The Affiliate Economy: How Coupon Sites Get Paid (And Why That Matters to You)

Every coupon site you've ever used — from the household names down to the SEO farms ranking on the third page of Google — was built on a single mechanic: last-click attribution. A shopper clicks a link, lands on a merchant, buys something, and the network running the tracking awards a commission to whoever owned the last click. The shopper sees a discount. The merchant sees a sale. The aggregator sees a payout. Everybody, in theory, wins.

Understanding the plumbing of that mechanic — who pays whom, in what amounts, and on what terms — explains almost everything about why coupon listings look the way they do. It explains why expired codes stay up. It explains why some sites push you toward a "click to reveal" page even when the code is already visible. It explains the Honey scandal of 2024 that cost the company hundreds of millions of dollars in goodwill with creators. And it explains why two listings for the same store, on two different sites, can look identical down to the typography but produce different outcomes at checkout.

This is not a screed against affiliate marketing. Affiliate marketing is how a large fraction of the consumer-facing internet — including, full disclosure, CouponHive — actually pays its bills. But the incentives created by last-click attribution are real, and shoppers benefit from understanding them.

The networks

There are five affiliate networks that handle the bulk of US coupon and cashback traffic. In rough order of US retail share:

CJ Affiliate (formerly Commission Junction, now owned by Publicis). Long-tenured, large catalog, particularly strong in apparel and home goods. Standard cookie window of 30 days for most merchants, with some offering 7 or 14.

Rakuten Advertising. The advertising arm of the Japanese conglomerate, separate from the consumer-facing Rakuten cashback site (which is itself a publisher on the network). Strong in retail and travel. Cookie windows vary widely.

Awin (which absorbed ShareASale in 2017 but still runs ShareASale as a separate brand). Two networks under one ownership; ShareASale is more SMB-focused, Awin handles enterprise. Both are common for direct-to-consumer brands.

Impact (sometimes called Impact.com). Newer than the others, popular with tech-forward merchants and creator partnerships. Often offers more flexible attribution windows.

Pepperjam and the smaller independents (Sovrn Commerce, formerly VigLink; Skimlinks; etc.) round out the long tail.

A coupon site is, in network terms, a publisher. The merchant is an advertiser. The network is the matchmaker, tracking middleman, and clearinghouse for payments. When a shopper clicks a coupon link, the network drops a tracking cookie. When the shopper buys something within the merchant's defined window, the network attributes the sale to the publisher and the merchant pays a commission, typically a percentage of the order total minus tax and shipping.

The commission structures

Most affiliate commissions are CPS — cost per sale, paid as a percentage of the order. For US retail, the typical range looks like this:

• Mass merchants and big-box retailers: 1 to 3 percent. Walmart, Target, Best Buy, and similar tend to pay at the low end because their margins are thin and their conversion rates are high.
• Apparel and accessories: 4 to 8 percent. Macy's, Nordstrom, and most direct-to-consumer apparel brands pay in this band.
• Beauty and personal care: 6 to 12 percent. Sephora, Ulta, and DTC beauty brands run higher because margins are higher and the lifetime value of a new customer is large.
• Home goods and furniture: 3 to 8 percent, with bigger orders often capped at a flat dollar amount.
• Specialty and niche DTC: highly variable. Some pay 15 to 25 percent on first orders, especially in subscription categories.

A few categories pay flat fees rather than percentages: financial products (a credit-card sign-up might pay $50 to $200), software subscriptions, and travel bookings.

There are also CPL (cost per lead) and CPA (cost per action) variants — paid for an email signup or a free-trial start — but in the coupon space, CPS dominates.

Last-click attribution and why it matters

Here is the mechanic that shapes the entire industry: most networks attribute the commission to the last publisher whose tracking cookie was set before the conversion. This sounds neutral. It is not.

Consider a typical shopper journey. A creator on Instagram posts about a sweater from a DTC brand and includes their affiliate link. The shopper clicks, browses, doesn't buy, and closes the tab. Three days later, the shopper is on a coupon site looking for a code for that brand. The shopper clicks the coupon site's outbound link to the brand's site, applies a code (or doesn't — sometimes there's no code, just a click that drops a fresh cookie), and completes the purchase. The commission goes to the coupon site. The creator who originally introduced the shopper to the brand gets nothing.

This is not a bug. It is the explicit design of last-click. The argument for last-click is that the coupon site provided the final value-add (a discount, or at minimum a re-engagement) and so deserves the commission. The argument against is that the creator did the actual marketing work and the coupon site captured the credit.

The 2024 Honey scandal — when investigative video creators documented that the Honey browser extension was, in effect, swapping out creator affiliate cookies with its own at checkout — was an extreme version of this dynamic. Honey wasn't just being the last click in the natural sense; it was actively rewriting the cookie state in the user's browser to ensure it was the last click, even when the user had not clicked through Honey for that specific session. The economic harm to creators was real and concrete, and the resulting class actions and platform bans materially changed how browser extensions in this category operate.

But the underlying incentive — get to be the last click — exists for every coupon site, not just the ones running browser extensions. It explains a lot of behavior.

