Affiliate Program Economics: The Commission Structure Decision
By Diosh — Founder, AHAeCommerce | eCommerce decision intelligence for $50K–$5M GMV operators
The pitch for an affiliate program is that you only pay when you make a sale. That framing is wrong, and the gap between the pitch and the actual cost stack is where most brands lose money on affiliate programs without realizing it. Commission is roughly 60% of the all-in cost of running a program at $1M GMV. The other 40% — platform fees, fraud, management labor, brand-dilution costs, and tax compliance overhead — never appears in the affiliate dashboard the operator uses to evaluate the program. Brands routinely report "20% commission, ROAS positive" while running an effective 32–38% commission program when the hidden costs are added back.
This isn't an argument against affiliate programs. It's an argument that affiliate is a structural decision about your brand and your channel mix, not a performance-marketing decision you can make based on tracked conversions. Get the structural call wrong and you've built a margin-eating overlay on your existing acquisition that takes 12–18 months to unwind. Get it right and you've added a high-leverage channel that fits the gaps your paid acquisition can't reach.
The Default Assumption (and Why It Fails)
The default operator framing for affiliate programs is that they are pay-for-performance — the brand commits no spend until a sale closes, the affiliate carries the marketing risk, and the brand wins on every transaction. This framing is the basis for most affiliate platform marketing copy and most operator decision-making. Statista's 2024 affiliate marketing report places median program commission rates at 8–15% across DTC categories, with high-end fashion and supplements running 15–25%, and operators evaluate affiliate decisions almost entirely on the visible commission percentage.
The framing fails because affiliate is not a single cost — it's a stack of costs, and the visible one is the smallest of the operationally important ones.
The platform fee is the first hidden cost. Networks like Impact, Refersion, ShareASale, and AvantLink charge transaction fees ranging from 1–3% of attributed revenue, plus monthly platform fees of $200–$2,000 depending on tier. For a brand running $40,000 of monthly affiliate revenue at a 12% commission rate, the visible cost is $4,800. The actual platform-and-commission cost, with a 2.5% network fee and a $500/month platform fee, is $6,300 — 31% higher than the dashboard reports.
The second hidden cost is fraud. The Federal Trade Commission's 2024 affiliate marketing report estimated that 8–14% of attributed affiliate conversions in unmanaged programs are fraudulent — coupon abuse, click stuffing, cookie stuffing, brand-bidding violations, or fake influencer accounts. A program that does not actively police fraud pays commission on transactions that would have happened anyway (cookie stuffing on existing high-intent buyers) or on transactions that don't represent real value (return-fraud rings). Brands that skip fraud monitoring overpay by 6–12% of total commission spend on average.
The third hidden cost is management labor. A program with 20 active affiliates can be managed in 2–4 hours per week. A program with 200 affiliates needs 15–25 hours per week of active management — recruiting, onboarding, content review, dispute handling, payout processing, fraud auditing. At a fully-loaded cost of $35–$60 per hour for an affiliate manager, that's $2,100–$6,000 per month of operating expense not visible in the commission line.
The fourth hidden cost is brand dilution and channel cannibalization, which is the structural cost most operators never measure.
What the Decision Actually Hinges On
The True All-In Cost Per Affiliate Order
The first decision input is the all-in commission rate, which is the visible commission plus platform fees plus fraud loss plus management overhead expressed as a percentage of attributed revenue. For a typical mid-stage DTC brand running a managed affiliate program, this calculation generally produces a 28–38% all-in rate when stated commission is 12–18%. McKinsey's 2024 retail incentive analysis found that brands operating affiliate programs without active fraud and quality management saw 32–45% deviation between stated and effective commission rates within 18 months of launch.
The implication is direct. A brand evaluating affiliate against paid acquisition with a 22% blended CAC-to-revenue ratio should compare it against an affiliate program at its actual all-in rate, not its stated commission rate. If the all-in is 32% and the gross margin is 42%, the affiliate channel produces 10% contribution margin per order versus paid acquisition's 20%. The dashboard suggests parity. The real comparison shows affiliate is half as profitable per order.
The Brand Cannibalization Rate
The second decision input is the cannibalization rate — the percentage of affiliate-attributed orders that would have happened anyway through other channels. The standard tracking model gives the affiliate the credit when their tracking link was the last touch before purchase, but the customer often had brand awareness, was on the email list, or was already in the consideration set before the affiliate touchpoint. Forrester's 2024 attribution research estimated that 20–35% of affiliate-attributed sales in DTC programs are cannibalized from other channels, meaning the brand pays commission on revenue it would have captured at zero marginal cost.
