How crypto exchange affiliate commission works: CPA vs lifetime revshare vs hybrid, with break-even and cohort math, tiers and the clauses that cut your pay.
Two affiliates send the same exchange the same trader. One took a $150 CPA. One took 50% lifetime revenue share. Fourteen months later the CPA affiliate has $150 that they collected on day 32, and the revenue share affiliate has $245 that arrived $17.50 at a time. Both made a defensible decision. Only one of them can tell you why.
Choosing a crypto exchange affiliate commission structure is a cash-flow decision dressed up as a rate comparison. The percentage matters much less than three things: how long your referred users actually trade, how quickly you need the money back to fund more traffic, and what the contract defines as revenue in the first place. Get those wrong and a 50% deal can pay you less than a 30% one.
What follows is the full arithmetic — a single-trader comparison, a twelve-cohort stacking model, the two distinct crossover points that people conflate, and then the parts of the contract that quietly move all of it: tiering, sub-affiliate layers, inactivity resets, fee rebates and VIP carve-outs.
CPA. A fixed amount per qualified user. Qualification normally means completed identity verification plus a minimum first deposit, sometimes plus a first trade or a volume floor. Paid once, usually after a hold period of 30 to 60 days. You have no further claim on that user.
Lifetime revenue share. A percentage of the net trading fees your referred users generate, paid monthly for as long as the attribution holds. No cap on total earnings, no guarantee of any earnings.
Hybrid. A reduced CPA plus a reduced revenue share. Something like $60 plus 25%, or $80 plus 20%. You get partial recovery of acquisition cost up front and keep a tail.
There is a fourth arrangement worth naming because it is common and rarely discussed: CPA with a revenue-share fallback, where you get the CPA only if the user meets a volume threshold, and drop to revenue share if they do not. It looks generous and mostly functions as a way for the advertiser to avoid paying full CPA on weak traffic.
Before any comparison means anything, you need to know what the percentage is applied to. On exchange programs it is nearly always net fee revenue, and net is doing a lot of work.
A trader running $70,000 of notional volume a month at a 0.05% taker fee pays $35 in fees. At 50% revenue share your commission is $17.50. Now change one variable at a time:
None of these are scams. They are standard mechanics. But an affiliate comparing "50% on Exchange A" against "40% on Exchange B" without checking the net revenue definition is comparing nothing at all.
Take one referred trader who generates $35 a month in net fees and stays active for 14 months.
Under the CPA deal: $150, received once, roughly 30 to 45 days after the deposit clears qualification.
Under 50% lifetime revenue share: $17.50 a month for 14 months, which is $245 total. That is 63% more than the CPA — but the last dollar arrives more than a year later.
Under a hybrid at $60 plus 25%: $60 up front plus $8.75 a month for 14 months, which is $122.50, for $182.50 total.
| Model | Total per trader | Received by month 2 | Received by month 9 | Time to full value |
|---|---|---|---|---|
| $150 CPA | $150.00 | $150.00 | $150.00 | ~1 month |
| 50% revshare | $245.00 | $17.50 | $157.50 | 14 months |
| $60 + 25% hybrid | $182.50 | $68.75 | $138.75 | 14 months |
The revenue share catches the CPA on this single trader at month nine — $17.50 × 9 = $157.50 against $150. Before that point the CPA affiliate is ahead and has had the cash available to reinvest for eight months. After it, the revenue share affiliate pulls away and keeps going.
That break-even is straightforward: $150 ÷ $17.50 = 8.57 months. Which gives you a clean decision rule. If you believe your referred traders stay active for longer than nine months, take the revenue share. If you do not, take the CPA. Everything else is refinement.
And be honest about the input. Fourteen months of continuous activity is a good outcome for a crypto trader, not an average one. A large share of funded accounts go quiet inside a quarter. If your realistic average is five months of activity, the same 50% deal pays $87.50 — well under the CPA — and the comparison flips entirely.
Single-trader math tells you which model wins. Cohort math tells you what your business actually looks like month to month, and it is the part that makes new affiliates give up on revenue share three months too early.
Assume you send 10 qualified traders a month, each generating $35/month in fees for 14 months, on the same 50% deal.
Under CPA at $150, your income is flat: $1,500 every month, forever, as long as volume holds.
Under revenue share, each monthly cohort contributes 10 × $17.50 = $175 a month while it lives. Cohorts stack:
| Month | Active cohorts | Monthly revshare income | Cumulative revshare | Cumulative CPA |
|---|---|---|---|---|
| 1 | 1 | $175 | $175 | $1,500 |
| 3 | 3 | $525 | $1,050 | $4,500 |
| 6 | 6 | $1,050 | $3,675 | $9,000 |
| 9 | 9 | $1,575 | $7,875 | $13,500 |
| 12 | 12 | $2,100 | $13,650 | $18,000 |
| 14 | 14 | $2,450 | $18,375 | $21,000 |
| 17 | 14 (steady) | $2,450 | $25,725 | $25,500 |
| 24 | 14 (steady) | $2,450 | $42,875 | $36,000 |
Two crossover points, and conflating them is the classic mistake.
