Neither payout model pays more on its own. Cost per action pays a fixed amount once, when a referred user completes a qualifying action. RevShare pays an ongoing percentage of the revenue that the same customer generates afterwards. Which structure puts more money into an affiliate account depends on how long customers stay, how the traffic is bought, and who absorbs the loss when a sale reverses. That makes the choice a marketing decision as much as a financial one. In affiliate marketing, the same offer can look strong under one model and weak under the other.
That is why the usual comparison asks the wrong question. A payout model is not a price list. It is a contract about risk, and the risk sits in a different place under each structure. This article defines both models, compares them on the dimensions that matter, and sets out a procedure for choosing between them. An affiliate who understands it can negotiate terms that match the traffic. Readers who want the wider picture first can start with the basics of the affiliate model.
What CPA means in affiliate marketing
CPA stands for cost per action, and it is also called cost per acquisition. An advertiser pays a fixed amount each time a referred user completes a defined action, and that amount does not change afterwards. The action is agreed in advance and written into the offer terms. That fixed commission structure is the main appeal for media buyers who need a predictable marketing cost.
Almost anything measurable can serve as the trigger. Cost per sale pays on a confirmed purchase. Cost per lead pays when someone submits contact details. Cash on delivery pays after a courier confirms that the order arrived and was paid. Cost per install pays when a mobile app lands on a device. Registration, a verified account, an approved application and an activated trial are common triggers too. Each trigger uses a different signal, so the affiliate must check which one the program counts.
What unites these variants is the fixed amount. If a hypothetical sixty-dollar CPA is agreed, the affiliate receives sixty dollars whether the customer buys one item or three, whether the basket is small or large, and whether that customer returns next year. Upsells, repeat orders and subscription renewals add nothing to the payout. The commission does not change when the order value changes, which keeps the arithmetic simple.
That predictability is the appeal. A media buyer can work out the margin before spending anything: known payout minus known cost per conversion. If the arithmetic holds, the campaign scales. If it does not, the traffic gets cut. Nothing that happens after the conversion changes the result.
What RevShare means in affiliate marketing
RevShare, short for revenue share, replaces the fixed fee with a percentage. The affiliate takes a share of the revenue that a referred customer generates, normally paid monthly and normally for as long as that customer keeps paying. The percentage applies to a revenue base defined in the contract, which is rarely the same as gross sales. That percentage turns the affiliate into a long-term marketing partner rather than a paid source of conversions.
A percentage deal only produces meaningful sums when customers come back. A subscription billed monthly, a software plan renewed every year, a marketplace that takes a fee on repeat orders: these are the natural homes of RevShare. Each renewal adds to the affiliate total from a single referral, so the payout curve keeps rising long after the traffic has stopped running.
The trade-off is timing. Income arrives slowly at the start and often looks disappointing in the first months. It also depends on the advertiser's ability to keep users, which the affiliate cannot control. Both features point the same way: RevShare suits publishers who can afford to wait. A marketing budget that needs same-month payback will struggle with that shape.
CPA vs RevShare: the dimensions that differ
Side by side, the two commission models diverge on more than the moment of payment. The table below puts the main criteria in rows. Each row shows a different marketing trade-off, not just a different payment schedule.
| Criterion | CPA | RevShare |
|---|---|---|
| Payment trigger | One defined action | Revenue from the referred customer |
| Timing | On validation of the action | Monthly, as revenue is recognised |
| Payout shape | Fixed amount per action | Percentage of a revenue base |
| Predictability for the affiliate | High | Low at first, higher once cohorts mature |
| Cash flow | Front-loaded | Back-loaded |
| Budgeting for the advertiser | Easy to forecast | Harder, because the total depends on retention |
| Quality incentive | Weak after the action | Strong, tied to customer value |
| Risk of refunds and churn | Sits with the advertiser | Sits with the affiliate |
| Reporting burden | Low | High, needs transparent revenue figures |
| Natural fit | Short, trackable conversions | Recurring or high-lifetime products |
A hypothetical example, not market data, shows how the timing gap works. Suppose 100 customers each pay $20 a month and the RevShare is 20%, while the CPA on the same offer is $40. On CPA the affiliate collects $4,000 in month one and nothing afterwards. On RevShare the affiliate collects $400 a month for as long as those customers stay, so the two curves cross somewhere in month ten. Change the retention assumption and the crossing point moves. That is the whole argument in one line. The example uses round numbers only to show the arithmetic, not a promised return.
