Amazon affiliate site sales collapsed 2020-2026
Plain Reckoner

Amazon-affiliate businesses were 45% of verified sales in 2020. In 2026 they are 4%

551 words · about 3 min

Across 1,819 verified sales, sites that earn commission on Amazon links fell from 45% of everything the broker Empire Flippers sold in 2020 to 4% in 2026. Businesses that sell their own stock through Amazon — same marketplace, same broker, same buyers, but not living on search traffic — went from 35% to 53% over the same period.

How it was measured
We counted every sold listing in Empire Flippers' public record for 2019 to 2026, one year at a time. Anything counted twice was removed by its listing number. Each year was checked against the count the record itself reports, and all eight matched. A business can be tagged with more than one way of making money, and we count every tag rather than only businesses with a single one — the broker started applying several tags per listing during 2021, so counting single-tag businesses would measure that change in filing rather than the market.
Last checked
Listen to this Playing read from this page, nothing added
Watch it plot 37s · narrated · the same figures as above
Businesses tagged Amazon Associates, sold by the broker Empire Flippers, per year. Counted from their own public record. 2019: 117 2020: 134 2021: 124 2022: 118 2023: 54 2024: 22 2025: 15 2026: 5
Businesses tagged Amazon Associates, sold by the broker Empire Flippers, per year. Counted from their own public record.

We already published that sales of sites that live on display advertising fell 93% while Amazon stock businesses held. This is the same test run on the largest category of all, and the fall is steeper.

Empire Flippers is a broker: it sells websites, and it checks the traffic and revenue figures itself before a listing goes up. Its record of completed sales is public and needs no login. Counting it by how each business earned and by year gives a clean comparison — one broker, one pool of buyers, one period, and categories that differ mainly in whether the business depends on Google sending it people.

What the count says

Share of everything sold that year. A business can be tagged more than once, so the columns do not add up to 100.

20192020202120222023202420252026
Commission on Amazon links43%45%37%37%30%16%9%4%
Commission on other links18%16%21%28%25%13%13%4%
Display advertising25%18%33%44%43%25%18%8%
Selling stock through Amazon29%35%33%26%28%33%37%53%
Own online shop10%11%10%10%8%15%24%22%
businesses sold272297335315178141168113

Every category that lives on search traffic falls. Every category that does not, holds or grows. Selling stock through Amazon and running your own shop did not merely survive a shrinking market — they took it over, going from roughly a third of sales to three quarters between them.

The total is falling too: 335 businesses sold in 2021, 113 in 2026. So a rising share is not the same as rising sales. Amazon stock businesses went from 109 down to 60 in absolute terms. The point is the difference between the lines, not any single one: in a market where everything sold less, the categories that depend on search sold almost nothing.

What this does not show

The record returns what was sold, not what was put up for sale or what was offered and turned down. A category can vanish from this table because owners stopped selling, because buyers stopped bidding, or because the broker stopped taking the listings. We cannot separate those, and none of them is good news for this kind of business.

You can filter the record by sale date but the date is not returned with a listing, so each year here is a separate query against the broker's own filter rather than something we can work out again from a single download. We checked every year against the count the record reports for that window; all eight matched.

There is no comparable published series anywhere. Empire Flippers publishes total figures for marketing, not a series by category and year with a stated method, so this is measured here or not at all.

Tracked as AFFILIATE-EXIT-COLLAPSE, MONETISATION-TAG-DRIFT. If a number here looks wrong, tell us and we will check it against the source.

One email per finding

Get these as they land

One email per finding — the number, the method, and the date it was true. No schedule, because measurements do not arrive on one.

Data controller: Plain Reckoner. Contact contact@plainreckoner.com.