Halfway through September, the same question keeps surfacing in content team meetings, now that AI answers are taking clicks away, how should the content spend be accounted for. Right at that moment, Search Engine Journal and Digiday both reported on September 14 that Google is testing a paid pilot, when site content contributes significantly to AI answers, Google will pay the site, and the settlement entry point sits inside Search Console, a panel named AI contribution.

Once the news spread, three common readings followed. One treats it as a new revenue source and starts doing the math. One treats it as a formal Google commitment. Another pushes further and treats it as the start of a long-term commercial relationship between Google and content owners. All three run ahead of the reported terms. According to both outlets, this is an invite-only, small-scope pilot, and the only characterization Google confirmed to Digiday is that it is an early-stage learning pilot.

What makes this small item worth its own piece is not the money but the measurement. For the first time, Google has turned "did AI use your content" into a dial with numbers on it, sitting inside Search Console, the tool SEOs open every day. On the content investment ledger, the blank cell between being clicked and being adopted has started to fill in.

The Panel and Its Boundaries, First

Put the two reports together and the confirmable facts read like this. The pilot is invite-only, and only selected sites can see the AI contribution panel; the panel shows a monthly amount plus some historical data; participants can leave at any time from settings. Google describes the effort as exploring how to reward high-quality content on top of the traffic and tools it already provides; per Digiday, the invitations go well beyond news organizations, a wider net than Google's earlier collaborations with news publishers.

The billing runs around one word, significant contribution. Per the help text Digiday obtained, payment requires that content significantly contributed during the generation phase of an AI answer, across three surfaces, the Gemini app, AI Overviews, and AI Mode. One detail deserves underlining, anything that happens after the answer exists does not count, content being linked or used to confirm facts does not qualify. What gets measured is adoption, with the answer itself as the reference point. Impressions and link counts do not enter this ledger.

The boundaries are just as important. As of now, Google has not disclosed the eligibility criteria, the participant list, the calculation method, or the rates; the panel gives you a number without the derivation. Per Digiday, one person familiar with the program called it quite a black box. One more easily missed point, earnings and payments are two different things, and the help text notes that the amount actually paid can differ from the displayed figure due to taxes and local payment thresholds.

There is one more item worth keeping for export teams. The panel's neighbor, the generative AI performance report, opened to all sites at the end of August (per SEJ) and shows exposure inside AI features without click data. One is available to everyone, the other is invite-only, and reading them as the same thing leads to wrong conclusions.

From Clicked to Adopted, One More Cell on the Ledger

For the past two years, content teams have lived with two sets of numbers, clicks from traditional search and AI exposure from the generative AI report. AI exposure has no clicks, and traditional clicks keep getting intercepted by AI answers. Beyond those two, a place to price adoption has been missing. The pilot fills that cell.

For content investment decisions, it carries three implications.

The first lands on KPIs. Content performance used to be read through traffic, rankings, and conversions; being adopted by AI now has its own observation surface. It does not move in the same direction as traffic, pages that get adopted the most are often the ones with the steepest click declines, because users get the answer without clicking through. The old ledger recorded a zero for those pages, the new ledger starts recording positives, and together the picture of content value gets fuller.

The second lands on how value gets confirmed. If the direction continues, part of the value of quality original content will be confirmed through adoption rather than depending on visits alone. The change favors long-consideration categories, professional content, and clearly structured Q&A pages, since those are what AI answers absorb.

The third lands on language. Once AI adoption becomes a term people can discuss, content teams, SEO teams, and leadership talking about the same content line gain a dimension that fits into one shared table. What to watch, what to report, and what to do next quarter all become easier to move on once the vocabulary is shared.

Whether or Not the Pilot Expands, Start with Three Things

First, build a 28-day baseline with the generative AI report. The report is free and available to everyone, so start by mapping your exposure inside AI features. Take the most recent 28 days as the baseline, then pull the same view each month, watching the trend rather than any single number.

Second, find the pages that get adopted heavily while clicks stay flat. The method is to cross two datasets, pages with high exposure in the generative AI report, and pages whose search clicks are stalled or declining; if you use third-party tools to monitor AI citations, add high-citation pages to the pile. The resulting list is the set of pages being absorbed by AI while the old ledger shows nothing, review it monthly. They may not produce revenue soon, but they deserve the top spot on the watchlist.

Third, add citable structure to your content acceptance criteria. At the operational level, that means a handful of items, conclusion sentences that can be extracted without losing meaning, key facts with visible sources, update dates and authorship on display, and one clear question per piece. Write these four into the acceptance checklist of your content briefs, alongside word counts and keyword coverage. The reason for now is that adoption only registers when contribution can be recognized, and clearly structured content fares better under that kind of judgment; at the same time, admit the judgment is a black box, so these four are the moves that stay right no matter how the rules change. We covered citation mechanics in our GEO pieces; here the focus is the acceptance step.

Uncertainty and What Not to Do

Start with the uncertainty, there is more of it than certainty. The pilot is invite-only and small; the criteria, participants, and calculation method are undisclosed; even participants do not know where the number comes from. The reporting holds two more voices worth noting, one from participants who feel it is better to be inside the room early than to stand outside waiting for the old traffic to return; another from critics who call the terms opaque and the early returns small next to ad revenue, with some reading it as an early hedge at the industry level. These are observations from media and practitioners, not official terms, recorded here as reference for decisions, not as the basis for them.

Then a few things not to do right now.

Do not put this number into any revenue forecast. The scope, rates, and eligibility are undisclosed, the calculation is a black box, and budgeting against it is borrowing against air.

Do not change your content roadmap for possible payouts. Writing filler because it looks like it might get adopted inverts the logic; adoption follows being needed, and reversing the order means nothing.

Do not make moves on team size or budget structure. Structural decisions need three things, a wider pilot, disclosed terms, and at least two or three quarters of data; until then, the right position is the watchlist.

There is one thing worth doing, track two signals. Whether invitations widen, and whether the calculation gets more transparent. It is also worth reading Google's June policy note linked in the references, where the language on grounding partnerships lives.