How do I calculate customer lifetime value?

Calculate customer lifetime value by forecasting the receipts and costs associated with a defined customer or customer group, placing them on their actual expected dates, and discounting the resulting cash flows to a stated valuation date. Say whether acquisition cost is included. Revenue alone leaves out the work of serving the customer; a forecast is not money already collected.

My first question about a large customer-value number would be: value measured how? A lifetime revenue total, gross-margin estimate, discounted customer contribution and whole-company profit are different things. The label CLV does not tell you which one a spreadsheet has produced.

Define the number before calculating

Write down the customer group, valuation date, forecast period, cash flows included and treatment of acquisition cost. State whether the calculation begins before the first transaction or after some work and payment have already happened. Keep completed historical activity separate from remaining future value.

For costs, identify the relevant materials, fulfillment, support, payment charges and other resources needed to supply the forecast work. If the result excludes corporate overhead, tax, financing or another cost, name the limit. Do not quietly present contribution before those costs as complete profit.

A fixed three-year projection should be labeled that way. It is not a complete lifetime estimate unless the relationship really ends then or the remaining value has been appropriately addressed. A terminal value is another forecast requiring support, not an automatic bonus added to improve the result.

In their note on CLV formulas, Peter Fader and Bruce Hardie show why first-payment inclusion and cash-flow timing change the answer. They also discuss the limits of constant retention and the difference between contractual and noncontractual settings. That is a reason to establish the model's meaning before using a convenient formula.

Keep the customer group intact

A cohort is a specified group followed through time. Customers acquired later do not belong in the original group's future receipts merely because the company has grown. Make the inclusion rule clear enough that another person can reconstruct it.

For an annually renewing service, you may observe which contracts renew. For a shop, a customer who has not bought recently may simply be between purchases. Do not treat a quiet shopper as an observed contractual cancellation without a sound basis.

The retention guide separates customer count from recurring-value retention. Lifetime valuation adds future costs, timing and assumptions; a retention percentage alone does not supply those.

Use appropriate records and protect customer information. You need a reliable group and transaction history, not unnecessary sensitive details. Identify refunds, credits, collection problems and service costs rather than retaining only the largest sales total.

Build a bounded three-year case

Consider an entirely fictional annually renewing service with one hundred customers in its starting cohort. The planning case assumes one hundred customers buy in year one, seventy in year two and fifty in year three. No new customers enter this cohort. These are supplied forecasts, not observed renewals or industry retention rates.

Each active customer pays $120 at the end of the relevant service year. Supplying that year's service requires $80 per customer, paid at its beginning. The complete $80 includes $45 materials, $25 paid fulfillment work and $10 support and administration. The paid-work resources are available without displacing accepted obligations.

Acquiring the original cohort costs $1,200 immediately: $900 of channel payments and seven and a half paid preparation hours at $40, costing $300. Acquisition is counted once. It is separate from the $80 service cost.

The comparison is before corporate overhead, tax and financing. It contains no capital purchase, additional acquisition, refund, bad debt or other cash movement. Those exclusions define this simplified contribution model; they should not be carried into a real business if those events matter.

The three-year service receipts are $12,000, $8,400 and $6,000, totaling $26,400. Service payments are $8,000, $5,600 and $4,000, totaling $17,600. Before acquisition and discounting, the difference is $8,800. After the $1,200 acquisition cost, it is $7,600.

Dividing receipts by the original one hundred customers gives $264 of three-year revenue per starting customer. That is not $264 of profit, cash available today or permission to spend $264 acquiring the next customer.

Put cash on its actual assumed date

The service costs occur before the corresponding customer receipts. Keep that fact in the valuation. The first $8,000 service payment and $1,200 acquisition payment occur now, at time zero.

At the end of year one, $12,000 arrives. Immediately afterward, at the same modeled date, the business pays $5,600 to serve the seventy year-two customers. The net movement at that point is $6,400.

At the end of year two, $8,400 arrives and $4,000 is paid for the fifty year-three customers, leaving a net movement of $4,400. At the end of year three, $6,000 arrives. No later service, receipt or terminal value is included.

