Google Told Investors AI Answers Earn It Less. Here Is What That Means for Your Website.

Google earns less money from a search in AI Mode, where Google's own AI writes the answer instead of listing links. Google's chief business officer said that to investors on October 28, 2026. And Google's fix is more ads inside the AI answer, not more links out to your site. If search traffic matters to you, plan around that.

Both of those claims come from Google itself, on the same day.

What Google Actually Said on October 28, 2026

Alphabet reported its third quarter on October 28, 2026. Google Search and other revenue rose 12% to $63.35 billion (Alphabet Q3 2026 earnings release, October 28, 2026).

That is growth. It is just slower growth. The same line was 19% in the first quarter of 2026 (Alphabet Q1 2026 release, April 29, 2026) and 17% in the second (Alphabet Q2 2026 release, July 22, 2026).

A JP Morgan analyst asked about the slowdown on the call. Philipp Schindler, Google's chief business officer, gave three reasons. The third one is the story.

"It's worth noting that the strong growth in AI Mode queries had an impact because of the lower monetization we were seeing in AI Mode," he said (Alphabet Q3 2026 earnings call transcript, October 28, 2026).

Monetization is industry shorthand for money earned per search. In plain English: when someone uses AI Mode, Google earns less from that question than it would from an ordinary page of blue links.

Then the fix, in the same breath. "Our ad experiments have shown promising results and we're now scaling ads within both AI Overviews and AI Mode to more users and more queries; including, as I mentioned, international markets and multiturn ads in the U.S." (same source and date).

That was not just talk on a call. Google announced the rollout the same day. Ads in AI Mode are scaling to "Australia, Canada, India, Kenya, Malaysia, New Zealand, Nigeria, Pakistan, the Philippines, and Singapore." Multiturn ads are launching in the US, with Brazil, Mexico and the United Kingdom planned for the fourth quarter of 2026 (Google blog, October 28, 2026).

Read the shape of that. The question volume moved into AI answers, the money did not follow, and Google is closing the gap by selling ad space inside the answer.

Read the Whole Answer, Not Just That One Line

Fair is fair. AI Mode was Schindler's third reason, not his first.

His first: "there was FX. It was the single largest driver of Search deceleration relative to Q2." FX means currency exchange rates. His second was specific to a few industries. Retail advertisers in Asia-Pacific were hit by the EU dropping a tax exemption. Entertainment spending settled down after a World Cup quarter (same source and date).

Alphabet's chief financial officer, Anat Ashkenazi, answered a similar question on the same call and named currency and a tough comparison against late 2025. She did not mention AI Mode at all (same source and date).

So "AI answers are eating Google's ad business" is too strong. The accurate version is narrower and more useful. Google named AI Mode's weaker monetization as one of three drivers. It is also the only one of the three that is structural rather than temporary.

And this is not a story about people searching less. Sundar Pichai said on the same call that "our year-over-year query growth rate is the highest it's been in the last five years" (same source and date). People are asking Google more questions than ever. Google is earning less per question.

The Thing Google Tells Investors and the Thing It Tells You

Here is the part worth sitting with for a minute.

To investors, on October 28, 2026: AI Mode has "lower monetization."

To website owners, in its own site-owner documentation: clicks from a results page carrying an AI Overview are "higher quality (meaning, users are more likely to spend more time on the site)" (Google Search Central, checked November 5, 2026).

Both of those can be true at once. Fewer clicks that convert better is a coherent position. But notice who each sentence is written for, and notice that only one of them is backed by a number.

Google's fullest public defense on traffic is still from August 6, 2025. Liz Reid, who runs Search, wrote that "total organic click volume from Google Search to websites has been relatively stable year-over-year." She added that average click quality had gone up (Google blog, August 6, 2025).

Google published no sample, no date range, and no definition of "relatively stable" behind that. It is an assertion on a company blog, not a study. Fifteen months later, we could not find a 2026 Google post that restates it with a method attached.

Reid did concede the part that actually lands on one business. "The web is vast, and user trends are shifting traffic to different sites, resulting in decreased traffic to some sites and increased traffic to others" (same source and date).

Stable across the whole web is not stable for you.

What the Outside Measurements Actually Show

Three findings, oldest first, each with its limits attached.

Pew Research Center, July 22, 2025. Pew tracked the browsing of 900 US adults for the month of March 2025, covering 68,879 Google searches. When an AI summary appeared, people clicked a normal search result on 8% of visits. With no AI summary, 15%. Clicking a link inside the summary happened on 1% of those visits (Pew Research Center, July 22, 2025). The caveats matter. That is March 2025 data, from before AI Mode scaled. It measures the chance of a click per visit, not any site's total traffic. Pew makes no claim about cause.

A randomized experiment, published August 18, 2026. Researchers assigned 1,100 people to three search setups for a week each. Forcing people into AI Mode cut click-through to websites by 18.8 percentage points. Taking AI Overviews away raised it by 8.8 points (arXiv preprint, August 18, 2026). The authors list their own limits. The paper is a preprint and not peer reviewed. The sample skewed young and highly educated. And the AI Mode group was forced into full adoption rather than choosing it.

That same paper found ad clicks fell 42.7 points in the AI Mode group, "because AI Mode did not surface ads at the time of our experiment" (same source and date). That is Google's monetization problem, measured from the outside, months before Google named it on a call.

