated by software gets quoted with a confidence it has not earned, while the same number calculated by hand comes with its assumptions attached.