RevShare vs CPA: Which Model Fits Your Traffic

RevShare wins when your referred users stay active for months and generate recurring revenue; CPA wins when you need predictable cash per signup or your traffic converts once and leaves. The crossover depends on referral lifetime value - a simple spreadsheet comparison over 24 months beats any rule of thumb.
The core trade-off
RevShare is a bet on lifetime value: you earn a slice of every fee your referrals generate, forever (if the terms say lifetime). CPA is a bet on volume: a fixed, known amount per qualified signup, with no tail.
Neither is universally better. The right model depends on three things you can estimate from your own data:
- Conversion quality: what share of your clicks sign up AND deposit.
- Referral activity: how long referred users stay active and how much revenue they generate per month.
- Your cash needs: revshare income starts small and compounds; CPA pays from the first qualifying signup.
A worked comparison
Take a crypto casino as the example (the math is identical for exchanges, with trading fees instead of house edge):
- You send 1,000 visitors a month. 3% sign up and deposit (30 players).
- Average deposit: $200. Expected house-edge revenue per depositor: roughly $40 in month one, decaying over time.
- RevShare offer: 40% of net gaming revenue, no negative carryover.
- CPA offer: $80 per first-time depositor.
Month one: RevShare pays 30 × $40 × 40% = $480. CPA pays 30 × $80 = $2,400. CPA looks 5x better - but that is only month one. If those 30 players keep losing $40/month to the house for an average of six months, lifetime revshare from that single cohort is 30 × $40 × 6 × 40% = $2,880 - and cohorts stack: after a year of consistent traffic, you are earning on twelve cohorts at once while the CPA column stays flat.
The crossover in this example sits around month four to six. Run your own numbers with the RevShare vs CPA calculator - it takes your traffic, conversion and revenue assumptions and shows the cumulative curves for both models.
When CPA is the right answer anyway
- You cannot verify retention. If you have no data on how long referred users stay active, you are guessing at the revshare tail. CPA prices that uncertainty away.
- Your traffic is one-shot. Airdrop hunters, bonus hunters and one-time visitors rarely become recurring revenue. For incentivized or deal-seeking traffic, CPA (if the program allows that traffic at all) is the only honest model.
- The program is young or unproven. Locking in CPA reduces your exposure to a program that might close (see the scam archive). A hybrid - small CPA plus reduced revshare - splits the difference.
The traps that change the math
- Negative carryover turns a good revshare month negative and can freeze payouts. Under carried carryover, the lifetime-value math must include the possibility that a hot streak by one player wipes out your entire book. Payout speed differs just as much between programs - several crypto-native operators settle daily [2], while the classic default is monthly - so the same headline rate can land at very different cash-flow intervals.
- Minimum activity clauses can close your account before the compounding starts. If a program closes inactive accounts and forfeits balances, a slow-building revshare strategy carries execution risk, not just market risk.
- Retroactive rate cuts exist: some programs have reduced affiliate percentages with notice periods measured in days. Prefer programs with published, dated rate cards.
How to decide in practice
- Estimate referral lifetime revenue from your own data (or conservatively: assume decay to near zero after 6 months for casinos, 12 for exchanges).
- Compute cumulative income for both models over 24 months with the calculator.
- Read the program's carryover and dormancy terms before signing anything - they change which model actually pays.
- If in doubt, negotiate a hybrid: the CPA floor covers acquisition cost, the revshare tail keeps the upside.
The asymmetry most comparison tables miss
The two models price different risks. RevShare hands the operator two risks - churn and volatility - in exchange for a bigger long-run share; CPA hands both risks to you. A $80 CPA means the operator keeps the entire lifetime tail of every player you bring who turns out to be a whale. If your audience's retention is above the program's implied break-even, you are selling that tail too cheaply - the calculator makes the implied valuation visible.
Negotiating from data
Affiliate managers move on three data points: your traffic quality, your historical conversion numbers, and your tracking discipline. Bring a 30-day sample of click-to-deposit conversion and you can negotiate a tier instead of accepting a default; bring nothing and the default is what you get - usually the floor rate with carried carryover. Programs publish their tier structures more readily than they publish their negotiable ranges; both belong in your notes before the first call [3].
What we document on program pages
The model mix (revshare, CPA, hybrid), the published rate floor and ceiling where the operator states both, the carryover policy, the cookie window, payout coin/minimum/schedule, and the sub-affiliate rate. Fields the operator does not publish stay marked "not documented" - that is information too, because programs that hide terms usually hide the ones that hurt you.
Sometimes. Most networks allow model changes at the start of a billing period, but the switch usually resets your referral attribution - read the switching clause before assuming your existing referrals carry across.
What is a "hybrid" commission?
A combined deal: a reduced CPA (for example $30) plus a reduced revshare (for example 15%). You give up part of each model's upside in exchange for a floor plus a tail. It is the default answer for traffic the program cannot classify yet.
Which model suits Telegram or YouTube traffic?
Audience-based channels tend to favor CPA or high-front-loaded hybrids: referrals arrive in bursts, attribution windows are short, and retention of campaign-driven signups is usually worse than SEO traffic. The EPC calculator converts any offer into earnings-per-click so you can compare programs on the same axis.
Do all programs pay lifetime revshare?
No. Lifetime is the crypto-native standard, but fiat-era operators and several exchanges cap revshare to a fixed period (typically 3 to 12 months) [1]. The cap is one of the highest-impact fields on each program page - [source 1](#src-1).