Spotting affiliate income mistakes before they cost money

The failures in this income line are mostly structural. They are decisions about what to write, what to build on and what to ignore, and every one of them looks sensible on the day it is made.

Nine of them, with what each one actually costs and what to do instead.

What to take away

1. Picking the category by what it pays

Someone tells you a category pays well, so you write about it. You do not use those products, you have no view on them, and neither does your audience.

What it costs is the only advantage you had. Your recommendation is worth something because it comes from someone with a track record in a subject. Moved to a category you know nothing about, it is worth what any stranger's is.

Instead: start from the decisions your readers are already making. If nothing in your subject has a program attached, that is information about the subject, not a reason to change subject. This is ordinary market research at a small scale, and the Small Business Administration's guide to market research describes the direct version that fits a one-person business.

2. Building on one merchant

One program provides most of the income, so it gets most of the pages, and the arrangement works until it does not.

Programs close, rates get revised, accounts get terminated, and merchants move networks. Any of those can arrive without warning and take the majority of this income line with it, along with hundreds of links that now go nowhere.

Instead: know your concentration figure before you need it, and treat a very high one as the risk it is. That does not mean joining more programs for the sake of it. It means not letting one merchant become load bearing without noticing.

3. Treating traffic as intent

A page has plenty of readers and produces almost nothing, so the conclusion is that the audience is bad or the link is in the wrong place.

Usually neither. People who arrived to read something are not people who arrived to decide something. Volume tells you about interest in a topic. It tells you nothing about whether anyone was about to buy.

Instead: work out which pages sit at a decision and which sit at a curiosity, and stop expecting the second kind to convert. A small number of decision pages will do more than a large number of popular ones.

4. Publishing volume before publishing anything good

The plan is to produce many pages quickly and see what sticks. It is appealing because it feels like progress and each page is cheap.

What it costs is the thing that would have worked. Thin pages assembled from specification sheets compete against thousands of identical ones, hold nothing a reader could not get elsewhere, and drag down the reputation of the pages that took real work.

Instead: make one page that could only have been written by you, and see what it does over a few months. If that does not work, twenty versions of it will not either. What genuine research behind a recommendation looks like is set out in the affiliate income guide.

5. Believing the dashboard in both directions

Two opposite errors, from the same mistake.

The first is treating the reported figure as what you earned. It is provisional until the hold period closes, and reversals arrive weeks later.

The second is treating a low figure as proof you had no influence. Tracking fails routinely: blocked or shortened cookies, an app that opens without the tag, a purchase on another device, a coupon site that fires last, a reader who decides three weeks later. Your recorded number is a lower bound on your actual influence and always will be.

Instead: treat the dashboard as one imperfect instrument. Judge a page on whether it answered the question well, and use the revenue figure for direction rather than truth.

6. Never reading the clauses that decide the outcome

The rate is on the front page of the program. Everything that determines whether you keep the money is further down: the attribution window, the exclusions, the hold period, the reversal triggers, the payout threshold, and what happens to an unpaid balance if the account is terminated.

Rate changes and terminations are where people lose real money, and both are usually covered explicitly in terms nobody read. How that chain works, and where it breaks, is covered in the guide to affiliate income.

Instead: read the terms once, properly, before joining, and write the four numbers that matter onto your own record: window, hold, threshold, and what happens on termination. Keep the program's statements alongside them, since those are the documents that explain any figure you later have to account for, which is the point of the Internal Revenue Service's recordkeeping guidance.

7. Letting old pages speak for you

You published something two years ago. You have not thought about it since. It still tells readers to buy a product you would not buy now.

This is the most common and least visible failure in this business, because nothing breaks and no alert fires. The page just quietly becomes wrong, and it keeps making a recommendation in your name.

Instead: an inventory and a review cycle, run on a schedule rather than when you remember. The twelve checks in the affiliate audit are the version of this that takes fifteen minutes a page.

8. Competing with yourself without noticing

You recommend something on a commission, and it is a near substitute for something you sell. Or a sponsorship in the same category conflicts with a program you are already in.

The cost here is not lost revenue so much as an incoherent position. A reader who sees you recommend a competing product and sell your own is left working out which one you actually mean, and that ambiguity is expensive.

Instead: keep a short map of what you sell, what you recommend, and where they overlap. Where they do, say plainly which is which and why, and where your own thing genuinely is not the right answer, say that too.

9. Optimizing the link instead of the decision

Endless tinkering with placement, wording, button color and the number of links, on a page whose underlying recommendation is vague.

The gain from that work is real and small. The gain from making the recommendation genuinely useful, specific about who it is wrong for, and honest about the drawbacks, is much larger and much less fun.

Instead: fix the decision first. A page that tells a reader exactly who should not buy something will outperform any amount of link arrangement, and it will keep doing it in two years.

The thread running through them

Every one of these is a version of paying attention to the transaction instead of the decision.

The transaction is the click, the window, the rate, the placement, and the decision is what the reader came to make. Only one is yours to improve; the other largely belongs to the merchant's checkout.

Creators who do well are not better at links; they are better at being useful at the exact moment somebody has to choose.

Common questions

How do I know which of these I am making?

Look at where your effort went last month. If most of it went to the mechanics rather than to the recommendation, you have some combination of the fourth and the ninth.

Is a low conversion rate my fault?

Often not. Conversion is mostly a property of the merchant's checkout, their pricing and their availability. That is why it is worth comparing programs on how they convert your traffic, and why a well written page can be wasted on a badly run merchant.

Should I compare my results against what other people publish?

They cannot tell you much, for the same reasons set out under published benchmarks: different subjects, different programs, different traffic, and no way to check any of it.

What if the honest answer is that nobody needs to buy anything?

Then write that, and take the loss. Some subjects have no version of this that survives contact with your conscience, and the effort is better spent on an income line where you are selling something of your own, such as memberships.