An affiliate store that gets approved and stays approved
We build the site, get you approved with Amazon and the networks, place the links so Google and the FTC are both satisfied, and do the search work that brings the clicks.
Would an affiliate layer earn anything on your traffic?
Four numbers. It works out what the links would earn, and whether the account would survive its first six months.
Monthly visitors to the pages that could carry links
Not the whole site. The pages about things people buy.
Typical price of what you would recommend
In dollars. A rough average is fine.
What are those pages?
This decides how many readers click through to buy.
Commission rate
Amazon pays by category. Software and services pay far more.
The answer
Put in four numbers
Commission a year
—
Do this next
Show the numbers
Before you apply
Want someone to look at whether the account would survive?
Send the numbers over and you get back whether the site would be approved as it stands, which pages would carry the links, and what the first 180 days would need to look like.
Goes to one inbox.
In short
An affiliate store earns when three things are true at the same time: content someone actually wanted to read, an account that survives its first 180 days, and links tagged so that Google and the FTC are both satisfied. Most stores fail the second and third quietly, months after launch, and the owner finds out when the account closes. This is built in the order that keeps all three true.
What we take on
The site itself
Built or rebuilt: fast pages, real product research, and writing a person did. Since April 2026 Amazon requires commentary, analysis or transformation rather than its own product copy reworded, and it has been closing accounts that ignore that.
Approval, and the 180 days after it
The application prepared and submitted in your name, then the traffic work that produces the first three qualifying sales inside 180 days. That deadline, not the application, is what closes most new Associates accounts.
Links placed and tagged
Every affiliate link marked rel sponsored so Google reads it as a paid placement, and disclosed next to itself rather than in a footer. Live pricing pulled from the source, because static prices break the agreement and mislead the reader.
Search and answer engines
A link earns nothing without a click. Technical cleanup, the terms buyers actually type, and page structure that answer engines can quote, so the pages carrying the links are the ones people reach.
Where affiliate stores actually die
Almost none of them are killed by traffic. They are killed by the account. Amazon gives a new Associate 180 days to produce three qualifying sales and closes the account if they do not arrive. A store launched with no audience and no search plan spends that entire window invisible, and loses the program before it ever had the chance to earn from it.
The second killer is the link itself. Google asks that paid placements carry rel sponsored. The FTC asks that the commission is disclosed next to the recommendation, not in a footer, and says plainly that a disclosure page on its own is not enough. Both are small pieces of markup. Both are missing on most stores, and both are the kind of thing that gets discovered late.
The third is content the program will not accept. The agreement now asks for commentary, analysis or transformation that adds value, and rewritten product descriptions do not clear that bar. A store is a publication that happens to check out through somebody else's warehouse. It has to be worth reading before it is worth anything else.
So the order is fixed: something worth reading, then approval, then links placed correctly, then the search work that brings the clicks the account needs to survive its first six months. Done in any other order, the earlier steps get undone by the later ones.
Questions we get
Do I need a website already?
No. If there is one we work with it. If there is not, building it is part of this.
Amazon will not approve an application against a placeholder or an unfinished site, so the site has to be live and worth reading before anything is submitted. That sequencing is most of why applications get rejected.
Which programs do you set up?
Amazon Associates first, because almost everything is on it and it converts.
Then whichever networks fit the subject, among Impact, Awin, ShareASale, CJ and Rakuten, plus direct programs where a brand runs its own. Software and business tools pay far more per sale than physical products and are worth going after wherever the audience fits.
What happens if the application is rejected?
We fix what it was rejected for and reapply.
Rejections are almost always one of three things: too little original content, an unclear account of where the traffic comes from, or a site that is not finished. All three are cheaper to settle before applying than to argue about afterwards.
Who owns the accounts?
You do, all of them.
The Associates account and every network account is opened in your name against your own tax details, because the commission is your income and it has to be. We work inside them. Nothing is held at handover, and nothing has to be moved later.
How long before it earns?
Approval takes days. The first three qualifying sales have to land within 180 days, and that is the deadline that decides whether there is anything to build on.
Revenue follows traffic and traffic follows search, so the honest unit of planning is a quarter. Anyone promising a month is describing paid traffic, which Amazon disqualifies.
Is this passive income?
No, and it is worth being blunt about that.
Prices move, products go out of stock, links break, and programs change their terms, as Amazon did in April 2026 when it narrowed both the attribution window and what counts as original content. A store left alone stops earning quietly. What can be handed over is a documented routine for keeping it alive, which is the part almost nobody builds.
More in the guides and every answer in one place.
Shaheer Shaikh, operations lead at LARVOL, a San Francisco AI company working with global pharma on clinical trial data and model benchmarking. Six Sigma on the process side, Anthropic certified on the Model Context Protocol, ten years across eight industries. More about the firm