Product5 publishers3 min readPublished
Google's AI Overview pilot pays small publishers about a tenth of a percent of their ad revenue
Google's pilot paying about 100 publishers for AI Overview use returns small and mid-sized members roughly a tenth of a percent of their ad revenue. For most members the check cannot stand in for lost clicks, so the case for joining rests on the usage data and on what a site covers.
The Product Desk · Product desk
What happened
- Some smaller sites in the pilot have earned less than $1,000 over several months, according to The Information's reporting.
- One publisher that joined early is on track for more than $1 million a year, a big chunk of its revenue, according to Ars Technica.
- Several larger publishers have reportedly declined to join, hoping their refusal pushes Google toward a more generous deal.
- Google had long resisted paying publishers directly, arguing that sites were compensated with search traffic for letting its bots crawl them.
Compiled by The Product DeskSomething wrong?How this is made
Why it matters
- contradiction Tiny checks for most members and a $1 million pace for one sit in the same report, so averages about the pilot say little about what any single site will earn from it.
- precedent Paying per contribution, even in a test, moves Google off its traffic-for-crawling position, so any wider program will be negotiated over the rate.
- constraint Google has not explained how it prices a contribution, so a publisher can only guess from early patterns what a given page will earn.
A small publisher in Google's pilot gets two things back. One is a count of how often AI Overviews drew on its content [8]. The other is a payment, and at the reported rate it comes to about a dollar for every $1,000 the site earns from ads [1].
Google's pitch, as Ars Technica describes it, is that a site gets paid when its content "materially contributes" to a Gemini-powered result such as an AI Overview [9]. Members measure that payment against their traffic. 9to5Google reports traffic is down almost universally, though Google insists AI is not to blame [12]. For the smallest sites, Ars Technica wrote that the payments "won't come close to offsetting declining traffic" [11]. The payment figures support that verdict. The reports do not say how much ad revenue any member has lost, so the shortfall cannot be put in dollars.
The top of the range is harder to fit into the same complaint. The early member's pace of more than $1 million a year works out to about $2,740 a day [2], more than the smallest sites have collected in several months [2]. A more recent member has earned $50,000 to $60,000, a small slice of its business [5]. The top earner's early start is one candidate explanation [4]. Topic is another: 9to5Google reports that less widely covered topics with strong reader interest, anime and gaming among them, currently earn more [7].
Larger publishers staying out are betting on Google's own needs. According to 9to5Google, The Information reports that Google "needs these web publishers to stay active" for its models and answers to stay current and accurate [13]. Small sites that join now get the usage count and accept a rate the holdouts are trying to raise [6].
I think most small and mid-sized sites should join for the count and book the payment at zero. The tradeoff is that the count shows how often Google uses a site's pages [8] and brings none of the clicks those pages used to earn. Two facts about your own site place you in the grid: whether your topics are thinly covered with strong demand, and whether you are large enough that staying out is a credible position.
- Small site, niche topic: join. This is the profile the early reporting says the pilot pays best [7]. - Small site, widely covered topic: join for the usage count and plan as if the payment rounds to nothing. - Large publisher, widely covered topic: the holdouts' position [6]. If the rate resembles what small members report, declining costs almost nothing [3]. - Large publisher, niche topic: the hard cell. The payment could be real money, and every large site that signs lowers the pressure on Google to raise the rate.
What to watch
- Whether Google expands the pilot beyond its roughly 100 members, which 9to5Google says remains unclear.
- Whether the larger publishers that declined are offered a better rate, the first sign that refusing to join works as leverage.
- Whether Google discloses how it prices a contribution, so sites can test the anime-and-gaming pattern against their own topics.