Google AI Overviews and AI Mode for commercial queries: what to do now
AI answers now sit above some of the queries that feed your pipeline, and fewer people click through. Which commercial queries are exposed, how to be cited, how to measure it and what to do with the visitors who still arrive.
- Commercial queries trigger AI Overviews far less often than informational ones. Ahrefs found an AI Overview on 4.3% of commercial and 2.1% of transactional keywords, against 21.4% of informational ones. The exception is the comparison query.
- Where a summary appears, clicks fall for organic and paid. In Seer Interactive’s 2026 data, organic click-through was 2.36% with an AI Overview against 3.82% without, and paid was 16.2% against 21.8%.
- Being cited is ordinary SEO done well. Google’s guide says optimizing for generative AI search is “still SEO”, with no special files or markup, and it adds Merchant Center feeds and Business Profiles for product and local answers.
- Measure impressions, then work on the visit. Use the Generative AI performance report in Search Console for visibility, and treat the landing experience as the lever for the fewer visitors who still click.
For commercial queries, do three things now. Find which of your comparison, “best” and pricing queries show an AI Overview, because most product and transactional queries still do not. Get your pages cited on the ones that do, which Google says is still SEO. Then put the remaining effort into the landing experience, because fewer people click and the ones who do arrive further into the decision.
The live question for a marketing leader is no longer how AI Overviews pick sources. It is which revenue queries are exposed, what the exposure costs in organic and paid clicks, and what a visit is worth once Google or AI Mode has already done the comparing. This article answers those in order, with dated sources, and says where the advice does not apply.
The conversion event used in the worked reasoning is the lead (a demo request or enquiry). Where a study or a customer result measures purchases instead, it says so.
Which commercial queries get an AI answer
Most commercial queries still show no AI Overview. The exposed ones are the long, comparative, question-shaped queries buyers type just before they shortlist. Ahrefs’ analysis of 146 million results pages (published November 2025, data from September 2025) found an AI Overview on 20.5% of all results pages. By intent, the rate was 21.4% for informational keywords, 4.3% for commercial, 2.1% for transactional and 0.9% for navigational. Only 7.9% of local searches and 3.2% of keywords in the shopping category triggered one.
Seer Interactive’s 2026 update (April 2026, 53 brands, 5.47 million queries, January 2025 to February 2026) reports higher levels and the same order: informational queries showed an AI Overview 36% of the time, commercial queries 8% and transactional queries 5%. The two studies classify intent differently, so read the ranking, not the exact rates.
The exception matters more than the average. In the Seer data, comparison queries had a 95.4% AI Overview rate: “nearly every ‘X vs Y’ informational query in this study has an AI Overview”. Wording drives exposure too. In the Ahrefs data, 57.9% of question queries and 46% of queries of seven words or more triggered one, and non-branded keywords did so at 24.9% against 13.1% for branded ones. A buyer who types “best payroll software for a 50 person company” is asking a long, non-branded question, whatever intent label a tool gives it.
| Query trait | AI Overview rate | Source |
|---|---|---|
| Comparison (“X vs Y”) | 95.4% | Seer Interactive, April 2026 |
| Informational intent | 36% (Seer), 21.4% (Ahrefs) | Both |
| Commercial intent | 8% (Seer), 4.3% (Ahrefs) | Both |
| Transactional intent | 5% (Seer), 2.1% (Ahrefs) | Both |
| Local searches | 7.9% | Ahrefs, November 2025 |
| Shopping category | 3.2% | Ahrefs, November 2025 |
| Navigational intent | 0.9% | Ahrefs, November 2025 |
AI Mode is a different case, because the user chooses it and every query gets an AI answer. Google said in May 2026 that AI Mode had “surpassed a billion monthly active users globally” and that “the average AI Mode search is triple the length of a traditional Search query”. Google does not publish how many of those searches are commercial, so treat AI Mode as a growing surface you measure on your own site, not one you can size from outside.