Scroll horizontally if needed. Costs are placed at the beginning of service, not moved to the collection date.

The undiscounted movements still total $7,600. Rearranging the table by payment date has not created value. It has made the timing explicit enough to perform the next calculation correctly.

Discount with an explicit convention

Use an assumed annual discount rate of 10% solely for this illustration. It is not a recommended rate for your company or a forecast investment return. Choosing an appropriate rate in practice requires reasoning about the actual valuation and risk.

Divide a cash movement one year away by 1.10, two years away by 1.10 squared, and three years away by 1.10 cubed. The immediate $9,200 payment is already at the valuation date.

The cohort's modeled present value is −$9,200 + $6,400 ÷ 1.10 + $4,400 ÷ 1.21 + $6,000 ÷ 1.331 = $4,762.43. Calculations retain precision internally; displayed money is rounded to cents.

Across the original one hundred customers, that is $47.62 per starting customer after acquisition cost, for the specified three-year horizon and exclusions. Before the $1,200 acquisition cost, cohort value would be $5,962.43. Report which convention you use rather than comparing those two figures as if one calculation were wrong.

Do not discount each year's $4,000, $2,800 and $2,000 service surplus at year end and then subtract acquisition. That would place the service costs later than this case actually pays them. It produces $6,253.04, overstating this particular correctly timed value by $1,490.61.

Check funding separately from value

Give the case $10,000 of opening cash, separately assigned after protecting household and other business obligations. Paying $9,200 now leaves $800. The first year's receipts then take cash to $12,800; paying for year two leaves $7,200.

At the end of year two, receiving $8,400 takes cash to $15,600. Paying $4,000 for year three leaves $11,600. The final $6,000 receipt brings closing cash to $17,600. The $7,600 increase agrees with the undiscounted total.

The plan can fund the stated payments. The $4,762.43 present value is not another deposit to add to the $17,600. One is a valuation at time zero; the other is the cash balance after following the assumed schedule.

The existing acquisition-budget guide examines what can actually be spent and funded to win customers. A projected positive customer value does not establish an affordable acquisition allowance or supply evidence that a particular campaign will work.

Test the service-cost assumption

Suppose the complete service payment is $90 instead of $80 per active customer, with counts, prices, dates and acquisition cost unchanged. The required service payments become $9,000, $6,300 and $4,500.

The immediate outlay becomes $10,200, followed by net movements of $5,700, $3,900 and $6,000. At the same assumed discount rate, present value is $2,712.85, or $2,049.59 less than the original case.

It is still positive. Yet the original $10,000 opening cash is $200 short of the initial payment. That case is not executable without another supported funding arrangement or a legitimately different payment schedule. Favorable lifetime economics cannot pay a bill before cash is available.

Customer counts and collection dates also deserve examination. If fewer renew, service costs may fall too; if some costs remain committed, they may not. Model the actual relationship rather than reducing revenue while assuming every cost disappears proportionately.

Use the estimate within its limits

The fictional seventy and fifty customers are not a measured survival model. In real work, examine comparable cohorts, their age and changing service conditions. A new group observed for six months does not supply three years of actual behavior.

Keep forecasts versioned and compare them with later results. Investigate differences in price, renewals, collection, support demand and fulfillment costs. Do not rewrite the old forecast to make it look as though the later outcome was always expected.

For an existing customer, decide whether you are valuing the remaining relationship from today or including the original acquisition and first transaction. Costs already paid can matter for evaluating a past campaign without being future payments in a new decision.

A high estimated customer value also does not establish the effect of a proposed retention gesture. The repeat-business guide compares what might happen with and without that additional action. The whole relationship's value is not automatically the benefit caused by another expense.

Keep the model readable

1. Define the cohort, valuation date, horizon and inclusion of acquisition cost.
2. Separate historical observations from future assumptions.
3. Forecast service receipts and complete relevant resources by period.
4. Place each receipt and payment on its expected date.
5. Apply the stated discount convention and retain precision until display.
6. Check funding and meaningful changes independently of the value estimate.
7. Compare later evidence with the original forecast and revise the next decision honestly.