The monthly crossover is month nine. That is when revenue share income per month ($1,575) exceeds the flat CPA income ($1,500). This is the point where your dashboard starts to feel good.
The cumulative crossover is month seventeen. That is when total money received under revenue share ($25,725) passes total money received under CPA ($25,500). Until then, the CPA affiliate has had more cash in hand the entire time — a lot more, at the worst point roughly $5,600 more around month nine.
From month 14 onward the model reaches steady state: cohort 1 rolls off exactly as cohort 15 arrives, so income flattens at $2,450 a month. That is the ceiling for this traffic level, and it is 63% above the CPA line — the same 63% we saw on the single trader, which is what you would expect once the system is full.
If you are funding traffic out of your own pocket, the eight-month cash gap in the middle is the whole problem. An affiliate spending $800 a month on media to produce those 10 traders needs to cover roughly $9,600 of spend in year one against $13,650 of revenue share receipts — survivable. An affiliate spending $1,400 a month does not survive it without capital. That is not a strategy failure, it is a financing failure, and it is why plenty of good media buyers take CPA deals they know are worth less in total.
Two adjustments make revenue share workable when cash is tight. Take a hybrid so something arrives in month one. Or run CPA on your paid channels and revenue share on your organic ones, where there is no cash outlay to recover. Splitting deal types by channel rather than picking one for the whole business is underrated.
Most exchange programs move your rate with volume. Typical shape: 20 or 25% base, stepping to 35% at some monthly threshold of new qualified users or referred fee revenue, and 40–50% above that. Some tiers are permanent once earned; more often they are recalculated monthly, so a slow month drops you a tier.
Things worth checking before you treat a tier as real:
The top tier is also where negotiation actually happens. Published tiers are defaults. If you are consistently landing in the second tier, a bump to the third for the next quarter is a normal thing to ask for, and going through an agency with direct advertiser relationships is usually how it gets granted rather than acknowledged.
Nearly every exchange program has one: recruit other affiliates, earn a percentage of what they generate. Rates commonly sit in the single digits to low teens on one tier, occasionally with a smaller second tier below it.
The important structural point is that sub-affiliate commission is normally paid by the advertiser out of their own margin, not deducted from your sub's earnings. Your sub earns the same either way. That makes it a genuinely non-adversarial arrangement and worth being open about.
Where it goes wrong is expectation-setting. A sub-affiliate layer at 10% means you need ten productive subs to equal one of yourself, and most people you recruit will produce nothing. It compounds beautifully for anyone running a community, a course, or a Telegram channel full of aspiring affiliates. For a solo content operator it is usually a rounding error. Treat it as a bonus on a business you were building anyway, not as a business.
Read the clause. Lifetime is a marketing word, and there are four standard mechanisms that end it early.
Inactive-user reset. If a referred user does not trade for 90 or 180 days, they detach from your account. If they come back later, they come back unattributed — or attributed to whoever gets them next. This one is common and rarely highlighted.
Affiliate dormancy. If you send no new qualified referrals for a stated period, often six months, your existing revenue share can be reduced or terminated. It is a retention mechanism aimed at affiliates who want to stop working and keep collecting.
Unilateral term changes. Nearly every affiliate agreement lets the advertiser change rates with notice, often 14 or 30 days. Lifetime describes the attribution, not the rate attached to it. A program that cuts everyone from 50% to 35% has not breached anything.
Corporate events. Acquisition, restructuring, exit from a market, or a regulatory change that makes your geo ineligible. Existing balances usually get paid; future revenue does not. This is why concentration risk on a single exchange is an operational risk, not just a portfolio preference.
Beyond the net-revenue definition, these are the ones that show up repeatedly and cost real money.
VIP carve-outs. Your best users become the exchange's best users, and above a certain VIP tier they are either excluded from affiliate commission entirely or moved to a reduced rate. The logic from the advertiser's side is that a whale is expensive to service and would have found them anyway. The effect on you is that the single account carrying your cohort stops paying at exactly the point it becomes valuable.
Fee rebate deductions. If the exchange runs a promotional period with zero fees on certain pairs, or issues fee vouchers to your referred users, net revenue for that period is near zero and so is your commission. Promotional generosity to users is frequently funded out of the affiliate pool.
Sub-account and institutional exclusion. Volume routed through API sub-accounts, market-making agreements or institutional desks is often carved out. A referred user who upgrades to an institutional account can disappear from your reporting entirely.