What happens to organic and paid clicks
Where an AI summary appears, fewer people click anything, and that includes your ads. Pew Research Center (July 2025, browsing data from 900 US adults in March 2025) found that users “clicked on a traditional search result link in 8% of all visits” when an AI summary appeared, against 15% without one. A click on a link inside the summary itself “occurred in just 1% of all visits”, and the browsing session ended on 26% of pages with an AI summary against 16% without.
Seer Interactive has tracked the same effect in client data twice. Its November 2025 study (3,119 search terms, 42 organizations) saw organic click-through on queries with an AI Overview fall from 1.76% in June 2024 to 0.61% in September 2025, and paid click-through fall from 19.70% to 6.34%. The larger 2026 update shows a partial recovery: organic click-through on those queries climbed “from a floor of 1.3% in December 2025 to 2.4% in February 2026”. It is still below queries without an AI Overview (3.82%), and paid click-through shows the same gap, 16.2% against 21.8%. The two Seer studies use different samples, so their levels are not comparable with each other.
The number to plan around is the difference between being cited and not. Seer reports that “being cited in the AIO delivers +120% more organic clicks per impression versus when you are not cited”, while a cited brand “still underperforms No AIO by -38%”. On the paid side, cited brands held a click-through of 13.99% to 17.95% through 2025, an advantage of more than four percentage points over uncited brands in every month.
Three cautions before you put this in a forecast. Seer states that it “cannot claim causation”, because stronger brands are more likely to be cited in the first place. Its per-impression figures are for informational queries. And Ahrefs’ click study (April 2025, updated June 2026), which measured a 34.5% lower click-through for the top position on AI Overview keywords, also used informational keywords. Nobody has published the same cut for commercial queries alone.
For paid search, Google is also changing the inventory. In February 2026 it said it was “testing a new ad format” in AI Mode to showcase retailers, “clearly marked as sponsored”. Whether that restores paid volume on your queries is something only your account data will show.
This is the zero-click pattern applied to the queries that used to feed your pipeline. How to optimize for zero-click searches covers the wider version, and whether ChatGPT is taking traffic from Google covers the question of where the searches are going.
How to be cited on commercial queries
Google’s answer is that there is no separate discipline: a page is eligible for AI Overviews and AI Mode when it is indexed and can be shown with a snippet, and it is chosen when it answers one of the searches Google runs behind the scenes. Google’s guide to optimizing for generative AI features (published May 2026, last updated 10 July 2026) puts it this way: “From Google Search’s perspective, optimizing for generative AI search is optimizing for the search experience, and thus still SEO.”
“Best”, “X vs Y”, pricing, product or service.
Google fans the query out into related searches and writes one answer.
Links to indexed pages, plus product listings and local businesses.
A visitor who has already compared. Lead or no lead is decided here.
The mechanism that changes the work is query fan-out, which the guide defines as “a set of concurrent, related queries generated by the model to request more information and fetch additional relevant search results”. One commercial query becomes several: price, alternatives, reviews, fit for a use case. Your content competes for each of those separately, which has two consequences that still hold from the original version of this article. The answer is assembled from passages, so a page that answers one sub-question precisely can be cited without ranking first for the head term. And coverage counts, so a cluster of pages that answers the whole set of sub-questions is cited more often than one page aimed at one keyword.
For commercial queries, that translates into five moves:
- Stay eligible. The guide says a page “must be indexed and eligible to be shown in Google Search with a snippet”. Check that comparison, pricing and category pages are crawlable, not behind a login, and not restricted with snippet controls you set years ago for another reason.
- Answer the sub-questions buyers ask. State pricing logic, limits, integrations, who the product suits and who it does not, each in a passage that makes sense on its own. Lead with the answer.
- Publish what a summary cannot copy from ten other sites. The guide asks for “non-commodity content” that provides “unique expert or experienced takes”. For a vendor that means your own data, tested comparisons and real customer results.
