Google AI Mode ads are getting a bigger push, and on Oct. 28 Google's own executives said the searches behind them earn the company less than traditional ones. Search Engine Journal covered it on Nov. 9 in a column by Greg Jarboe. The details matter to any BC business that pays for Google clicks.
Two statements came out the same day. Shivani Mohan, who leads data science and user research for Search at Google, said on the Ads Decoded podcast that people now arrive earlier in the buying process and ask long, detailed questions, many of a kind nobody typed before. That afternoon Philipp Schindler, Alphabet's chief business officer, told analysts that the strong growth in AI Mode queries had an impact on Search revenue because of the lower monetization Google sees there.
As the column points out, most Google Ads accounts are still judged at the bottom of the funnel, on the phone call or the booking form, while the searches are arriving at the top. That gap is the planning problem for 2027.
Why AI Mode pays Google less
The quarter's numbers show the squeeze. Google Search & other revenue was $63.35 billion in Q3 2026, up 12% from a year earlier. Paid clicks on that line grew 8%, slower than the 11% growth over the first nine months of the year. Cost per click rose 4%. Both measures cover Google Search & other as a whole, which also includes Gmail, Maps, Play and partner sites, so they are wider than Search alone. Sundar Pichai said query growth was the highest in five years, so people are searching more while paid clicks grow slower.
Schindler named currency as the biggest driver of the slowdown against Q2, then added AI Mode. Other factors came up too, including retail advertising challenges tied to the EU's removal of its de minimis exemption. His optimistic reading is that many AI Mode queries sit high in the funnel and most are incremental, which gives Google more ways to earn later.
Google has not explained the mechanics, and Alphabet's filing does not split out AI Mode paid clicks or cost per click. So nobody outside Google can say how much less each AI Mode click earns, or whether AI Mode pays an advertiser back at the same rate as a regular search click.
The one performance figure comes from Google's own U.S. data for Aug. 20 to Sept. 3, 2026: click-through rates were already similar between sponsored and organic product listings inside AI Mode. That is two weeks of data, measured by Google, and it compares ads with organic listings rather than with regular Shopping ads. Treat it as a claim until your own account confirms it.
Why the mobile comparison falls apart
Pichai compared this moment to the shift to mobile, when Google focused on the user experience first and on monetization later. The column checks that against Reuters: in Q3 2016, paid clicks rose 33% and cost per click fell 11%, as searches moved to smaller screens with cheaper ads.
This time cost per click across Google Search & other is still up 4%, and the gap shows up as queries growing faster than paid clicks. A BC advertiser gets new top-of-funnel inventory at the same auction prices and carries the job of proving it pays off. No discount comes with the move up the funnel.
One limit on all of this: multi-turn ads, which let ads appear in follow-up questions, were announced for the U.S. On Oct. 28 Google also named Canada among 10 additional countries where it is bringing ads in AI Mode. Check what your own account shows before you plan around it.
What a BC advertiser can do before Black Friday
The column turns the podcast advice into four jobs. Black Friday falls on Nov. 27 this year, so the order matters.
1. Check your match types
Per Google's help documentation as the column cites it, ads inside AI Overviews can serve only on broad match or keywordless targeting: AI Max for Search, Performance Max, Shopping and Dynamic Search Ads. Exact match can still serve above or below the overview. If your non-brand campaigns run only exact and phrase match, you are outside the placements Google is scaling.
Keep exact match where you need control. Give broad match or AI Max a defined slice of non-brand budget, with Smart Bidding and clean conversion data behind it. Automation learns from whatever you count as a win, and we covered what happens when that count is wrong in Why Bad Conversion Data Now Destroys Your Ad Strategy.
2. Group search terms by length
Export the search terms report and sort terms into one to three words, four to seven, and eight or more. Compare conversion rate, cost per customer and how often each group appears earlier in the path to a sale.
Say long queries convert badly on last click but show up as early touches. Then you hold the upper-funnel evidence finance will ask for. If they show up nowhere, broad match or AI Max is spending on discovery you cannot justify yet, and negatives and brand controls tighten it. A blended average hides that difference.
3. Fix the page the ad lands on
A person who typed a 25-word question about a specific situation will not convert on a generic category page. With AI Max final URL expansion, Google picks the landing page from your site, so a thin page can get chosen for a nuanced question.
Audit your highest-spend landing pages for the details a follow-up question would ask about: comparisons, sizing, pricing, use cases. Much of that work sits with whoever runs your content and website, so bring them in alongside the person running ads.
4. Online stores: fill the new Merchant Center fields
Google added six optional product fields in May, among them question_and_answer and related_product. They do not affect product approval, and Google recommends supplying them through a supplemental data source so the main feed stays untouched. Start with your top sellers.
Why the organic side matters too
Mohan described what shoppers do when they get close to a decision. When people compare two products, they want more than a summary. They look for reviews, podcasts and videos from people in situations like theirs.
Pages that answer those follow-up questions lift an ad landing page, and they also give regular results and AI assistants something to quote. How AI Recommends Local Businesses covers how those assistants pick which local companies to name.
NG checks every month what ChatGPT, Gemini, Perplexity and Claude say about a client's business and which sites they cite. On one BC industrial cleaning account, 76% of AI answers to buyer questions cited the site on Sep 24, 2026, up from 47% on Jul 7 (AI answer check, 3 assistants, 15 questions). That is one account and one check, and it shows what is measurable today.
Content work takes time. In our SEO and AEO service, which starts at $800 a month on the Starter plan, the first moves come in 6 to 10 weeks and real traffic in 3 to 6 months. A page started now gets that clock running.
My view
Searchers are clearly moving up the funnel, and I do not doubt Google on that. I doubt the expectation that advertisers will pay for the move before they see results. Until Google publishes AI Mode ad performance that holds up the way Mohan says its user statistics do, I would run AI Mode and AI Overview reach as a test line with its own budget and its own success measure. Keep it out of the blended number that decides whether Google Ads survives the next review, and wait for proof that it pays.
On the NG side, Google Ads management starts at $800 a month (Starter) and $1,200 a month (Growth), with ad spend paid directly to Google, so a test slice comes out of money you already control. If you want to see how your account would split between the bottom of the funnel and an upper-funnel test, request a quote.