Affiliate Marketing
Affiliate Content Strategy
Affiliate content has one job: help a reader make a decision they will not regret. Pages that do that earn commissions. Pages that exist to hold links do not — and readers can tell the difference in seconds.
Content that converts vs content that ranks
New affiliates usually write one kind of page: the review that is really an advertisement. It lists features, repeats the manufacturer’s claims, and ends with a link. It ranks poorly and converts poorly, for the same reason: it helps no one decide. Content that converts starts from the reader’s decision, not from the product. The reader is choosing between options, or deciding whether to buy at all, or working out which version fits their situation. A page that answers that decision — honestly, with evidence — is the page that earns the click and the sale. A page that only describes the product is redundant: the merchant’s own page already does that.
Content that ranks and content that converts are not enemies, but they are different jobs. A page can rank because it answers a common question thoroughly, and still convert nothing because its readers are researching, not buying. The strategy is to build both kinds deliberately: informational pages that earn traffic and trust, and decision pages that earn commissions — and to link them so the research reader finds the decision page exactly when they are ready. Our search intent guide explains how to tell the two apart before you write a word.
The five page types that do the work
Almost all affiliate income comes from five page types, and a content plan should be built around them rather than around whatever is easiest to write.
1. The single-product review
A review earns trust by being specific. What you tested, how you tested it, what surprised you, what disappointed you, who it suits and who it does not. The most valuable sentence in a review is usually a negative one, stated fairly: the feature that is weaker than advertised, the case where the product fails, the buyer who should choose something else. A review with no weakness is not a review; it is a brochure, and readers treat it as one. If you have not used the product, say so — and do not publish a review that pretends otherwise. Our affiliate disclosure policy requires honesty about whether a product was actually used, and the FTC treats a fake hands-on claim as a misleading endorsement.
2. The comparison
Comparisons are the highest-converting affiliate pages, because their readers have already narrowed the choice to two or three options. A good comparison is structured around the reader’s decision criteria, not the products’ spec sheets: price, the feature that matters most for the use case, reliability, support, and the total cost of owning the product. It names a winner — and it names the conditions under which the winner changes. A comparison that declares every option excellent helps no one decide, and earns nothing.
3. The “best for” list
A “best” list is a comparison with a verdict per use case: best for beginners, best for a small budget, best for heavy use. Its honesty depends entirely on its method. A list with no stated method — no testing, no criteria, no reason for the order — is exactly the kind of page the FTC and search engines both treat with suspicion. State how you chose, what you tested, and what would change the ranking. A method paragraph takes five minutes to write and is the difference between a useful list and a ranking ordered by commission.
4. The tutorial that uses the product
A tutorial teaches the reader to do something, and the product appears as the tool that does it — in context, doing real work. Tutorials convert well because the reader is already doing the job and can see the tool working. They also rank well, because they answer a genuine how-to question. The rule is proportion: the tutorial must be genuinely useful without the product. If removing the product leaves nothing, the page is an advertisement wearing a tutorial’s clothes.
5. The problem page
A problem page starts from the reader’s problem — a slow website, a leaking budget, a room that is too hot — and works through the solutions, of which a product is one. These pages attract readers who do not yet know a product exists, which makes them earlier in the buying journey and slower to convert, but they build the trust and the traffic that the decision pages feed on. Every cluster needs problem pages; a site of only reviews and comparisons has no way to reach a reader before they have already chosen.
How much content, and in what order
A realistic starting plan is small and decision-first: a handful of problem pages that bring in readers, a comparison for the niche’s main choice, and two or three reviews of the products you know best. Ten thorough pages beat a hundred thin ones — a principle that holds across blogging, as our blogging guide explains. Publish the comparison early: it is the page your other pages will point to, and it is where your first commissions are most likely to come from. Then add problem pages that link into it, so each new page sends its readers toward a decision instead of toward an exit.
Do not write fifty pages before publishing any of them. Publish in small batches, watch which pages get traffic and which get clicks, and let the data choose the next pages. A page that earns traffic but no clicks may need a clearer decision section. A page that earns clicks but no sales may be sending the wrong readers to the wrong product. The conversion optimisation guide covers how to read those signals and fix the page instead of writing a new one.
Writing a review that a reader trusts
Trust in a review comes from specifics that cannot be copied from a product page: the measurement you took, the photograph you shot, the time it saved or cost, the customer support exchange, the moment it failed. Write those down. Name the buyer the product suits and the buyer it does not. Say plainly whether you bought it yourself or received it, and whether the merchant had any say in what you wrote. Disclose the affiliate relationship at the top of the page, before the first link — the detailed rules are in our disclosure requirements guide. And keep the review current: a review that still recommends a discontinued model, or quotes a price from two years ago, is worse than no review, because it sends a ready buyer to a dead end.
Where links belong — and how many
An affiliate link belongs where a reader who has decided to buy would look for it: beside the product name in a comparison table, at the end of a review’s verdict, in a “where to buy” line. It does not belong in the first paragraph, repeated in every sentence, or attached to words that are not the product. There is no magic number of links per page, but there is a test: read the page with the links removed. If it is still useful, the links are in the right places. If it collapses into nothing, the page was built for the links, not the reader — and the reader will feel it.
Link text matters too. A link that says “check the current price” is honest about what happens next. A link that says “best deal” or “limited offer” makes claims you cannot verify and should not make. And a link must never be disguised: hidden links, links that do not look like links, and shortened links that hide the destination all break the trust the model depends on — and several programmes forbid them outright in their terms.
Updating beats publishing
Affiliate content decays faster than most writing, because products change: new models, price changes, discontinued lines, changed terms. A comparison that is a year out of date is a page that recommends the wrong product. Build updating into the plan from the start: re-check the top-earning pages every few months, re-verify prices and availability, and date the re-check on the page. Our editorial standard is a “last checked” date on every fee, policy and price claim, and affiliate pages need it most of all. An updated page also signals to search engines that the content is maintained, which is a real advantage over the abandoned review pages that fill most niches.
Last checked: 14 November 2026.