How customer lifetime value works
Lifetime value answers one question: how much is a customer worth to you in total, and therefore how much can you afford to pay to get one? Everything else is detail. The detail matters, though, because the three most common shortcuts all push the number in the same direction, which is up.
The first shortcut is using revenue instead of gross profit. Revenue LTV will cheerfully tell you a customer is worth $2,000 when the goods cost you $1,400, and it will justify an acquisition budget that quietly bleeds you. The second is ignoring time: money arriving in year five is not worth what money arriving today is worth, and a five-year LTV that has not been discounted is a forecast wearing the costume of a fact. The third is treating retention as one flat rate, which is the one nobody talks about, and it is covered further down.
This calculator handles all three. It projects an actual year-by-year retention curve, applies your margin, discounts each future year back to today, and shows you the shortcut version alongside so you can see exactly how much optimism you were carrying.
The formula
Annual gross profit is average order value × purchases per year × gross margin. The share still active is 100% in year 1, your first-year retention rate in year 2, and then multiplied by the ongoing retention rate for each year after. d is your annual discount rate, and year 1 is not discounted because that money arrives now. The familiar textbook shortcut, LTV = annual gross profit ÷ (1 − retention), is the same idea with no discount rate and an infinite horizon, which is why it always comes out larger.
Worked example
An online shop has an $80 average order, customers buy 2.5 times a year, gross margin is 60%, first-year retention is 40%, retention after that is 65%, over a 3 year horizon at a 10% discount rate. Acquisition cost is $60.
Annual revenue per active customer is 80 × 2.5 = $200, and annual gross profit is $120. Year 1 contributes $120. Year 2 has 40% still buying, so $48, discounted to $43.64. Year 3 has 40% × 65% = 26% still buying, so $31.20, discounted to $25.79.
Total: $189.42 of gross profit per customer, against $315.70 of revenue. That is an LTV to CAC ratio of 3.16 to 1, with the $60 acquisition cost repaid after about 6 months. The old shortcut ($120 ÷ 0.60) would have said $200.
The first-year cliff, and why one retention rate lies
Ask a business its retention rate and you will get a single blended number, usually somewhere in the 50s or 60s. That number is an average of two very different populations: brand new customers, who mostly leave, and established customers, who mostly stay. Getting someone to buy a second time is by far the hardest step in the whole relationship. Once they have bought three or four times, they behave like a different species.
Blending those two into one rate does real damage in both directions. It overstates how much a new customer is worth, which is exactly the number you use to set acquisition budgets, and it understates your loyal core, which is the number you use to justify retention spending. That is why this calculator asks for two rates. If you only know one, leave the second blank and the math falls back to a single rate, but the honest move is to pull the two numbers separately from your order history. It usually takes twenty minutes and it usually changes the answer.
Using LTV as a bidding input
If you are feeding values into Google Ads, Meta, or any bidding system that optimizes toward conversion value, use the profit-based, discounted figure, and use the same basis for every conversion action you report. Bidding algorithms do not care whether your absolute numbers are right; they care intensely whether your numbers are right relative to each other. A purchase valued on revenue and a lead valued on profit will quietly teach the system to chase the wrong one.
One warning about horizon. Feeding a 10 year LTV to a system whose feedback loop is 30 days is not wrong, but it does mean your reported return on ad spend describes a decade while your bank account describes a month. Most advertisers are better served by a 2 to 3 year horizon: long enough to capture the repeat business that justifies the spend, short enough that you would actually bet cash on the forecast.