- Feed the product and local surfaces. “Generative AI responses can include product listings, product information, and information about local businesses”, and the guide names Merchant Center feeds and Google Business Profiles as the way to be visible there.
- Earn the third-party mentions the answer draws on. “Best” answers lean on lists and reviews you do not own. Whether backlinks still matter for LLMs and how to show up in an AI answer you never ranked for cover that side.
Common mistake: spending the budget on AI-specific files and formats. Google’s guide says “you don’t need to create new machine readable files, AI text files, markup, or Markdown to appear in Google Search”, that structured data “isn’t required for generative AI search”, and that “there’s no requirement to break your content into tiny pieces”. Keep the structured data you already use for rich results. Do not expect it to buy a citation.
On agents, the guide is measured. It describes AI agents as “autonomous systems that can perform tasks on behalf of people, such as booking a reservation or comparing product specifications”, and suggests exploring agentic experiences “if this is something that’s relevant to your business and you have extra time”. For most B2B sites that is a watch item, not this quarter’s project. Online stores are closer to it: Google said in February 2026 that checkout inside AI Mode was rolling out for US shoppers with selected retailers.
The guide speaks for Google only. For how this differs from other engines and from the surrounding vocabulary, see GEO vs SEO and the difference between SEO, GEO, AEO and LLM optimization.
How to measure it
Measure visibility with impressions in Search Console, then judge the result on leads from the affected queries, not on click-through. Google’s guide tells site owners to “use the Generative AI performance report in Search Console” to “get an idea of how people are discovering your content”. Google announced the Search Generative AI performance reports on its Search Central blog in June 2026. According to PPC Land’s report of the announcement (3 June 2026), they cover AI Overviews and AI Mode, show impressions by page, country, device and date, and were rolling out to a subset of websites first. The announcement did not mention click data, and the AI impressions are also counted in the overall performance report.
That leaves a working method in four steps:
- List the revenue queries. Comparison, “best”, pricing and product or service queries that lead to demo requests or enquiries today. Twenty to fifty is enough.
- Record exposure. For each, note whether an AI Overview appears and whether your domain is cited. Repeat monthly, because the answer changes from run to run.
- Read impressions and clicks together. Flat impressions with falling clicks on those queries usually means you are being shown next to, or inside, an AI answer. It is a change in searcher behavior, not a sign the page got worse.
- Follow the visit to the lead. Report sessions, lead rate and leads for the landing pages behind those queries. That is the only number that tells you what the exposure cost.
The example uses Seer’s click volumes because they are the only published split of cited against uncited. With a 2.0% lead rate held constant, one million impressions produce 670 leads with no AI Overview, 189 when the brand is not cited and 415 when it is. Citation closes 226 of the 481 lost leads. The remaining 255 cannot be recovered on the results page at all. They can be recovered on the landing page: at 20,743 clicks, a lead rate of 3.2% returns 664 leads, close to the original 670.
The report shows that a page appeared in an AI answer, not what the answer said about you or whether a competitor was recommended instead. The lead rates in the example are assumptions, and Seer’s click volumes describe informational queries across 53 brands. Replace every number with your own before using it in a plan.
What to do for each query type
Sort your commercial queries by whether a summary appears, then do different work on each group. The matrix below combines the trigger data from the first section with the moves from the third. It is a starting order, not a rule: a query’s wording decides exposure more than its label.
Two readings follow from it. The first is that comparison content is where exposure is highest and where the buyer is closest to a shortlist, so it deserves the first round of work. The second is that the lower rows are mostly ordinary search and conversion work, which is easy to forget when AI search fills the agenda.
The remaining 255 cannot be recovered on the results page at all.