A regulator that has seen Google's internal files, January 28, 2026. On June 3, 2026 the UK Competition and Markets Authority ordered Google to give publishers real tools to keep their content out of AI features (CMA, June 3, 2026). The consultation document behind that order says something no outside study can: "Evidence from an internal Google experiment indicates that, for 'in slice' data (ie queries which would have triggered an AI Overview), clicks to publisher sites decrease by approximately [REDACTED]% when an AI Overview is shown on a search results page" (CMA consultation, January 28, 2026).

Google ran the trial itself. The direction is now on the public record. The size of it is blacked out, so nobody outside Google knows the number, and we are not going to guess one.

The same document records Google's reply, that AI Overviews "have sent higher quality traffic to websites." It also records what publishers told the regulator, which is the honest state of the art: they "cannot accurately assess whether referral traffic from search generative AI features is of higher value than that referred to publishers via traditional search links" (same source and date).

For scale, here is what is being argued about. Google said on June 3, 2026 that "AI Overviews now has over 2.5 billion monthly active users, while AI Mode has surpassed one billion monthly users" (Google blog, June 3, 2026).

One number from the publishing world, because it is the freshest available. Referrals from Google Search to hundreds of publishers fell 40% between July 2025 and July 2026, roughly twice the previous year's drop (Reuters Institute, October 13, 2026). Hold that one loosely. It covers news publishers rather than small businesses. And of the prior year's decline, the Reuters Institute said plainly that "it is not clear how much of this is down to AI overviews" (Reuters Institute, January 12, 2026).

There is one gap worth naming out loud. We could not find a single study measuring what AI answers do to a small business's search traffic. Everything above is publishers, or click-through rates in aggregate. Anyone who tells you they know your number is guessing.

Here Is Our Own Number

So here is ours, from the blog you are reading.

In the 28 days to November 3, 2026, this site drew 2,483 impressions in Google Search and 5 clicks. That is a click-through rate of 0.2%, at an average position of 24.7 (Ninth Degree Blog Ops, Search Console reading generated November 5, 2026).

Average position is the easy explanation, and for most of those impressions it is the right one. Position 24.7 is page three. Nobody clicks page three.

So look only at the pages that are not on page three. Seven of our pages averaged position 7.3 or better over those 28 days. Between them they drew 251 impressions and exactly 1 click. One of those pages sits at average position 5.0 with 33 impressions and no clicks at all (same source and date).

The trend is the part we did not expect. Compare the last two 28-day windows. The 28 days to October 18, 2026 gave us 693 impressions and 4 clicks. The 28 days to November 3, 2026 gave us 2,483 impressions and 5 clicks (same source and date).

Impressions more than tripled. Clicks did not move. For a longer baseline: the 90 days to September 21, 2026, our first reading, showed 257 impressions and 7 clicks.

Now the honest part, because we hold our own numbers to the same test as Google's. This is one small blog, live only since August 2026, and the click counts are low enough that a single reader's afternoon changes the percentage. It proves nothing on its own. What it is, is a measurement from inside the exact pipeline we sell to clients. And it points the same way every outside study above points.

What to Do This Week

Open Search Console and check for the same gap. Look at your pages that rank in the top ten, and compare their impressions to their clicks. Google says AI-feature traffic shows up in the Performance report under the "Web" search type (Google Search Central, checked November 5, 2026). If you rank well and get nothing, that is the pattern on this page. Worth knowing before you spend another month chasing rankings.

Decide whether you want to be in AI answers at all, before someone asks you. Google shipped a Search Console toggle for this on June 3, 2026. It is blunt. "Sites that opt out will not receive traffic or impressions from our generative AI features" (Google blog, June 3, 2026). There is no setting for cite me but send me the click. Google said it was starting with a subset of site owners in the UK, so check whether you have it yet.

If you run Google Ads, know you are already in the AI answers. Google's own help page says "Existing Search, Shopping, and Performance Max campaigns are automatically eligible to serve ads within AI Overviews and AI Mode" (Google Ads Help, checked November 5, 2026). That is worth a look at your placement reports.

Write the page that answers the question instead of teasing it. The AI answer is going to take the summary either way. The click you still get comes from somebody who wants the detail. Prices, specifics, your actual process. Pichai said AI Mode queries run "on average three times longer than traditional searches" (Alphabet Q3 2026 earnings call transcript, October 28, 2026). Longer questions are more specific questions.

One Thing to Skip

Skip buying an AI Mode optimization checklist.

Google's own documentation is unusually plain here. A page needs to be "indexed and eligible to be shown in Google Search with a snippet." Beyond that, there are "no additional technical requirements." The same page adds that "Indexing and serving isn't guaranteed" (Google Search Central, checked November 5, 2026).

There is no special markup to add. Anyone selling you one in November 2026 is selling you something Google says does not exist.

This post was drafted by Ninth Degree's agent pipeline and edited by a human before publishing. Two sources were deliberately left out. One was a New York Times report on AI Overview accuracy that we could not load at the source. The other was a second randomized experiment whose working paper would not open, and where two secondhand write-ups disagreed on the headline figure. Working out which search traffic is still worth chasing, and which has quietly stopped arriving, is the call we make for clients on every engagement.