The commission rate is the first number on every affiliate program page, and it tells you almost nothing about whether to join. A 40% cut of a product your readers don’t need is 40% of zero. Whether anyone buys depends on fit and trust, and neither of those shows up in the sign-up form.
If you’re picking your first programs, or auditing the ones already on your site, the order of your questions matters more than the answers. This piece lays out a sequence for how to choose affiliate partners that puts payout last.
The commission rate is the last thing you should look at
Think of the rate as a multiplier. Your earnings are roughly the number of people who buy, times the amount you earn per sale. The rate only touches the second part. The first part depends on whether your readers have the problem the product solves, and whether they believe you when you say it solves it.
Rates also carry less information than they seem to. A high percentage can reflect a healthy margin, and it can also reflect a product that needs heavy incentives to sell. The number alone doesn’t tell you which.
So run the vetting in this order:
- Does your audience have the problem this product solves?
- Have you used it, and can you name where it falls short?
- Would you recommend it if it paid nothing?
- Does the offer itself hold up when you read the terms?
- Only then: what does it pay?
A partner that fails an early question doesn’t get to the next one. A better rate can’t buy it back.
Gate one: does your audience actually have the problem this product solves?
Start with what your readers are asking, not with what the product can do. Look at the search queries that bring people to your site, the comments and emails you receive, and the questions that come up repeatedly in the communities where your readers spend time. If you’re still building an audience, define your reader and their main problem as precisely as you can. Understanding what people mean when they search is a skill in itself, and search intent goes well beyond the keyword.
Then write one sentence: “My reader struggles with X, and this product helps by doing Y.” If you can’t fill in X without stretching, the product fails this gate.
Here’s a hypothetical to show the pattern. Say your site teaches beginners to plan and publish content. A VPN program might pay generously and have a polished affiliate dashboard. But your readers haven’t asked about VPNs, and nothing in your content connects to one. You’d be inventing a need to fit the offer, which is backwards.
Adjacent relevance is the trap here. A product can sit in the same broad industry as your site and still be irrelevant to the person reading a specific page. Check relevance at the level of the individual article, not the niche. A scheduling tool might suit your post on content calendars and have no place in your post on writing headlines.
Gate two: have you used it, and can you say specifically where it falls short?
A recommendation is only worth something if it contains information the reader can’t get from the product’s own sales page. That information comes from using the product on a real task, not from watching a demo or skimming other people’s reviews. Rewriting someone else’s review gives your reader a second-hand opinion with your name on it.
Most tools offer a free tier, a trial, or a low-cost first month. Use that to do the actual job your reader would do, and pay attention to where it gets awkward. Slow exports, confusing settings, features locked behind higher plans, and support that takes days to answer are all things a reader would want to know before spending money.
The test for this gate is simple: name one specific limitation and say who it matters to. “The free plan caps the number of projects, so it suits someone testing one site and not someone running several” is a limitation a reader can act on. “It might not be for everyone” is not.
If you can’t name a real shortcoming, one of two things is true. Either you haven’t used it enough, or you aren’t saying what you found. Both mean the product isn’t ready for your recommendation yet. Using it first is free, and it costs you nothing but time.
Gate three: would you still recommend it if it paid nothing?
This is the gate that catches the other two being rationalised. It’s easy to convince yourself a product is relevant and well-tested when there’s money attached.
Try these checks:
- Strip the link. Write the recommendation as if there were no affiliate link, then read it. Does it still make a case, or does it only work as a lead-in to a button?
- Compare against the alternatives. If a competitor suits your reader better but pays less, which one are you actually recommending? The honest answer is the better fit.
- Ask the friend question. If someone you know had exactly your reader’s problem, would you bring this product up unprompted?
Notice whether the commission has ever changed your answer. If a product moved up your list when its rate went up, or dropped when it went down, the rate is doing the choosing. That’s a signal to rethink the selection, since the product hasn’t changed at all.
None of this means you can’t be paid for good recommendations. Earning from a product you’d have recommended anyway is the whole point of affiliate marketing done well. The free test only checks which came first, the recommendation or the payout.
Red flags in the offer itself: cookie windows, pressure tactics and vague claims you can’t verify
A product can pass all three gates and still come with an offer you shouldn’t accept. Read the terms before you join, and look for these.
Cookie windows and attribution. The cookie window is how long after a click a purchase still counts toward you. For a product people buy on impulse, a short window may be fine. For something a reader will think over for weeks, like business software or a course, a short window means you do the work of convincing someone and get nothing when they buy later. Also check the attribution rule: if a later click from another source replaces yours, your credit can disappear. Terms differ widely between programs, so read the actual document and don’t assume.
Pressure tactics. Countdown timers that reset, “only a few spots left” claims with no real limit, and launch schedules that push you to promote on the seller’s timeline all rely on manufactured urgency. If you’d have to pass that pressure to your readers to make the offer work, it’s a problem. Your site’s editorial standards rule out fabricated urgency anyway.
Claims you can’t verify. Watch for supplied swipe copy and marketing assets that promise fast results, guaranteed outcomes or specific income. You’d be repeating those claims under your own name. If you can’t check a claim against your own use of the product or a credible source, don’t publish it.
Terms that limit honesty. Be cautious of programs that forbid mentioning competitors, discourage you from describing limitations, or restrict how you disclose the relationship. A partner that needs you to be less honest is telling you something about the product.
Unclear money terms. Check payout thresholds, payment schedules, and whether the program can change commission rates or reverse commissions after the fact. Vague terms on any of these are worth asking about before you’ve linked to the product in twenty posts.
How to say no (and how to drop a partner you’ve already promoted)
Declining a program takes no explanation. If a company approaches you directly, a short reply is enough: the product isn’t a fit for your readers right now, and you only recommend things you’ve used. Leave the door open if you like, but don’t promise coverage.
Dropping a partner you’ve already promoted takes more care, because readers may have acted on your advice. Reasons to do it include a price jump, a product that has changed for the worse, a company that handles complaints badly, or a better option you’ve since found. The process:
- Find every placement. Search your site for the product name and the affiliate link domain. Links hide in older posts, resource pages and email templates.
- Remove or replace the link. If a better option exists, point readers to it and say why you switched.
- Add a short update note to any post where you strongly recommended the product, so readers who return know the advice has changed.
- Tell readers plainly if it matters. If the product has problems that could affect people who already bought it, say so. Silent deletion is easier, but it leaves earlier readers without information.
You’ll lose some income when you do this. That’s the real price of the filter, and it’s the reason the filter means something. A reader who sees you remove a paying link because the product slipped has good reason to trust the links you keep.
Disclosing the relationship so readers can weigh your recommendation
Disclosure gives your reader what they need to judge your advice: you may earn money if they buy through your link. Put it where they’ll see it before they click, in plain words, not tucked into a footer. A sentence such as “If you buy through this link, I may earn a commission at no extra cost to you” does the job.
Requirements vary by country and by platform, and affiliate programs have their own terms about disclosure too. Check the rules that apply to you and to the places you publish. Disclosure that meets the minimum is a starting point, but the stronger version also says why you chose the product, which is where your gate-two notes become useful.
Choosing partners carefully and disclosing them openly are two halves of the same practice. For the wider argument about why this matters for a site built on trust, see why your affiliate strategy needs a human filter.
Run one link through the gates this week
Pick one affiliate link that’s live on your site right now and run it through the three gates this week. Ask whether your readers have the problem, whether you can name a specific limitation from your own use, and whether you’d recommend it if it paid nothing. If it fails any of them, take it down, and notice what that tells you about how you chose it.