Who carries the risk after the conversion
The root difference is not the rate. It is what happens when the customer behaves badly. That is why the comparison is a risk question before it is a commission question.
Under CPA, the advertiser pays on the action and then owns everything downstream. If the buyer refunds, cancels or never returns, the advertiser has already paid the affiliate and absorbs the loss. That is why a fixed fee shifts risk to the advertiser, and why a merchant with weak retention may still prefer to buy on that basis: the budget line stays knowable. The affiliate gets paid once and moves on, while the marketing cost is already booked.
Under RevShare, the affiliate funds the acquisition effort and is paid only as revenue appears. If the referred users churn in month two, the affiliate carries that. In exchange, the affiliate keeps the upside when those users stay for years. The model rewards retention, not the click.
Cost per lead sits between the two. The advertiser pays for a qualified lead before any sale closes, so it holds the conversion risk and has to police lead quality. That structure is reasonable when the sales cycle is long and the affiliate cannot influence the close. The affiliate is paid quickly; the advertiser takes the harder part of the funnel. This commission model suits long sales cycles where the marketing team cannot verify a sale quickly.
Hybrid and tiered structures
Because each model carries a different kind of risk, many programs stopped choosing between them. A hybrid deal combines a smaller fixed payment with a share of future revenue: a reduced CPA or CPL on the action, plus a percentage of what the customer spends afterwards. The fixed part covers ad spend quickly. The percentage keeps the affiliate interested in quality. The hybrid commission plan gives the affiliate a known floor and a share of the long tail.
Tiered structures work on another axis. The rate rises as the affiliate delivers more volume, better quality or a larger share of the program total. Tiers reward the partners who matter most and pull supply towards the program. A program that pays per lead can tier payouts by how much of the overall lead portfolio a partner supplies, which keeps the biggest suppliers committed.
Both structures solve the same tension. Media buyers want a fixed number they can price into a campaign. Content publishers want a long tail that keeps paying. A program that offers only one of the two will lose one of the two groups. Whether the merchant runs its own program or works through a network changes who sets those terms, and white label versus affiliate program is the distinction that decides it.
When CPA is the right choice
CPA fits when the conversion is short, trackable and close to the click. Paid search, paid social and display campaigns need payback inside the billing cycle, and a fixed payout is the only shape that allows that calculation. If the cost of a click and the conversion rate are known, the affiliate knows the margin before the campaign goes live.
Three more conditions point the same way. A high refund rate punishes revenue share, because returns erase revenue that was already counted. A product bought once and never again gives a percentage deal nothing to work with. And a publisher with no retention data cannot judge a RevShare offer at all, so a fixed fee is the safer entry point. The affiliate can use the fixed fee to learn which traffic sources convert before taking on revenue risk.
When RevShare is the right choice
RevShare fits the opposite profile. The product is bought repeatedly or billed on a subscription, retention past the first few months is realistic, and the advertiser reports revenue in a form the affiliate can check. Under those conditions, the same referred audience can produce far more on a percentage deal than on any single fixed fee.
Organic traffic is the other half of the argument. Search, email and content channels keep producing visitors from work that was paid for once, so a slow-building revenue stream matches the cost structure. Paid traffic has the reverse profile. The ad bill arrives immediately, which makes a model that pays in month eight hard to finance. A marketing strategy built on search and content can wait for that slower return.
RevShare against the other payout models
CPA and RevShare are not the only two options, and the neighbouring models cover different stages of the funnel. Cost per click pays for the click itself, the weakest signal of intent and the easiest to fake. Cost per install pays for an app install and says nothing about whether the user ever pays. Cost per lead pays for contact details before a sale exists. Cost per sale pays for the completed purchase. Each commission model shifts the risk to a different point in the customer journey.