The cannibalization risk is highest for coupon-based and review-site affiliates, where the customer was already searching for the brand by name. It is lowest for content-driven affiliates whose audience would not have encountered the brand otherwise. Brands that don't segment their affiliate base by audience-overlap profile cannot identify which affiliates produce incremental revenue and which produce attribution-shifted revenue. They pay the same commission for both.
The Brand-Dilution Risk From Discount-Coupon Affiliates
The third decision input is the brand-positioning impact of discount-coupon affiliates. Coupon sites and deal aggregators (RetailMeNot, Slickdeals, Honey, Capital One Shopping) account for 30–45% of affiliate-attributed traffic in many DTC programs and convert at 2–4x higher rates than content affiliates. They also train customers to check for a coupon before completing checkout, which structurally lowers the brand's average order value and conditions price sensitivity into a previously price-insensitive customer base.
A direct measurement: Klaviyo's 2024 retention benchmarks show that customers acquired through coupon-affiliate channels have a 22–34% lower 12-month repeat purchase rate compared to organic and paid social acquisitions, because they were trained to wait for a discount. The cost of this isn't visible in the affiliate program P&L. It shows up in declining LTV over the following 12–24 months, which then forces continued discount-coupon dependence to maintain volume — a margin compression spiral that's hard to reverse.
The Cost Reality
The following table builds the all-in affiliate cost stack for three program scales. Assumes a brand with $80 AOV, 42% gross margin, and a stated affiliate commission rate of 15%.
| Program Scale | Monthly Affiliate Revenue | Stated Commission | Platform Fees (2.5% + $500/mo) | Fraud Loss (10%) | Management Labor | Cannibalization Loss (25%) | All-In Cost | Effective Commission Rate | |---|---|---|---|---|---|---|---|---| | Small (20 affiliates) | $10,000 | $1,500 | $750 | $150 | $1,200 | $2,500 | $6,100 | 61% | | Medium (75 affiliates) | $40,000 | $6,000 | $1,500 | $600 | $4,500 | $10,000 | $22,600 | 57% | | Large (200+ affiliates) | $120,000 | $18,000 | $3,500 | $1,800 | $10,000 | $30,000 | $63,300 | 53% |
The pattern is stark. The small program is the worst, not the best, because management labor is fixed-cost-heavy at low scale. The 61% effective commission rate at the small tier means a brand at 42% gross margin loses $19 on every $80 affiliate-attributed order on a fully-burdened basis. The medium and large programs improve only because fixed labor costs amortize across more revenue.
If you exclude cannibalization (treating those orders as net new), the effective commission rates drop to 36%, 33%, and 28% respectively — still substantially above the stated 15%, and only marginally profitable for a brand with sub-45% gross margin.
The threshold question for any brand is whether the marginal customer acquired through affiliate is genuinely incremental. If 25% of attributed orders would have come through anyway, the program needs to be evaluated as if commission is 33% of net-new-customer revenue, not 15% of attributed revenue.
The Trade-Off Map
Run a Wide Program (200+ Affiliates, Open Recruitment)
The benefit of a wide program is reach: more affiliates produce more attributed revenue and lower per-affiliate management labor as a percentage of revenue. The cost is fraud risk and brand dilution. Open recruitment programs typically attract more coupon and deal-site affiliates than content-driven affiliates, which structurally tilts the program toward cannibalization-heavy revenue. Forrester's 2024 affiliate research found that programs above 200 affiliates without active curation saw effective commission rates 8–14 percentage points higher than curated programs of equivalent revenue, primarily due to fraud and cannibalization concentration.
This trade-off is favorable when the brand has commodity-feeling products, broad demographic appeal, and gross margins above 55%. It is unfavorable for brands with premium positioning, narrow audience focus, or margins below 45%.
Run a Curated Program (20–50 Affiliates, Vetted Onboarding)
The benefit of a curated program is signal quality. Vetted content affiliates produce mostly net-new customers, exhibit lower fraud rates, and align with brand positioning. The cost is volume — curated programs typically produce 30–60% of the attributed revenue of an equivalent-effort wide program at the same brand stage, and management labor per dollar of revenue is higher at low affiliate counts. McKinsey's 2024 retail incentive data identifies curated affiliate programs as the highest contribution-margin variant for brands at $1M–$5M GMV with premium positioning, with effective commission rates running 18–24% (vs 28–35% for wide programs), but with revenue typically 40–55% of what the wide program produces.
