Affiliate Commission Structures That Reward the Right Behavior

A commission is not an accounting detail. It is a message. Whatever behavior you pay for is the behavior you will get more of, so the structure of your commissions quietly decides what kind of channel you end up with. When I rebuilt Chegg’s affiliate program, changing the commission logic was one of the highest-return moves I made, because the old structure was paying for the wrong thing.

This post is about designing a structure that pulls the channel in the direction you actually want.

The default structure is usually wrong

Most programs start with a single flat rate: every partner earns the same percentage on every order. It is easy to administer and easy to explain, which is exactly why it is so common and so limiting.

The problem is that a flat rate treats two very different partners identically. A content creator who introduces your product to a brand-new audience earns the same as a coupon site that intercepts a customer at checkout who was already going to buy. One expanded your business. The other taxed a sale you had already won. Paying them the same tells both that interception is fine, and you will get more of it.

The first principle of commission design is therefore simple: pay most for the behavior that grows the business, and least for the behavior that merely reprices existing demand.

Reward new customers over repeat

The single most useful distinction you can build into a commission structure is new customer versus returning customer. Acquiring someone new is worth far more to the business than paying a partner for a repeat purchase the customer would likely have made anyway.

A structure that pays a higher rate for a first-time customer and a lower rate for a returning one aligns partner incentives with the business. It rewards partners who genuinely expand your base, and it stops you overpaying for orders that were never really at risk. This one change also tends to shift your partner mix over time toward the creators and communities who bring new audiences, which is exactly the diversification you want. I go deeper on why that mix matters in diversifying partner concentration to de-risk a channel.

Use tiers to grow your best partners

A flat rate gives a partner no reason to do more. Tiered commissions do. When a partner earns a better rate as they drive more volume, you give your best partners a reason to invest in you rather than a competitor.

Tiers work best when they are:

Transparent. Partners should be able to see exactly what they need to do to reach the next tier. Hidden tiers do not motivate anyone.
Achievable but meaningful. The jump should be worth chasing without being impossible.
Reviewed on a sensible cadence. Monthly or quarterly, so a good month is rewarded and a slow one is not punished forever.

Tiers also give you a clean tool for partner development. When you are recruiting and onboarding new partners, a visible path to a better rate is part of the pitch. That connects directly to the system I describe in recruiting affiliate partners.

Match the model to the partner type

Not every partner should be paid the same way, because not every partner does the same job.

Content and review partners who build long-lived pages that keep sending traffic are often worth a standard revenue share, because the value compounds over time.
Influencers frequently need a blend of a flat fee and performance commission, because their upfront production cost is real and pure performance pay can scare off the good ones. The measurement side of this is its own topic, covered in measuring an influencer program beyond orders.
Loyalty and coupon partners deserve the most scrutiny. They can drive real incremental volume, but they can also skim. A lower rate, or a rate that only applies to new customers, keeps them honest.

The point is not to build a baroque matrix. It is to stop pretending that fundamentally different partners should be paid on identical terms.

Protect margin with caps and rules

A commission structure that ignores margin will eventually cost you more than it earns. A few guardrails keep the program healthy:

Category-aware rates. If your margin varies by product, your commission should too. Paying a flat rate across high and low margin products slowly erodes profit.
Exclusions where they make sense. Some products, like heavily discounted or loss-leader items, may not support a commission at all.
Attribution rules that reflect reality. Last-click attribution overpays the partner closest to checkout and underpays the one who created the demand. Even a simple rule that reduces payout when a paid or branded channel touched the order first will stop you paying twice for the same customer.

These rules are not about squeezing partners. They are about making sure the channel stays your lowest-CPA line rather than drifting into an expensive habit. That efficiency was central to the whole turnaround, which I tell in full in how to turn a declining affiliate program into your lowest-CPA channel.

Communicate changes like a product launch

Here is the part teams underrate: changing commissions is a trust event. Partners have built their businesses on your rates, and a clumsy change will cost you your best people.

When you change a structure:

Give notice. Sudden cuts read as betrayal. A clear runway reads as respect.
Explain the why. Partners understand a shift toward new-customer rewards far better when you tell them the reasoning.
Protect your top partners. Where you can, grandfather or cushion the people who drive real value, so a structural change does not read as a punishment for loyalty.

Treat a commission change the way you would treat a pricing change for customers: carefully, transparently, and with the relationship in mind.

A worked commission model

Principles are easier to apply against a concrete model, so here is one I would put in front of a program that wants to reward growth without bleeding margin.