Brand and navigational queries: an AI Overview appeared on 0.9% of navigational keywords in the Ahrefs data, so someone searching your name still gets your site. Transactional queries: 2.1% in the same data, so product and booking pages are won with classic SEO, feeds and ads. Local services: 7.9% of local searches, where the Business Profile matters more than any content change. If most of your leads come from these three groups, AI Overview work is a low priority and the SEO and CRO guide is the better use of the quarter. The same is true of pure informational content with no commercial follow-up: the summary absorbs the answer and there is no lead to protect.
What to do with the visitors who still click
The visitor who clicks through an AI answer has usually compared already. Treat the page they land on as the last step of a decision, not the first. Seer’s data makes the volume side plain: even a cited brand gets 38% fewer organic clicks per impression than it would with no AI Overview. Whatever lead volume you keep has to come from a higher lead rate on fewer visits.
How much higher is not settled, and it is honest to say so. No published study isolates the conversion rate of clicks from AI Overviews or AI Mode. The nearest evidence is traffic referred by AI assistants, and it points both ways. Ruler Analytics’ 2026 benchmarks (May 2026, conversion defined as a qualified lead or sale) put AI referral at 5.8% against 4.9% for organic search. Contentsquare’s 2026 benchmarks (March 2026, 99 billion sessions) put AI-referred traffic at 1.3%, low against other channels. The difference between those two numbers is mostly what the visitor found on arrival.
One customer result shows the upside in e-commerce, where the conversion event is a purchase, not a lead. Tinggly, an experience gifts store, converts AI traffic at 2.4% against a 0.3% site average.
For a lead-generation page, the practical changes are small and specific. Repeat the claim that earned the citation near the top, so the visitor can verify it. Put proof next to it. Offer one next step sized to a person who has compared options but not yet spoken to anyone. How to optimize a landing page for higher conversions without hurting UX goes through the page itself, and what to do with visitors who are not ready to book a call covers the ones who arrive earlier in the decision.
How Pathmonk fits
Pathmonk works on the two parts of the flow you control: being cited, and converting the visit. Its GEO automation and SEO agent draft answer-first content from your first-party data and Company DNA, and the Search Console connector keeps the affected queries in the same view as the visits they send.
On the page, Pathmonk reads first-party, cookieless in-session behavior in real time, predicts the buying stage of each visitor (awareness, consideration or decision) and serves a stage-matched microexperience without changing the page content or needing dev work. A decision-stage arrival from a comparison query sees proof and a path to the demo request. Lift is measured against a control group.
Across Pathmonk AI Search customers (September 2026), sessions from AI assistants rose 28% on average over the first three months, measured in Pathmonk analytics, and AI answer citations rose 122% on average. These figures cover AI assistants and AI answers broadly, not Google’s AI Overviews alone. Individual results from the same page:
| Customer | Sector | Result | Period |
|---|---|---|---|
| Tinggly | Experience gifts e-commerce | +121% AI-referred sessions | Against the previous 23 days |
| Yapa Explorers | Travel | +21% traffic from AI search | Months 1 to 3 |
| Care4Dental | Dental | 4 of 5 buyer prompts name the brand | Month 3 |
| Allure | Medical clinic | +148% AI answer citations | Months 1 to 3 |
Key takeaways
- Most commercial queries still show no AI Overview: 4.3% of commercial and 2.1% of transactional keywords in the Ahrefs data, against 21.4% of informational ones.
- Comparison queries are the exception, with a 95.4% AI Overview rate in Seer Interactive’s 2026 data. Long, question-shaped, non-branded queries are also more exposed.
- Where a summary appears, clicks fall for organic and paid alike: 2.36% against 3.82% organic, and 16.2% against 21.8% paid (Seer, February 2026).
- Being cited is associated with 120% more organic clicks per impression than not being cited, and still 38% fewer than with no AI Overview. Seer does not claim causation.
- Google’s guide says optimizing for generative AI search is still SEO. No special files, markup or chunking are required. Merchant Center feeds and Business Profiles cover product and local answers.
- Measure visibility with the Generative AI performance report in Search Console, and judge the outcome on leads from the affected queries.