Each one moves the risk further along the funnel. CPC and CPI leave almost all of the risk with the advertiser, because payment fires long before any revenue appears. CPL moves part of that risk back to the affiliate through the quality bar. CPA lands on a completed action. RevShare is the only structure tied to what the customer is worth over time, which is why it behaves differently from the rest.
Networks and platforms differ in which structures they support, so checking which networks offer each payout model belongs at the shortlist stage rather than after signing. The platform terms can also decide whether a hybrid or tiered model is available at all.
How the traffic source shapes the choice
Traffic decides the model more often than preference does. Paid sources need a fast return. Organic sources can wait. Influencer placements sit in between.
| Traffic source | Recommended structure | Reason |
|---|---|---|
| Paid search and paid social | CPA or hybrid | Ad spend needs payback inside the campaign cycle |
| SEO and editorial content | RevShare or tiered RevShare | Older pages keep producing visitors at no extra cost |
| Email and push lists | RevShare or tiered | Known audience with measurable repeat activity |
| Influencer and streamer placements | Hybrid | A burst of volume plus a long tail of followers |
| Mixed portfolios | Hybrid | Covers both the short and the long horizon |
| No data yet | CPA, CPL or hybrid | A fixed fee is easier to evaluate without history |
A review site or blog can build an audience once and keep earning from it, which is exactly the profile that suits a percentage deal. A paid campaign has to drive traffic today and recover the cost this month, so it needs a fixed number. That is the simplest version of the whole decision.
Vertical by vertical
The product decides how long a customer stays, and that decides the model. The vertical also shapes the marketing data that is available.
Subscriptions and software are the clearest case for RevShare. A plan billed every month keeps paying for as long as the user is active, so a share of that stream compounds while the affiliate moves on to new campaigns. Recurring-revenue business software behaves the same way, and online courses sold as memberships follow the same pattern. A software platform with recurring billing can report revenue events that the affiliate can check.
Mobile apps split by monetisation. Where the money comes from a subscription, RevShare or a hybrid fits. Where the money comes from in-app purchases, a CPA on the first purchase or a share of all purchases gives the affiliate a direct stake in sending valuable users. Where the app is free and monetised by advertising, paying for installs alone is the weakest option, because an install says nothing about revenue. The app platform may also set rules on tracking and attribution that limit what the affiliate can measure.
E-commerce and physical goods lean towards CPA. A single purchase with no repeat order gives a percentage deal nothing to compound. A brand that sells repeat consumables can justify a share of order value instead, and a hybrid covers both sides. Travel and hosting sit in the middle: a one-off booking behaves like retail, an annual hosting plan behaves like a subscription. The question to ask is always the same. Does this customer pay again without new traffic? That question connects the product, the marketing channel and the payout model in one line.
Gross revenue, net revenue and negative carryover
A percentage is only as good as the number it is applied to. Many programs calculate RevShare on a net figure rather than a gross one, and the difference is larger than most affiliates expect. The commission basis belongs in writing, because the affiliate cannot verify a number that is never defined.
Net revenue is gross revenue minus bonuses, chargebacks, processing fees and taxes. A formula of that kind is normal and legitimate. The problem is that affiliates often assume the percentage applies to the gross figure, and the gap between the two can be substantial. The deduction list belongs in the contract, not in a conversation. A marketing report that shows only gross sales can hide the deductions that reduce the payout.
Negative carryover is the second trap. If the revenue base for a month comes out negative, some programs roll that deficit into the next month, and the affiliate earns nothing until the balance is recovered. Programs that reset the balance every month remove that risk. A no-negative-carryover clause is worth negotiating before signing, and a written clause that zeroes the balance at the start of each period achieves the same thing. The affiliate should use the contract to check whether a losing month is carried forward.
Transparency and the terms worth checking
RevShare depends on numbers the affiliate cannot see directly, so the reporting relationship matters as much as the percentage. A program that will not show how a payout was calculated is asking for trust it has not earned. The platform or network should provide a report that the affiliate can reconcile.
