Google's Brendon Kraham argued on October 14, 2026 that generative AI in Search widens the path to purchase rather than narrowing it, pointing to Google-commissioned Ipsos research that found people using AI search visit 12 touchpoints across their buying process, nearly three times as many as in journeys without AI. The piece, headlined "Building brand authority in the Gemini era of Search," appeared on Think with Google, the company's publication for marketers, and is written for chief marketing officers.

Kraham is vice president of Search and Commerce for Global Ads Solutions at Google, a role on the advertising side of the company. He opens with a question he says marketing leaders put to him often: "In this new LLM world, how do we actually get discovered and stand out?" His answer is that the anxiety is misplaced. "A persistent myth suggests that new, intelligent Search experiences act as restrictive gatekeepers - narrowing choices, curating fewer responses, and shutting brands out of the conversation," he wrote. "The reality is the new era of Search doesn't shrink the consumer journey; it expands it."

The article links to a 34-page Google report dated February 2026, "The Rise of the Super-Empowered Consumer," which supplies the framing and much of the data. PPC Land covered that report in March. The October piece is also Kraham's second Think with Google essay on AI search this year. In June he told CMOs that good SEO is still all that is needed for AI.

In Short

A senior Google advertising executive says AI search makes shoppers look at more places before they buy, not fewer, and tells brand chiefs to publish specific, expert content and to judge results by leads and sales. The claim rests on a survey Google paid for, and it sits beside independent studies showing that AI summaries cut clicks to websites, so it matters to any business that depends on Google traffic. The measurement tools Kraham points to count impressions rather than clicks or sales, which leaves the outcomes he asks marketers to track hard to trace back to AI search.

Twelve touchpoints and a "persistent myth"

The central claim is about breadth. According to Kraham, AI search users "are exploring broader consideration sets, creating an unprecedented opportunity for brands to break through and be discovered." A second statistic is meant to show that the extra complexity does not slow anyone down: "On Google, 3 in 4 people report that AI features help them make faster, more confident decisions."

From those two numbers Kraham derives three priorities he says the CMOs he speaks with share: "Follow the super-empowered consumer," "Fuel brand authority with unique expertise," and "Focus on business outcomes."

The consumer he describes searches "with rich context," digs into niche topics and asks "highly specific questions that simply weren't possible a few years ago." Searches, he wrote, are now "multimodal, conversational, and deeply research driven." He quoted Daniel Shin Un Kang, head of organic and agentic search at Expedia Group, to make the point that the underlying task has not moved: "The fundamentals of delighting consumers haven't changed. What has changed is the medium of search, from keywords to conversations."

Google's position on how that medium works is stated plainly. "Google's generative experiences remain anchored in core ranking and quality systems to help organize information and make it accessible and useful," Kraham wrote, adding that "Google's core guidance remains durable and focused on user needs." The phrasing echoes the June essay, in which he argued that AI Mode and AI Overviews sit on top of Google's core ranking systems and that a separate playbook for AI search is unnecessary.

What the Ipsos numbers measure

Both statistics come from the same source, footnoted as "Google-commissioned Ipsos Global Consumer Journeys, Dec. 2025," an online survey of adults in 25 markets: Argentina, Australia, Brazil, Canada, Chile, Colombia, Germany, Spain, France, Indonesia, India, Italy, Japan, South Korea, Mexico, the Netherlands, Peru, the Philippines, Poland, Singapore, Sweden, Taiwan, Thailand, the United States and the United Kingdom. Figures are a global average of those countries.

The touchpoint comparison draws on two groups of online shoppers who had made a consumer goods purchase "requiring consideration" in the previous week. According to the footnote, 14,305 of them "used AI platforms or features in their journey" and 21,881 did not. By PPC Land's calculation, that puts AI users at about 40% of the 36,186 respondents in the comparison.

Several things the footnote does not say matter for how the number can be read. The group is defined by use of "AI platforms or features," which is broader than Google's own products, while the body text attributes the result to "people using AI search." The piece does not define a touchpoint, does not say how many of the 12 were Google or YouTube properties, and does not give the figure for non-AI journeys. "Nearly 3X" implies an average of a little over four touchpoints for shoppers who did not use AI, by PPC Land's arithmetic. Nor does the footnote say whether the comparison accounts for differences in category, price or market between the two groups. As published, the finding describes two different populations of shoppers; it does not show that AI caused anyone to visit more touchpoints.

The "3 in 4" figure has a narrower base. Footnote two puts the sample at 13,189 online shoppers who "used Google AI Overviews or AI Mode." Those are people already using Google's AI features, asked about Google's AI features. The February report draws on the same sample size and the same survey wave but reports two separate findings: 77% of AI Overviews and AI Mode users agreed they could make decisions faster because of them, and 75% agreed they could make more confident decisions. The October article folds the two into a single "3 in 4" statement about faster, more confident decisions.

Independent surveys point in different directions. In August, an RTB House study found that 42% of US shoppers said AI tools lengthen their purchase decisions, based on 1,840 respondents in four countries. The question differed from Google's, and so did the sample, but the result complicates the idea that AI makes decisions uniformly quicker. In May, Audits.com founder Carl Hendy argued that public claims of user preference for AI search trace back to Google itself.

Clicks are a different question

Kraham's "myth" concerns how many options a shopper weighs. The objection that has dominated the debate among publishers and search marketers concerns something else: whether those shoppers still click through to websites. The October piece does not address it.

