For about twenty years the affiliate publishing business ran on a single, reliable exchange. A site did the work of testing or comparing products, that work earned a high Google ranking, the ranking delivered a click, and the click carried a tracked link that paid a commission when the reader bought. Every part of that chain depended on the next. Break the link between ranking and click and the whole business stops.
That link is now broken. Google's AI Overviews answer the buying question on the results page, in Google's own words, drawing on the very review and comparison content that used to earn the click. The reader gets the recommendation without visiting the site that produced it. For a publisher whose entire revenue model is search traffic to product content, this is not a slow headwind. It is the removal of the thing the business was built on.
This is a publishing-business story, not a tactics story. The question is not how to get cited inside an AI answer. It is which affiliate content formats still have a business once Google answers the question itself, and what the surviving publishers have actually changed.
How the model worked, and why it was fragile
Affiliate publishing has a clean origin. A merchant cannot see which website convinced a buyer, so it pays a tracked link per sale instead. Amazon opened its Associates program in 1996, the open affiliate program became the default, and a generation of publishers built businesses on one move: rank a product page in Google, collect commissions on the clicks.
The model worked because of an assumption that held for two decades. People researched purchases by searching, reading several pages, and clicking. Someone typing "best robot vacuum" would land on a comparison, read it, and click through to buy. The publisher earned its cut for the genuine work of helping that person decide.
The fragility was always there, and it was specific. The publisher never owned the reader. It rented attention from Google one query at a time, and that attention could only be monetized if the reader left Google and arrived on the publisher's page. Remove the reason to leave Google and there is nothing left to rent. That is precisely what AI Overviews and the broader AI Mode do: they keep the reader on the results page and hand over the answer there.
What the traffic data actually shows
The decline is documented, and it is not subtle. Across publishers generally, Google search traffic fell by roughly a third in the year to November 2025, with the drop steeper in the US, down 38 percent, than in Europe, down 17 percent, according to a Press Gazette analysis of industry data. That is the baseline, and affiliate review content sits well below it.
When an AI Overview appears above the organic results, the page that used to rank first loses most of its traffic. An Authoritas study found sites in Google's top position losing up to 79 percent of referral clicks once an AI Overview sat above them, as Press Gazette reported. Ranking number one, the prize the whole affiliate SEO playbook chases, now wins a fraction of what it used to.
The starkest single case is Wirecutter, the product-review operation owned by The New York Times. Depending on the tracking tool, it lost between 65 and 72 percent of its Google search visibility against May 2025 levels, with two sharp drops in mid-2025, according to an analysis by SEO consultant Glenn Gabe. Gabe names no single cause: a reviews-system update, Google treating the Wirecutter directory as separate from the parent domain, and a robots.txt change all plausibly played a part, so this is not a clean AI Overviews story. The wider lesson still stands. When a brand with deep testing and a strong reputation can lose two-thirds of its search visibility in a few months, no affiliate publisher should assume its rankings are safe.
The revenue effect follows the traffic. Affiliate consultants told a Press Gazette event that AI Overviews had cut traffic to buyers' guides and review content by as much as 50 percent, with affiliate revenue at some publishers down 20 to 40 percent, Press Gazette reported. Close to 7 in 10 affiliate publishers say they are worried that Google's changes will harm their business, per eMarketer. The worry is rational.
There is a structural twist that makes the squeeze worse. AI Overviews are built partly from the review and comparison content publishers wrote. Every well-structured comparison page is both the product the publisher sells and raw material for the system that removes the need to visit it.
Which formats are exposed, and which survive
The useful move is to stop talking about affiliate content as one thing, because AI Overviews do not hit every format equally. The dividing line is simple. If the value of a page can be fully captured in a summary, AI will capture it and the click disappears. If the value cannot survive a summary, the page still has a reason to be visited.
The most exposed format is the specs-and-shortlist comparison. A page titled "best budget laptops" that lists five models with prices, key features, and a one-line verdict for each is, structurally, already an AI Overview: a list of entities with attributes. A retrieval system reads that format perfectly and rebuilds it as an answer in seconds, so the reader gets the shortlist on the results page and never needs the source. These pages were the core of affiliate SEO for a decade because they matched high-intent buying queries. They are now the most directly substitutable thing a publisher owns.
The self-promotional "best of" listicle is exposed for a second reason on top of summarization. After Google's December 2025 core update, SEO analysts tracked steep ranking losses at brands publishing review-style pages that crowned their own product as the top pick, often with the current year in the title and little real updating underneath. SaaS and business-software brands using that tactic recorded organic visibility drops of roughly 29 to 49 percent in January 2026, as Search Engine Land reported on Lily Ray's analysis. The signal Google is acting on, a biased list dressed up as an independent review, describes a large share of thin affiliate listicles too. A format that is both easy to summarize and under separate algorithmic pressure has the weakest position in affiliate publishing right now.
The thin, untested roundup is exposed worst of all. AI made it nearly free to generate a passable product roundup, so the supply exploded while quality fell. Such a page holds no first-hand information an AI cannot assemble elsewhere, so summarization costs it everything. The SEO analyst Glen Allsopp found that across 10,000 product-review search results, only 4 of the top 100 ranking domains were independent sites, a finding Ahrefs cited in its case against starting a pure affiliate site. An independent thin-content publisher is squeezed from above by media conglomerates and from below by infinite cheap pages, and now summarized away by AI on top.