The contract should settle several points in writing: the revenue base and every deduction applied to it, the carryover policy, the cookie or attribution window, the minimum payout threshold, the payment methods and schedule, the geographies excluded from the offer, whether sub-affiliates are allowed and on what terms, whether the model can be switched mid-campaign, and which traffic sources are permitted. Compliance clauses belong on the list too, because regulated products carry restrictions on what can be advertised and to whom. In the United States, the FTC expects publishers to disclose a commission relationship clearly and close to the recommendation, so the program terms and the publisher's own pages both have to allow for that.
Reconciling what was paid is the other half of the job. Tracking what you are actually paid for is a reporting problem before it becomes an accounting problem. The affiliate needs to see clicks, conversions and revenue events in one place, otherwise a dispute turns into one party's word against another's. That is why the tracker and the commission report must use the same event definitions.
The numbers to run before choosing
The decision should come out of arithmetic, not taste. On the affiliate side, four figures decide it: the cost of producing one sale, the average payout under each structure, the refund rate on the offer, and whether the product bills again. If it bills again, a percentage deal usually overtakes a fixed fee at some point on the curve.
On the advertiser side, the same logic runs in reverse: the average order value including upsells, the CPA that keeps the unit economics positive at that order value, the refund rate, and the type of partner the program needs to attract. Media buyers respond to fixed fees. Content publishers respond to recurring percentages.
Two further variables override the vertical defaults. Margin sets the ceiling on what can be paid, because thin-margin products cannot support a generous fixed fee and lean instead on cost per lead or on a percentage that scales with revenue. Data maturity sets the floor. A program that cannot track and verify a downstream event should start with leads, move to a completed action once conversion tracking is reliable, and only then offer RevShare. Medium maturity covers a verified action. High maturity covers revenue reported through the customer lifetime, and that is where percentage deals open up.
Getting more out of a RevShare deal
Once the model is chosen, the work shifts from volume to durability.
Pick offers that retain. Before committing, look at refund rates and average subscription length, which say more about the eventual payout than the headline percentage does. Build funnels that keep the user, using onboarding sequences, email, retargeting and push to reduce churn. Focus on the metrics that predict the outcome rather than the ones that flatter the report: earnings per click, churn rate and lifetime value by cohort. Test geographies and audience segments separately, because retention varies far more between them than conversion rate does. A marketing channel that brings low-intent visitors will show weak retention even if the initial conversion rate looks good.
The mental shift is simple. A fixed fee rewards the number of conversions. A percentage rewards the number of conversions that stay. Publishers who treat every referral as a long-term customer tend to do better than those chasing one-off sales. The affiliate who uses that lens will judge an offer by repeat revenue, not by the first payment alone.
Mistakes that cost money
The most expensive mistake is chasing the highest advertised percentage without checking the product behind it. A large share of a shrinking revenue base is worth less than a smaller share of a healthy one. A commission rate means little if the underlying product cannot retain customers.
Ignoring the carryover policy is the second. A deal that looks generous on paper can pay nothing for months if a deficit rolls forward. Choosing the model before looking at the traffic source is the third, because paid campaigns financed by a slow revenue stream run out of money before the curve turns. Depending on a single program or a single channel is the fourth, since a rate change or an algorithm update then removes the whole income at once. Weak conversion work and a careless contract round out the list.
A first-model path for beginners
With no data, a fixed structure is easier to evaluate. CPA or a hybrid gives a beginner a known amount per conversion, which makes it possible to tell whether the traffic is profitable at all. Cost per lead is the simplest event to verify when tracking is thin. That fixed commission model also limits the loss while the affiliate learns the traffic.
The path from there is a sequence. Collect conversion and refund data on a fixed fee first. Watch how many of those customers are still active after a few months. Once retention is measurable, test a percentage deal on the offers where the numbers support it. Affiliates with more experience and spare cash can start on RevShare. Newcomers usually cannot afford the wait.
Negotiating terms: five points of leverage
Once there is a track record, terms become negotiable.