The trade-off favors brands that already have meaningful organic and paid acquisition and view affiliate as supplementary. It does not favor brands looking to use affiliate as a primary acquisition driver.
Run a Referral Program Instead of an Affiliate Program
The third option is to skip traditional affiliate entirely and run a customer-referral program — existing customers earn discounts or store credit for referring new customers. The benefit is structurally lower fraud (existing customers have low motivation to coupon-stuff their own accounts) and stronger fit signal (referred customers exhibit 12–18% higher 12-month LTV than affiliate-acquired ones, per Klaviyo's 2024 retention data). The cost is volume ceiling — referral programs scale with customer base size, not with marketing reach, so they produce 5–15% of the volume of a comparable affiliate program for a similar-sized brand. Referral programs are the right primary choice for brands at $50K–$500K GMV with strong product-market fit and weak organic reach. Affiliate is the right primary choice for brands at $1M+ GMV with brand awareness sufficient to attract content-affiliate interest. The error is treating these as substitutes when their use cases differ structurally.
When to Run a Program (Specific Triggers)
Trigger 1: Gross Margin Above 50% After Variable Costs
If your gross margin (post fulfillment and packaging) is below 50%, the all-in affiliate cost stack will compress contribution margin to single digits or negative on attributed orders. The math is simple: at 45% gross margin and a 32% all-in commission rate (typical for medium programs), contribution margin is 13% before any fixed cost allocation. That's not enough to justify the operational complexity. Unless your gross margin clears 50%, run a referral program instead.
Trigger 2: You Have a Curation Capability or Budget for One
Affiliate programs without active curation degrade into fraud-and-coupon programs within 12–18 months because the highest-volume affiliate types are coupon-aggregators and review-site farms. If you don't have time, headcount, or agency budget to actively recruit and vet content affiliates, the program will drift toward the wide-program economics with their attendant cannibalization. Budget the curation cost — typically $2,000–$5,000/month at the mid-size tier — into the launch decision, not after.
Trigger 3: Your Customer LTV Tolerates the Discount-Trained Behavior
If your business model requires repeat purchases at full price (subscription replenishment, premium positioning, lifestyle brand), do not launch a program with discount-coupon affiliates. The conditioning effect on repeat purchase rate compounds for 24+ months and is structurally hard to reverse. Statista's 2024 retail eCommerce data shows that brands shutting down coupon-heavy affiliate programs experience a 12–18 month period of compressed AOV before customer behavior normalizes back to pre-program patterns.
What Operators Get Wrong Most Often
Mistake 1: Comparing Stated Commission to Paid Acquisition CAC Ratio
The most common analytical error is comparing the stated affiliate commission percentage to the brand's paid-acquisition CAC-to-revenue ratio and concluding that affiliate is "more efficient." A 15% stated commission compared to a 22% blended paid CAC ratio looks like a 7-point efficiency gain. The corrected comparison — 32% all-in affiliate cost compared to 22% paid CAC — flips the conclusion. Operators who never construct the all-in calculation routinely scale affiliate programs that are net-margin-destructive while reporting strong "ROAS" on the affiliate dashboard.
Mistake 2: Failing to Distinguish Incremental Affiliates From Attribution-Shifters
The second mistake is treating all affiliate-attributed revenue as equivalent. A content affiliate whose audience would not have heard of your brand otherwise produces incremental revenue. A coupon site that intercepts customers searching for your brand name produces shifted attribution at zero marginal customer acquisition. Both look identical in the affiliate platform dashboard. Brands that don't segment affiliates by audience-overlap profile and incrementality testing pay full commission rates on attribution-shifted revenue, which is one of the highest-impact and lowest-effort optimizations available — typically reducing effective commission cost by 15–25% with no revenue impact.
The Verdict
Affiliate programs make structural sense for DTC brands at $1M+ GMV with gross margins above 50%, premium-or-lifestyle positioning that benefits from content-driven discovery, and the operational capacity to actively curate the affiliate base. They make a margin-destructive structural sense for brands below those thresholds, even when the affiliate dashboard suggests positive ROAS. This week, calculate your all-in affiliate cost stack — visible commission plus platform fees plus fraud allowance plus management labor plus cannibalization estimate — and compare the result to your paid acquisition spend CAC-to-revenue ratio. If the all-in number is more than 1.5x your paid CAC ratio, you are running an affiliate program that's destroying margin on every attributed order while telling you it's working.
