To work out customer lifetime value for revshare, multiply the average revenue a referred customer generates each month by the share of customers still paying that month, add the months up over a realistic period, apply any deductions the program makes, then multiply by your revshare percentage. That figure is what one referral is worth to you. Compare it with the CPA on offer: if your lifetime value is higher, after allowing for risk and the wait to be paid, revshare wins.

The model below uses hypothetical figures to show the method. Swap in your own data from your dashboard or the program's cohort reports.

What customer lifetime value means in revshare

For a merchant, lifetime value (LTV) is the total revenue a customer produces. For an affiliate, LTV is your share of that revenue. Three inputs drive it:

  • Spend per period – a subscription price, average rebill, or average monthly spend on tokens, credits or deposits.
  • Retention – the percentage of referred customers still paying in each month after sign-up.
  • Your share after deductions – the revshare rate applied to the revenue base defined in your terms (gross, or net of refunds, chargebacks, processing fees, taxes or bonuses).

On cam, dating credit and gaming-style offers the same calculation is often called player value calculation: average spend per active player per month, multiplied by the share of players still active.

Step 1: Gather your inputs

  1. Find the revenue base. Read the revshare definition in your agreement. A 50% share of net revenue can be worth less than 40% of gross. Our guide to what revshare is explains common definitions.
  2. Pull cohort data. Ideally, revenue per customer by month since sign-up, grouped by the month they joined. Many programs show this; if not, ask your affiliate manager.
  3. If you have no history, ask the program for typical retention for your traffic type and GEO, then run your model with a pessimistic version of their figures.
  4. Note payment terms. Payment frequency, minimum payouts and any negative carryover clause all change what the revenue is worth to you.

Step 2: Build a month-by-month LTV model

Hypothetical example. A membership site charges €30 a month. The program pays 50% revshare on net revenue, and deducts around 15% for refunds, chargebacks and processing. That leaves €30 × 0.85 × 0.5 = €12.75 per paying customer per month to you. The CPA alternative is €35 per sale.

Retention per 100 referred customers, and your cumulative earnings per customer:

MonthStill paying (of 100)Your revenue per referralCumulative LTV
1100€12.75€12.75
245€5.74€18.49
330€3.82€22.31
424€3.06€25.37
520€2.55€27.92
617€2.17€30.09
715€1.91€32.00
813€1.66€33.66
912€1.53€35.19
1011€1.40€36.59
1110€1.27€37.87
129€1.15€39.02

The pattern of a steep drop after the first rebill, then a long, slowly declining tail is typical of subscriptions, but your own curve may look quite different.

Step 3: Find the revshare break-even point

In this example the cumulative LTV passes the €35 CPA in month 9. Over 12 months revshare earns €39.02 per referral, about 11% more than CPA. Two other break-even figures are useful:

  • Break-even revshare rate. Divide the CPA by the 12-month net revenue per customer. Here, 3.06 average payments × €30 × 0.85 = €78.03 of net revenue, so €35 ÷ €78.03 ≈ 45%. Any rate below that loses to the CPA over a year.
  • Break-even horizon. How long you must keep referrals active, and the program must keep paying, before revshare overtakes CPA. Nine months is a long time to rely on one partner.

Rule of thumb: if revshare only overtakes CPA late in the second half of the year, the extra earnings may not compensate for the risk. If it overtakes within the first few months, revshare is usually the stronger deal for quality traffic.

Step 4: Adjust for risk and the wait to get paid

A euro paid in month 12 is worth less to you than a euro paid now. Before you decide, apply some realistic haircuts:

  • Program risk. Terms change, programs close and revshare can be cut. If you discount the model by 20% for that risk, the 12-month LTV falls to about €31, below the CPA.
  • Cash flow. CPA usually pays within a month or two; revshare pays over a year or more. If you reinvest in paid traffic, the faster CPA cash may compound faster.
  • Tracking and attribution. Customers who later switch device, or rejoin through another link, may drop out of your stats.
  • Lifetime terms. A truly lifetime revshare adds the tail beyond month 12. A capped term (say, 24 months) does not.

Step 5: Test how sensitive the result is

Run the same model with retention 25% lower and 25% higher from month 2 onwards. In the example:

Scenario12-month LTVVersus €35 CPA
Lower retention (−25%)€32.45CPA wins
Base case€39.02Revshare wins slightly
Higher retention (+25%)€45.58Revshare wins clearly

When the decision flips within a plausible range, you do not have a clear winner. That is a sign to test both, or negotiate a hybrid deal that pays a smaller CPA plus a lower revshare.

When revshare beats CPA

Revshare tends to beat CPA when:

  • Your traffic is high-intent and loyal: review-site readers, email subscribers, returning visitors from SEO.
  • The product has strong retention and recurring rebills.
  • The program is established, reports transparently at customer level and pays reliably.
  • You can afford to wait for income to build.

CPA tends to win for paid traffic where you need fast cash to fund the next buy, for low-retention or impulse traffic, and with programs you have not yet tested.

Track real LTV once you start

Your model is a forecast; your earnings per referral are the reality. Each month, divide total revshare earned from a cohort by the number of customers in it, and compare with the forecast. Earnings per click (EPC) over a long window is a useful cross-check across programs; see how to calculate EPC. Tag traffic with sub-IDs so you can see which sources produce the most valuable customers; our guide to sub-IDs shows how.

Common LTV mistakes affiliates make

  • Using the program's best-case figures. Averages across all affiliates can include brand and email traffic that retains far better than yours.
  • Ignoring the revenue definition. Modelling on the advertised price instead of the net revenue you are actually paid on overstates LTV.
  • Mixing cohorts. Customers from a seasonal spike or a discount promotion often churn faster. Model them separately.
  • Forgetting traffic costs. If you buy traffic, compare LTV with your cost per acquisition, not just with the CPA offer. A revshare deal that breaks even against CPA in month 9 may not cover your media spend until much later.
  • Never updating the model. Revisit it quarterly with real cohort data; retention can shift when a merchant changes pricing or its product.

Frequently asked questions

How long should an LTV model run?

Twelve months is a sensible default, because forecasts beyond that become unreliable. Extend it to 24 months only if you have cohort data that far out and the program pays for life.

What if the program will not share retention data?

Run a small revshare test alongside CPA for a few months and build your own curve. A program that refuses to share any cohort data at all is a weak candidate for a long-term revshare commitment.

Does negative carryover affect LTV?

Yes, mainly on gaming-style offers where a customer's winnings or refunds can make monthly revenue negative. If negatives carry over, they reduce future earnings, so model them in as a deduction.

Next steps

Build the model in a spreadsheet with your own inputs, then use the break-even rate as your negotiating floor. For a refresher on the trade-offs between models, read CPA, RevShare and hybrid deals explained.