Affiliate Site Not Converting? Why Readers Click But Don't Buy (and the Seven Fixes)
The Orders report shows 400 clicks this month and three orders. The site is ranking, readers are clicking, and nobody's buying. That's a specific problem with a short list of causes, and the right one is almost always findable from the reports you already have, before you touch the pages.
This post covers how to read the click-to-order ratio in Amazon's reports and what's normal, the seven causes of an affiliate site that doesn't convert, one diagnostic for each, and the fix. It's the troubleshooting companion to the conversion rate post, which covers the levers on a site that's converting normally and could do better.
The short answer
Split Amazon's Orders report by tracking ID (one per guide or page type) and look at clicks against ordered items. Normal is 5% to 10% on guide traffic; under 1% with real volume is a problem.
Then test the seven causes in order: informational intent, price band, dead listings, button placement, geography, title mismatch, bot clicks. Each has a two-minute check and the first three cover most cases.
Reading the ratio
In Associates Central, the Orders report shows clicks and ordered items; the conversion column is one divided by the other. It only becomes useful once you're using separate tracking IDs for different parts of the site (the tracking IDs post covers the setup).
With one ID for everything, a 1% site-wide figure could be a 6% guide and a 0.2% informational post averaged together, and you can't see which.
What's normal: 5% to 10% on buyer guides for mid-priced products, higher for cheap consumables, 1% to 3% on informational pages that happen to carry links, lower for expensive items. Under fifty clicks on a page is too little to judge.
The seven causes
| Cause | Diagnostic | Fix |
|---|---|---|
| 1. Informational intent | The page's query is "how to", "what is" or "vs"; conversion under 2% is normal there | Write the "best X" guide and link to it; keep the info page as a feeder |
| 2. Wrong price band | Query implies budget ("under $50", "cheap"); picks are $120 | Re-select products to the query's price; one premium pick at most |
| 3. Dead or unavailable listings | Click your own buttons; "currently unavailable", "see similar", or a different product | Replace; refresh the guide; check monthly |
| 4. Buttons on the wrong products | Heatmap or outbound-click data shows clicks on products 7 to 10, not the pick | Put the verdict and its button in the first screen; button per section |
| 5. Geography | Analytics shows a big share of visitors outside the store your links point to | OneLink, or accept the loss and target the right country |
| 6. Title mismatch | High click-through from search, short time on page, clicks then no order: the page isn't what the title promised | Rewrite the page to the title's promise, or the title to the page |
| 7. Bot and self clicks | Amazon clicks far exceed analytics outbound clicks; clicks at 3am; your own testing | Separate ID for testing; ignore the inflated count; check link cloaking isn't attracting crawlers |
1. Informational intent
The commonest cause and the least fixable on the page itself. A reader on "how to clean a cast iron pan" clicks the pan link out of curiosity and buys nothing, because they already own a pan.
The intent post covers the distinction; the fix is to make sure every informational page links to the buyer guide for the readers who are shopping, and to stop judging the info page by its conversion.
2. Wrong price band
"Best bike lights under $50" with picks at $90 converts badly because the reader told you their budget and the page ignored it. The check is the query against the picks' prices. The fix is re-selection: products inside the band, with one stretch pick labelled as such.
3. Dead listings
Products go out of stock, get delisted, or the ASIN starts pointing at a different variant. The reader lands on "currently unavailable" and leaves. Click every button on your top ten guides once a month; it takes fifteen minutes and it's the most common fixable cause on sites over a year old.
4. Buttons on the wrong products
If the clicks are landing on the products at the bottom of the guide, the pick isn't getting the click, and the pick is the product you chose because it converts. A verdict box with its button in the first screen, and a button under each product section rather than one at the end, sends the clicks where the conversion is.
5. Geography
A UK reader clicking a link to amazon.com usually doesn't buy from a US store. Analytics shows where visitors are; if a third are outside the store you link to, that's a third of clicks that can't convert. OneLink redirects them; or, if the site is clearly for one country, target the keywords that way.
6. Title mismatch
A title that promises "tested" or "for beginners" or "under $30" and a page that delivers something else gets the click from search, loses the reader's trust in the first screen, and sends them to Amazon to look for themselves rather than to buy what you said. The signature is high click-through, low time on page, clicks, no orders. Make the page keep the title's promise.
7. Bot and self clicks
Amazon counts every hit on a tagged link, including crawlers, link checkers, preview fetchers and you. If Amazon shows 300 clicks and your analytics shows 60 outbound clicks, the ratio is being calculated on a number that's mostly not people. Use a separate tracking ID for anything you click yourself, and read conversion against your own outbound-click count instead.
Doing it with AffSERP
Causes 2, 3 and 4 are layout and selection, and they're built into guides from AffSERP's X Best Review tool: products are selected to the keyword's price band, each product gets its own button and the verdict's sits in the first screen, and the Refresh option replaces listings that have gone unavailable.
Intent, geography and the tracking-ID split are decisions about which keywords and IDs to use; the reports above tell you which one's the problem.
FAQ
Why am I getting Amazon affiliate clicks but no sales?
Seven usual causes: the traffic is informational and never meant to buy, the products are priced above what the query implies, the links go to unavailable or dead listings, the buttons are on the wrong products, visitors are in a country your links don't serve, the page doesn't match its title so readers bounce back to Amazon unconvinced, or the clicks are mostly bots and your own. Amazon's Orders report split by tracking ID narrows it to one or two in ten minutes.
What is a normal click-to-order rate for Amazon affiliates?
Amazon's conversion figure (ordered items divided by clicks) runs around 5% to 10% on buyer-guide traffic and 1% to 3% on informational traffic, with wide variation by category and price. Under 1% with hundreds of clicks means one of the seven problems; 0% with under fifty clicks usually means not enough data yet.
Does the 24-hour cookie cause low affiliate conversion?
It causes some loss on expensive products, where readers research on one day and buy on another, but it doesn't take a site from normal to zero. If conversion is low on a guide for $30 items, the cookie isn't the reason. Check the product and the intent first.
How do I know if my affiliate clicks are bots?
Compare Amazon's click count to your analytics' outbound clicks or your link manager's count; bots and preview-fetchers often hit links without rendering the page. A page with 200 clicks a month and 50 pageviews is being clicked by something that isn't reading. Clicks from your own testing count too; use a separate tracking ID when you test.
Keep reading