Volume and earnings per click are the first lever, because proven traffic that converts is worth a better rate. Customer quality is the second, and retention data is the proof. Exclusivity in a specific market is the third, since it is valuable to the advertiser. Contract length is the fourth, because a longer commitment can justify a better percentage. Competitive benchmarks are the fifth, and quoting a rival's published terms is a normal part of the conversation.
Whatever is agreed should be written down. The revenue formula, the carryover rule, the attribution window, the payout threshold and the sub-affiliate terms are the clauses that decide what actually gets paid.
What different programs do in practice
Public programs show how the choice plays out, because their products and channels differ. Amazon Associates, for example, keeps what it pays in a separate commission schedule that its operating agreement refers to, so what an affiliate earns depends on what is sold. Programs for subscription products tend to pay a recurring percentage, which attracts creators who keep promoting after the first sale. Programs for lead-driven services tend to pay per lead and tier payouts by how much of the lead volume a partner supplies. The pattern holds: programs match the payout shape to how long their customers pay.
Where the model is heading
Three shifts are visible. They touch tracking, marketing analytics and the way partners are paid.
Better analytics and AI-driven prediction are making lifetime value estimateable before a campaign launches, which makes percentage deals easier to price. Hybrid structures keep spreading, because they let a program attract both performance marketers and content publishers without splitting the offer in two. Privacy rules are pushing tracking towards consent-based models, which favours publishers who have a direct relationship with their audience rather than those relying on third-party signals.
Compliance shapes the model in a second way. Where advertising rules restrict what can be said about a product, both sides need a structure that survives scrutiny, and a percentage tied to a verifiable revenue base is easier to defend than a rate nobody can explain.
A decision checklist
Work through these questions in order. The first answer that fits usually settles the model.
- Does the customer pay more than once? If not, lean towards CPA.
- Is retention past the first three months realistic? If yes, RevShare becomes viable.
- Is the traffic paid? If yes, CPA or a hybrid keeps the campaign financeable.
- Is the traffic organic, email or content? If yes, RevShare or a tiered structure fits better.
- Is the revenue base net of deductions, and is the deduction list published? If not, negotiate before signing.
- Does the contract allow a model switch, and under what conditions?
- Can the affiliate's own tracking reconcile the revenue events?
- Is the refund rate measurable yet? If not, start with a fixed fee.
Choosing between two ways of getting paid
The question in the title has no single answer, and that is the useful part. CPA sells certainty: a known amount for a known action, with the advertiser holding the risk afterwards. RevShare sells duration: a smaller share of something that keeps paying, with the affiliate holding the risk. A product with repeat customers and honest reporting rewards the second. A campaign financed by ad spend usually needs the first. The right commission model is the one that matches the product, the traffic and the data the affiliate can actually verify. The wider context for these choices sits in the affiliate marketing overview.
FAQ
Which model pays more, CPA or RevShare?
Neither is ahead in general. CPA pays a known amount immediately and caps the upside at that amount. RevShare pays a percentage over time and can exceed a fixed fee when customers stay, but it pays less in the early months and puts churn risk on the affiliate. Retention, traffic type and available cash decide the outcome.
Who carries the risk in each model?
Under CPA the advertiser pays on the action and carries the risk of refunds, churn and weak customer quality afterwards. Under RevShare the affiliate is paid only as revenue appears, so the affiliate carries the risk that users leave early. Cost per lead pushes even more risk onto the advertiser, because payment happens before any sale closes.
Can one program offer both models at once?
Yes, and many do. A hybrid deal pairs a reduced fixed payment with a percentage of future revenue, while tiered deals raise the rate as a partner delivers more volume or better quality. A program that offers a fixed option and a percentage option can serve media buyers and content publishers at the same time.
When should an affiliate switch from CPA to RevShare?
Once retention data exists. A fixed fee is the safer starting point while refund rates and repeat behaviour are unknown. After a few months of data showing that customers stay and pay again, the arithmetic on a percentage deal becomes checkable, and the switch turns into a calculation rather than a guess.