Start with a base revenue share that is sustainable at your average margin. Then layer three modifiers on top. First, a new-customer premium: a meaningfully higher rate when the order comes from a first-time customer, because that is the behavior that grows the business. Second, a returning-customer discount: a lower rate on repeat purchases, because you are largely paying for a sale that was already likely. Third, a volume tier: partners who cross a monthly threshold move up a band, giving your best partners a reason to invest more in you.

Run a partner through it. A content creator who consistently brings new customers earns the base plus the new-customer premium plus, once they scale, the volume tier. They are richly rewarded, exactly as you want, because they expand your base. A coupon site that mostly intercepts returning customers at checkout earns the base minus the returning-customer discount. They still participate, but they are not overpaid for demand you already had. The structure has quietly tilted the whole channel toward the partners who grow it.

The elegance is that you did not have to police behavior manually. The incentives did the sorting for you, which is the entire point of designing a structure rather than negotiating every partner one at a time.

Negotiating and rolling out rate changes

The best-designed structure still has to survive contact with real partners, and the rollout is where programs win or lose trust. A few things make rate changes land.

Give real notice. A change announced weeks ahead reads as respect; a change that hits without warning reads as betrayal, and your strongest partners have the most options when they feel betrayed. Explain the reasoning in terms the partner cares about: a shift toward rewarding new customers is far more palatable when you show that top partners who bring new audiences will actually earn more, not less. Where a change would hurt a valued partner, cushion it, grandfather them for a period, or pair the change with a path to a better tier so it reads as opportunity rather than punishment.

Expect pushback from the partners a change is designed to affect, and hold the line where the structure is right. A coupon partner unhappy about a lower returning-customer rate is the structure working as intended, not a problem to solve by caving. But listen genuinely, because sometimes the pushback reveals a real flaw, such as an attribution rule that is misfiring. The goal is a structure that is fair and defensible, communicated like the relationship matters, because in an affiliate channel the relationship is the asset.

Keep it simple enough to explain

A final principle that keeps a structure healthy: a partner should be able to understand how they get paid in about a minute. It is easy to design a clever structure with so many modifiers and exceptions that no partner can predict their earnings, and unpredictable earnings suppress the behavior you are trying to encourage. If a partner cannot see the connection between doing more of the right thing and earning more, the incentive does not work no matter how elegant it looks on your side.

So favor the smallest structure that expresses your priorities. New-customer premium, a returning-customer discount, and a volume tier is usually enough to tilt the whole channel toward growth without becoming inscrutable. Every additional rule should earn its place by changing behavior in a way you can name. Complexity is not sophistication here; a structure partners can reason about will always outperform one they cannot, because incentives only work when the people they target can see them.

How commissions interact with your other channels

A commission structure does not live in isolation. It sits inside a marketing mix, and if you ignore that, you end up paying twice for the same customer. The classic case is a customer who sees a paid ad, searches your brand, and then clicks an affiliate coupon link at the last moment. Last-click attribution hands the affiliate full credit for a sale your paid and brand channels already earned.

The fix is attribution rules that reflect reality. A simple and effective one: reduce or remove the affiliate commission when a paid or branded channel demonstrably touched the order first. This is not about shortchanging partners; it is about not paying a partner to stand at the finish line of a race someone else ran. Partners who genuinely create demand are unaffected, because there was no prior touch to credit. Partners who only intercept see their real contribution reflected.

Brand bidding is the sharpest version of this. A partner who bids on your own brand terms in paid search is often intercepting customers who were already searching for you, and charging you for the privilege. A clear policy against brand bidding, enforced with monitoring, closes that leak. The same monitoring discipline catches the fraud patterns covered in spotting and stopping affiliate fraud.

The broader point is that commission design and channel strategy are the same conversation. A structure that looks generous in isolation can quietly cannibalize your other channels, and a structure designed with the whole mix in mind keeps affiliate as the efficient, incremental line it should be. That efficiency, the lowest CPA of any channel, was the entire goal of the affiliate turnaround, and it is impossible without attribution that tells the truth about who actually earned each order.

The short version

A commission is a message; pay for growth, not interception.
Reward new customers over repeat purchases.
Use transparent tiers to grow your best partners.
Match the payment model to the partner type.
Protect margin with category-aware rates, exclusions, and honest attribution.
Roll out changes like a launch, with notice and a clear reason.

Get the structure right and the rest of the program becomes easier, because every partner is now pulling in the direction the business actually needs.

I am Deepanshu Grover, a Growth Product Manager in Paris. I rebuilt Chegg’s affiliate commissions and payouts on the way to 150% order growth. If you are redesigning a commission structure, connect on LinkedIn or get in touch.