Bonus and incentive clawback. Deposit bonuses, trading competitions and airdrops given to your users may be deducted from your revenue before your percentage is applied. This works similarly to how bonus costs are handled in gambling deals — the negative carryover mechanics familiar from iGaming are the closest analogue, and while true negative carryover is rare on crypto exchanges, deduction of promotional cost is not.
Reporting opacity. If the dashboard shows aggregate revenue with no user-level or sub-ID-level breakdown, you cannot audit any of the above. You cannot tell a dead cohort from a VIP carve-out from a reset. Insist on granular reporting, and pass a sub-ID on every link so your own tracker has an independent record — the S2S postback setup is what makes that independent record possible.
| Traffic type | Typical user behaviour | Better fit | Why |
|---|---|---|---|
| Beginner SEO ("how to buy bitcoin") | Small deposit, low frequency, short life | CPA | Users rarely last past month six |
| Comparison and review content | Mixed, some serious | Hybrid | Splits the difference on unknown quality |
| Trader communities, Telegram, Discord | High volume, longer life | Revenue share | Cohort life easily exceeds break-even |
| Paid search and native | Unknown quality, cash-intensive | CPA or hybrid | Cash recovery funds the next flight |
| YouTube tutorials and strategy | Engaged, moderate volume | Revenue share | Long attribution tail, no media spend |
The general principle from the wider revenue share versus one-time CPA comparison holds here: the more capital you have tied up in acquisition, the more you should value early cash, regardless of which model has the higher theoretical total. Where your users come from drives their expected life more than anything you can do with a landing page, which is covered in the crypto traffic source breakdown.
Comparing percentages without comparing net revenue definitions. Already covered, and still the most expensive error in the vertical.
Abandoning revenue share in month four. Month four of the cohort model shows $700 against $6,000 cumulative CPA. It looks like a disaster. It is the expected path. If you switch models at that point you get the worst of both — no CPA cash and no accumulated tail.
Assuming your user life is the advertiser's quoted average. Ask what the median active life of a referred account is, not the mean. If they will not tell you, assume it is shorter than you hope and price accordingly.
Ignoring the qualification clause when comparing to CPA. A $150 CPA with a $10,000 volume requirement may convert on a third of the users a $90 CPA with a $100 deposit requirement pays out on. The effective per-deposit value is what matters, and calculating it properly is part of the wider affiliate KPI discipline.
Concentrating everything on one exchange. Every mechanism in the "lifetime" section applies at once when a single program goes bad. Spread across at least three, and read the terms on each — the offer-type breakdown in the crypto affiliate programs guide is a reasonable starting map.
We work with crypto affiliates on exactly this decision, and it is usually less about finding a higher number than about reading a contract properly before volume goes to it. Our partners run on a 50% revenue share, with tracking infrastructure and landing pages built and maintained for them at no cost, which removes the setup cost that makes the early months of a revenue share deal painful. Direct relationships with the platforms mean we can push for higher caps and custom payout bumps rather than accepting published tiers, and we can tell you where a program's net revenue definition or VIP carve-out will bite before you find out from a payout report.
You get a dedicated manager who will actually answer, and everything runs through Telegram. Start with the Telegram bot, or see how we work first.
Revenue share wins on total earnings if your referred users trade actively for more than about eight to nine months, and CPA wins if they do not. CPA also wins on cash flow, which matters if you are funding paid traffic. Cold beginner traffic usually suits CPA; experienced trader traffic usually suits revenue share.
Almost always as a percentage of the net trading fees your referred users pay, not their deposit or trade volume. Net means after VIP tier discounts, maker rebates, any fee kickback you have configured for the user, and sometimes after promotional costs. Read the definition of net revenue before comparing two percentage rates.
In practice it means for as long as the user remains attributed to you under the current terms. Common limits include a 90 or 180 day inactivity reset that detaches dormant users, a dormancy clause on your own account if you stop sending referrals, and a right for the advertiser to change rates with short notice.
You earn a percentage of what affiliates you recruit generate, paid by the advertiser out of their own margin rather than deducted from your sub's earnings. Typical rates are single-digit to low-teens percentages on one tier, occasionally with a smaller second tier. It compounds well but only if your subs stay active.
A working breakdown of exchange, wallet, tool and tax offers in crypto — payout mechanics, qualification gates, attribution windows, and choosing programs that hold up when volume dries up.
Read postMost affiliates pick a payout model out of habit. Here is the arithmetic that should decide it: retention thresholds, cohort curves, capital velocity and the risks on both sides.
Read postSeven traffic sources for crypto offers, compared on intent temperature, funnel length and failure mode — plus the cost-per-qualified-deposit math for three of them side by side.
Read post