What survives is content whose value does not fit in a summary. Three kinds hold up.
First, content built on first-hand testing and original data. A review that reports what actually happened when the writer ran the vacuum for three months, with measurements, photographs, and failures, contains information that exists nowhere else. An AI can state the conclusion, but a reader weighing a real purchase often wants the proof, the edge cases, and the photos, so a reason to click through remains.
Second, long-tail, implementation-specific content. The query "best air fryer" is answered fine by an AI Overview. The query "non-toxic air fryer that fits a small apartment kitchen and has a viewing window" is specific enough that a generic summary disappoints and the reader keeps looking. Publishers are deliberately moving down the long tail toward narrow questions where conversion is stronger and summarization is weaker.
Third, formats that are awkward to summarize at all. Video leads this group: it answers a buying question in a way a text snippet cannot easily reproduce, and AI systems increasingly point to it rather than replace it. Interactive tools, configurators, and calculators give a reader a personalized result a static summary cannot, so the page itself is the product. None of these is immune, but all are harder to compress into an answer box than a specs table.
What surviving publishers have actually changed
The encouraging part of the data is that some affiliate publishers are losing traffic heavily and growing commerce revenue at the same time. How they do it is the real lesson, and it is consistent across the cases.
The first move is to optimize for conversion instead of volume. Apartment Therapy reported commerce revenue up 10 percent year over year even as traffic fell about 20 percent, with conversions up 34 percent and its Amazon clickthrough rate doubling, after it introduced dynamic pricing and refreshed inventory on older pages, Digiday reported. Forbes told the same outlet that site traffic was down 37 percent year over year in the first quarter of 2026, while revenue at its Forbes Vetted commerce arm rose 2 percent because the Vetted conversion rate had doubled. Visitors who arrive after an AI Overview has done the early filtering tend to be further along and readier to buy, and a smaller, higher-intent audience can hold revenue flat if every page is built to convert it.
The second move is to leave low-margin organic affiliate deals for higher-margin direct ones. Trusted Media Brands described shifting away from lower-margin affiliate deals driven by organic search toward higher-margin direct and performance deals across web, social, and email, per the same Digiday reporting. A direct relationship with a brand does not depend on a Google ranking surviving the next AI expansion.
The third move is the structural one: stop renting readers from Google and own the relationship. The publishers holding up are rebuilding around channels they control. Newsletters, logged-in products, communities on Reddit and Discord, podcasts, and video all reach an audience without a search query in the middle, as eMarketer describes the diversification underway. Forbes paired its commerce arm with a dedicated audience and loyalty team using a customer data platform. The shorthand for this is a shift from traffic economics to relationship economics, and it is the shift the old model never had to make.
Wirecutter shows where this leads. The New York Times put it behind a metered paywall in 2021, years before the AI Overviews crisis, as its history records. A review operation that once lived purely on affiliate commissions now runs on subscriptions plus affiliate revenue. That is the template the category is being forced toward: affiliate income as one revenue line among several, not the whole business.
Where this goes
Two things will shape the next few years. The first is that the search surface keeps moving against summarizable affiliate content. Google has been extending AI Mode further into shopping, with commerce tools and in-experience checkout for some retailers, as Digital Commerce 360 reported. The more the buying journey completes inside Google's own AI experience, the less room there is for a publisher whose only asset is a ranked comparison page.
The second is that the floor under good content is real but lower than it was. AI shopping answers still draw heavily on independent review and comparison content, because that is where genuine product testing lives. But being the source is not the same as being paid. A citation inside an answer box the reader never clicks does not pay a commission. The publishers that survive will be the ones that turned that authority into something they own, a subscriber, a newsletter reader, a direct brand deal, before the citation became the only thing the AI left them.
The honest summary is that the era of building an affiliate site as a pure Google-traffic play is over. What is left is a smaller, harder, more durable business: produce review content good enough that an AI cannot fully replace it, convert the high-intent readers who still arrive, and own enough of the audience that the next algorithm change is an inconvenience rather than an extinction event. The publishers already doing this are not thriving. They are surviving, which in this category now counts as a win.
Council summary
This post argues that AI Overviews break the ranking-to-click link affiliate publishing was built on, and that a page's format now decides its fate: anything a summary fully captures loses its click, while first-hand testing, narrow long-tail content, and video hold up. The council verified the load-bearing figures against primary reporting, including the roughly one-third global Google traffic drop, the Authoritas finding of up to 79 percent click loss for top-ranked pages under an AI Overview, Wirecutter's 65 to 72 percent visibility decline, the 20 to 40 percent affiliate revenue hit, and the Apartment Therapy and Forbes commerce numbers. Two attributions were corrected: the Wirecutter section now reflects that Glenn Gabe names no single cause and does not call it an AI Overviews story, and the listicle-suppression statistic was re-sourced from Search Engine Land's coverage of Lily Ray's analysis, with its scope narrowed to the SaaS and business-software brands the data covers. The practitioner takeaway is concrete: treat affiliate income as one revenue line among several, shift toward content a summary cannot replace, and build owned channels before a citation is all the AI leaves you.
Comments