Rules
Part of Affiliate income for 2027: disclosures, real reviews and a launch plan that holds
An affiliate income framework built on conditions, not enthusiasm
Affiliate income framework: three conditions a recommendation must meet, where the model does not fit, and rules to write down before money is at stake.
The useful question is not how to earn commissions. It is which of your recommendations should carry one at all, and what this income line is doing in your business next to everything else.
Answer those two and the tactical questions mostly answer themselves. Skip them and you end up with a site full of links and no position.
What to take away
- Three conditions before a recommendation carries a linka real decision, real evidence, and a program that actually pays.
- Affiliate income rewards volume and search traffic. A small audience with deep trust usually earns more from something you own.
- Write your rules once, while nothing is at stake, and follow them on the day they cost you money.
Three conditions
Apply these to a specific recommendation, not to your site in general.
Three conditions before you link
Is there a real decision?
check evidence
no recommendation, no link
There is a real decision. Somebody is choosing between options and it matters to them which one they pick. No decision means no recommendation, and a link on a page where nobody was deciding anything is decoration.
You have real evidence. You have used it long enough to know how it fails, or you have done work that a reader could not do themselves. Specifications are not evidence, and a strong opinion formed from other people's reviews is not either.
The program pays reliably. Not generously, reliably: a workable attribution window, exclusions that do not remove the products people actually buy, a hold period you can live with, and a threshold you will clear. A high rate on a program that never converts a sale is worth nothing at all.
Fail the first two and there should be no link. Fail the third and the recommendation still belongs on the page, without one.
Where this model fits, and where it does not
Where this model fits
| Your situation | What this suggests |
|---|---|
| Large audience, low individual trust, decisions in your subject | A reasonable fit. This is what the model is built for |
| Small audience, deep trust, specific problems | Usually the wrong line. The same trust converts far better into something you make |
| Search traffic arriving at the moment of a decision | The strongest position available, and worth protecting with quality rather than volume |
| A subject where the best answer has no program | Say the best answer anyway, and find the income elsewhere |
| A category dominated by coupon and cashback partners | Expect to lose the last click on a large share of sales, whatever you do |
| A subject where the honest advice is to buy nothing | Not an affiliate subject. Accept it early |
The most common misfit is the second row: a few hundred committed readers, monetized through commissions on other people's products, when the same people would happily pay for something of yours. That comparison is worth making explicitly rather than assuming, and digital products is the version of the argument worth reading next to this one.
Where the affiliate model fits
Fits
- Audience
- Large, low trust
- Traffic
- Decision moment
- Best answer
- Has a program
- Category
- Few coupon rivals
- Honest advice
- Buy something
Does not fit
- Audience
- Small, deep trust
- Traffic
- No decision moment
- Best answer
- No program exists
- Category
- Coupon dominated
- Honest advice
- Buy nothing
Against your own products
When a recommendation could point at something you sell or something someone else does, decide by which is better for the reader, and then be explicit about what you did.
Three positions, all defensible.
Recommend your own, and say why. Fine when it genuinely is the better fit, and only when you say plainly that it is yours.
Recommend theirs, on a commission, and disclose it. Fine, and it builds more credibility than the first option, because you demonstrated you would send people elsewhere.
Recommend theirs with no link at all. The strongest position, and worth keeping available for the cases where you want no question about the motive.
What does not work is holding all three implicitly and choosing whichever pays on the day. Readers notice the pattern long before they can articulate it.
Rules you write once
Write these while no money is at stake. They are worth almost nothing agreed in the abstract and a great deal when applied on a day they cost you something.
- The unpaid testwould you recommend this if nobody paid you?
- No ordering by payout, ever, in any list.
- Every recommendation names who it is wrong for.
- Drawbacks appear in your own words, not softened.
- You will link to something that pays you nothing when that is the right answer.
- No program's representative writes or approves your conclusion.
- The disclosure sits before the link, in plain language, in the piece itself.
Keep them somewhere you will actually see them when writing. The third rule is not only yours to set: the Federal Trade Commission's guide to disclosure for influencers states the principle plainly for United States creators, and its questions and answers on the endorsement guides covers the awkward cases.
Most failures in this line are not cases where somebody lacked a rule. They are cases where a rule existed and was quietly not applied on the one day it mattered.
What share of your income this should be
There is no correct number, and anyone offering one is guessing. There is a useful test.
Ask what happens if the largest program terminates your account tomorrow with an unpaid balance. If the answer is a bad month, this line is sized correctly. If the answer is that the business stops, it is not, and no amount of growth in that program fixes it, because the underlying problem is that you do not hold the terms.
Apply the same reasoning to platform programs in the control axes: rank income lines by how much of the arrangement is yours, not by what they currently pay.
Commissions score low on every axis. You do not hold the customer or set the price, and the other side can change the terms. The income stops when the traffic does.
None of that makes it a bad line. It makes it a line to run alongside something you own, rather than under.
The quarterly decision
Once a quarter, twenty minutes, on the structure rather than the numbers.
- Which programs are you in, and are you still in each of them for the reason you joined?
- What share of the income comes from the largest one?
- Which recommendations would you drop if there were no commission attached? Drop one.
- Where does this line compete with your own products, and is that stated on the page?
- Is the maintenance load still one you can carry?
The last question is the one that decides whether the line survives. Pages you cannot keep current stop being an asset and become a standing risk, and the honest answer is often to have fewer of them.
Common questions
Is affiliate income compatible with sponsorships in the same category?
Sometimes, and it needs checking rather than assuming. Sponsorship agreements frequently include exclusivity language broad enough to catch an existing affiliate arrangement, so look before you sign either one.
Should I disclose an affiliate link when the recommendation is negative?
Yes. The disclosure is about the relationship, not about the verdict, and a reader is entitled to know the arrangement exists whichever way your conclusion went.
What if I would fail my own unpaid test on pages I have already published?
Then those pages are the audit queue. Rewrite or retire them, starting with the ones that get the most readers.
How does this line fit a career that is mostly something else?
As a top up on work you were doing anyway, not as a thing to build around. Which shape of business you are actually in, and what that means for where the effort should go, is the subject of the career shapes framework, and the mechanics of the commission chain itself sit in affiliate income.