Why expired codes stay up

The economic logic of last-click attribution does not actually require the code to work. Here is what happens when a shopper finds an expired code on a coupon site:

1. Shopper sees listing on the coupon site.
2. Shopper clicks the "Get Code" or "Reveal" button. This action triggers an outbound click to the merchant's site, often through a tracked redirect URL. The network's cookie is set in the shopper's browser.
3. Shopper arrives on the merchant's site, tries the code, and it doesn't work.
4. Shopper, now already on the merchant's site, browses around. Maybe finds a different sale. Maybe just decides to buy at full price. Maybe applies a different working code from a different listing on the same coupon site.
5. Shopper completes a purchase within the cookie window.
6. Network attributes the sale to the coupon site, even though the original listing was wrong.

The coupon site got paid for a code that didn't work. The shopper still paid full price (or near it). The merchant paid a commission for a sale they would have likely gotten anyway, since the shopper was already shopping. This is the "affiliate tax on the consumer base" critique that has been made by some retail economists for the better part of a decade — that a substantial fraction of coupon-site commissions are paid on transactions where the coupon site provided no actual value to either the shopper or the merchant.

So why don't merchants stop the practice? Some have. Several large retailers — most visibly Nordstrom and a few of the direct-to-consumer giants — have, over the past few years, restricted or terminated their participation in coupon-site affiliate programs specifically. They concluded that the incremental sales attributed to the coupon channel were not, in fact, incremental. Most merchants haven't gone that far, because untangling the data is hard and because the coupon channel does drive some genuinely incremental business, particularly from price-sensitive shoppers who would otherwise abandon.

But the structural incentive — that an expired listing still earns commission on a passive shopper — is why expired listings persist. A coupon site with no editorial review will tend toward more listings, not fewer, because every listing is a potential cookie drop, and the cost of being wrong about freshness is borne almost entirely by the shopper.

The "click to reveal" pattern

You have probably noticed that many coupon sites hide the code itself behind a button labeled "Click to Reveal" or "Show Code." There is a UX argument for this — it lets the site display many listings compactly. There is also an attribution argument: the click on the reveal button often opens a new tab to the merchant's site, dropping the affiliate cookie before the shopper even sees the code. This means the publisher is now in the attribution pool for any purchase the shopper makes, regardless of whether the code works.

Some sites are honest about the second tab — they open the merchant in a clearly visible new tab. Some are sneakier, opening a popunder or a series of redirects that the shopper barely notices. The end state is the same: cookie set, attribution claimed.

This isn't necessarily nefarious. If the code works and the shopper saves money, everyone benefits. But if the code doesn't work and the shopper buys anyway, the publisher has been compensated for a phantom service. As a shopper, you can tell the difference at home: pay attention to whether your tab count went up after clicking "reveal." If it did, the cookie was set.

Editorial vs. paid placement

A separate question, often conflated with affiliate commissions, is whether a listing's prominence on a coupon site is paid for. The answer is: sometimes.

Affiliate commissions are paid per sale. They do not, by themselves, change the order of listings on a page.

Premium placement — the brand showing up at the top of a category page, or in a "featured" carousel, or in an emailed newsletter — is sometimes additionally paid, either as a flat sponsorship fee or as a higher commission tier. Reputable publishers disclose this with a "Sponsored" or "Featured" tag. Less reputable ones do not.

SEO-driven placement — what shows up when you Google "[store] coupon" — is typically not directly paid; it's a function of how good the publisher's SEO is. But the publisher's incentive to optimize for that traffic is, again, last-click attribution.

A useful tell: when you land on a coupon page and the same listing is repeated three times, with slightly different titles ("20% Off Sitewide," "Save 20% Today," "Exclusive 20% Deal"), the publisher is hedging. They want as many cookie-drop opportunities as possible. There is no editorial reason to repeat a code three times. There is a strong attribution reason.

The Wayback Machine test

A useful exercise, if you want to know whether a coupon site has always been an aggregator or whether it started as something more editorial: pull it up on the Wayback Machine and look at the same site five or seven years ago.

You will find that a substantial fraction of today's most-trafficked coupon sites began as something else — a personal blog, a niche enthusiast site, a deal-of-the-day curator with actual human curation — before being acquired by a private-equity-backed parent and consolidated into a programmatic aggregator. The branding usually stays the same. The independent voice usually doesn't. This is not unique to coupons; it has happened across the consumer-recommendation web, from product-review sites to recipe blogs to local-news directories. But it is especially pronounced in coupons because the unit economics of the affiliate model reward scale over depth.

What this means for shoppers

A few things worth carrying in your head the next time you use a coupon site:

The site is, almost certainly, getting paid by your purchase. That is not by itself a reason to distrust it. It is a reason to remember that the site's incentive is to get you to click and buy, not necessarily to save money.

If a code doesn't work the first time, don't be loyal to the site that listed it. The cookie has already been set; if you go elsewhere and find a working code from a different publisher, you may be helping that publisher capture credit they actually earned. Not your problem to solve, but worth knowing.

Browser extensions that aggressively swap cookies are, by 2026, increasingly restricted by browser vendors and merchant agreements, but they still exist. If you use one, you should know what it's doing on your behalf at checkout.

And the simplest test of any coupon publisher: do they tell you when a code was last tested, and by whom? Sites that have invested in actual verification will tell you. Sites that haven't, won't.