- Brand, navigational, transactional and local queries rarely show an AI Overview. If they bring most of your leads, this is a low priority.
- Citation recovers part of the clicks. The rest of the lost leads can only be recovered on the landing page.
FAQs on AI Overviews and AI Mode
Do AI Overviews appear on commercial queries?
Less often than on informational ones. Ahrefs found an AI Overview on 4.3% of commercial keywords and 2.1% of transactional keywords in September 2025, against 21.4% of informational keywords. Seer Interactive reported 8%, 5% and 36% in its 2026 update. Comparison queries are the exception, with a 95.4% AI Overview rate in the Seer data.
Do AI Overviews use the same ranking signals as classic search?
Largely, yes. Google’s guide says optimizing for generative AI search is still SEO, and a page must be indexed and eligible to be shown with a snippet. What differs is selection: Google runs several related searches behind one query and builds the answer from passages, so a page that answers one sub-question precisely can be cited without ranking first for the main term.
What is query fan-out and why does it matter?
Google defines it as a set of concurrent, related queries generated by the model to fetch additional relevant search results. One commercial query becomes several, such as price, alternatives and reviews. You compete for each one separately, which is why a cluster of pages covering the whole set is cited more often than a single keyword-targeted page.
Why did my click-through rate drop while impressions stayed flat?
That pattern usually means your page is being shown inside or next to an AI Overview. Pew Research Center found that people clicked a traditional result in 8% of visits with an AI summary against 15% without. It reflects a change in searcher behavior, not a decline in the page.
Does blocking snippets or AI crawlers hurt visibility in AI Overviews?
Removing snippets does. Google’s guide says a page must be indexed and eligible to be shown in Google Search with a snippet to appear in generative AI features. A page that cannot be shown with a snippet gives up that eligibility, and with it the higher click-through that cited brands see in Seer Interactive’s data.
Does structured data get you cited in AI Overviews?
No. Google’s guide says structured data is not required for generative AI search and that no special markup is needed. Keep the structured data you use for rich results, but do not expect it to earn a citation on its own.
How do you measure visibility in AI Overviews and AI Mode?
Google’s guide points to the Generative AI performance report in Search Console, announced in June 2026. It reports impressions for pages shown in AI Overviews and AI Mode. Pair it with a monthly check of your main commercial queries and with leads from the landing pages behind them.
When is AI Overview optimization a waste of effort?
When most of your leads come from brand, navigational, transactional or local queries, which rarely show an AI Overview, and on pure informational content with no commercial follow-up, where the summary absorbs the answer. In those cases classic SEO, feeds, Business Profile and the landing experience deserve the effort first.
- Google Search Central, guide to optimizing for generative AI features on Google Search (published May 2026, last updated 10 July 2026) (developers.google.com)
- Google, How AI Mode is changing and expanding the way people search (19 May 2026) (blog.google)
- Google, digital advertising and commerce in 2026, Vidhya Srinivasan (11 February 2026) (blog.google)
- PPC Land, report on Google’s Search Generative AI performance reports in Search Console (3 June 2026) (ppc.land)
- Ahrefs, What triggers AI Overviews? 146 million results pages analyzed (10 November 2025, data September 2025) (ahrefs.com)
- Ahrefs, AI Overviews reduce clicks (17 April 2025, updated 5 June 2026) (ahrefs.com)
- Seer Interactive, AIO impact on Google CTR: 2026 update (24 April 2026) (seerinteractive.com)
- Seer Interactive, AIO impact on Google CTR: September 2025 update (4 November 2025) (seerinteractive.com)
- Pew Research Center, clicks when an AI summary appears (22 July 2025) (pewresearch.org)
- Ruler Analytics, conversion rate benchmarks 2026 (26 May 2026) (ruleranalytics.com)
- Contentsquare, 2026 Digital Experience Benchmarks, conversions (9 March 2026) (contentsquare.com)