Affiliate marketing has an appealing premise. You pay only when a sale happens, so the channel cannot lose money. That premise is true in a narrow sense and misleading in a wider one, and the gap between the two is where most program budgets quietly leak.
Three jobs, one commission
Look at who is in a typical program and you will find partners doing three quite different jobs.
| Role | Typical partners | What they contribute |
|---|---|---|
| Introducers | Content sites, creators, newsletters | Put your product in front of someone who had not considered it. |
| Persuaders | Reviewers, comparison sites | Help someone who is choosing between you and a competitor. |
| Closers | Coupon, cashback and loyalty sites | Give a final nudge, or a discount, at the point of purchase. |
Most programs track on a last-click basis. The partner whose link was clicked most recently gets the whole commission. Since closers act last by definition, they win a disproportionate share of credit, including for customers an introducer found weeks earlier.
Last-click affiliate tracking pays the partner who finished the sentence, not the one who started the conversation.
The cost that does not show on the invoice
When a closer is credited with a sale that would have happened anyway, you pay twice: once in commission and once in the discount. Worse, the introducers who created the demand see nothing for their effort, conclude that your program does not pay, and promote someone else. Over time the program fills with partners who are good at capturing sales and empties of those who are good at creating them.
This is not an argument against coupon and loyalty partners. They have large, real audiences, and some shoppers genuinely will not buy without a deal. The argument is that their contribution is different and should be priced differently.
Pricing by contribution
A few adjustments move a program toward paying for what it gets:
- Different rates for different roles. Higher commission for content and review partners, lower for partners who operate at checkout.
- New-customer premiums. Pay more when the buyer has never ordered before. Most platforms can pass this flag.
- Attribution rules that recognise early touches. Some platforms allow credit to be split, or allow an introducer's click to take priority over a later coupon click.
- Exclusive codes. A code tied to one partner shows who actually drove its use and stops it spreading to every coupon site.
- Test it. Pause a partner type in one region for a few weeks and watch total sales. If they barely move, the commissions were not buying much.
Compliance is part of the value
A program is also paying for its reputation. In the United States, the Federal Trade Commission expects anyone who earns from a recommendation to disclose that clearly, close to the recommendation itself. The advertiser shares responsibility for making sure partners do. Add to that partners bidding on your brand name in search, or publishing expired or invented codes to capture traffic, and policing becomes a real job.
A well-run program has written terms on all of this, checks them, and acts on what it finds. The effect is to protect the honest partners as much as the brand.
A better question
The usual way to judge an affiliate program is by its cost per sale. A more useful measure is cost per additional sale: what you paid, divided by the orders you would not otherwise have had. It is harder to calculate and far more informative. Programs that look expensive on the first measure sometimes look excellent on the second, and the reverse.
More on how we approach this in affiliate and partnerships.