The third-party record on clicks is extensive. A Pew Research Center analysis published in July 2025 found that users clicked a link on 8% of Google searches showing an AI summary, against 15% without one, across 68,879 searches; Google disputed the study, saying its method and choice of queries did not represent Search traffic. In February 2026, Ahrefs found that AI Overviews correlate with a 58% lower click-through rate for top-ranking pages, with position-one CTR falling from 0.073 to 0.016 on keywords that triggered an overview. A randomized field experiment with 1,065 desktop Chrome users, revised in June 2026, found outbound organic clicks fell 39.8% when an AI Overview appeared.

Google's executives have responded with statements about click quality and scale rather than with comparable published data. John Mueller said in July 2025 that clicks from AI Overviews were of higher quality, with users more likely to spend time on the site. Liz Reid, head of Google Search, said a month later that total organic click volume was relatively stable year over year. Nick Fox wrote in July 2026 that Google was sending billions of clicks to websites every week through AI features, without giving a denominator. The February report repeats the quality claim, stating that AI Overviews "drive higher quality clicks to the web," and cites Alphabet's second-quarter 2025 earnings as the source.

Both things can be true at once. A shopper can consult more sources over a longer journey while any individual website receives fewer clicks from a given search. Kraham's data speaks to the first; the click studies speak to the second. Neither settles the other.

Expertise over "commodity-style content"

The practical core of the piece is an argument about content. Competition in search, Kraham wrote, "was often about going after a limited set of broad search terms and topics with generic, commodity-style content." His example is a running store that, instead of a generic "Top 10 Things to Consider" list, produces "a deep-dive video analyzing exactly why a specific customer's running shoes collapsed."

He was explicit that brand-owned pages are only part of what Google's AI surfaces draw on. "We're designing our generative features to highlight the best of the web, including blogs, videos, and forum discussions, especially when it comes to product discovery," he wrote. "You can't control all of those external conversations." For what brands do control, the instruction is that websites, apps and product feeds "must be fast, rich, and user friendly."

The case study is IKEA. Linda Ha, the retailer's deputy CMO, is quoted describing kitchen planning: "planning a kitchen is such a complex journey, but now it's so easy with the new era of Search. We can show our experiences by telling them what complements their complete kitchen." She added that the company is "working closely across paid, owned, and earned [media] to ensure that we show up in the way that we want to." No figures accompany the IKEA example.

The October piece never uses the acronym GEO, which carried the June essay. The substance is the same: the playbook Google recommends for AI search is its existing guidance on helpful content. That position has critics. In May, iPullRank's Mike King argued that Google's guide to optimizing for AI features serves the platform over the open web, contending that retrieval mechanics differ from traditional ranking more than Google acknowledges.

Kraham's closing lines acknowledge the dependency that sits under the whole debate. "New, intelligent, and responsive Search features depend on the open web to provide depth, accuracy, and trust," he wrote. "The demand for credible, authoritative brand perspectives has never been higher."

"Unproven proxy metrics" and the reports Google offers

The third priority is measurement, and it contains the sharpest line in the piece. Kraham told marketers to keep "continuing to measure and scale plans based on real business goals like leads or sales, not unproven proxy metrics." He acknowledged the pressure behind the question: "We know there's a lot of pressure to demonstrate readiness for this new era and to show returns to finance partners, executive leadership, and boards."

Which metrics count as unproven? Kraham did not say. Measures of AI visibility, such as how often a brand is named or cited in generated answers, are among the newer numbers offered to brands for AI search, but the article does not name any metric, vendor or method.

Instead, it points to Google's own reporting. According to Kraham, Google has been "improving our tools" so that marketers can see how people find them through visual search queries, how their social and video content performs in Search, and "how your generative AI traffic scales with dedicated Search and Discover reports." The links in the article lead to three Search Central announcements from 2026.

The first, the Search generative AI performance reports, arrived in June for a subset of sites. PPC Land reported at launch that the reports cover AI Overviews, AI Mode and Discover, with impressions, pages, countries, devices and dates, but no click or query data. When the reports reached every site on August 31, the gap remained.

The second, platform properties for social and video accounts, let owners of Instagram, TikTok, X and YouTube accounts see how that content performs in Google Search. Google introduced them on July 7 and opened them to every account on July 29. They measure performance on Google Search only, not on the originating platform.

The third, a multimodal search type in Search Console, separates searches made with Lens, Circle to Search, image uploads and Chrome's "Search this image" from text searches. Google added it in September without a queries dimension.

None of the three measures leads or sales. The generative AI report, which the article describes in terms of "traffic," counts impressions. That does not make the reports useless, and Kraham presents them as a way to "better understand" visibility in generative features. But the outcomes he tells CMOs to prioritize are not what Google's AI-specific reporting currently shows, and the piece does not explain how a marketer would connect the two.

The February report behind the "super-empowered consumer"

The phrase Kraham builds his first priority on comes from the February 2026 report, subtitled "Marketers' guide to reaching consumers at the speed of AI" and marked "Vol. 1." Its foreword is signed by Vidhya Srinivasan, Google's vice president and general manager of Ads and Commerce, and it speaks about advertising more directly than Kraham's essay does. "In this new environment, relevant, helpful ads and products aren't interruptions - they're answers," she wrote. Later the report states: "The traditional marketing funnel is obsolete."

The report's usage figures are Google's own: AI Overviews "used by over 2 billion people," AI Mode at "over 75 million daily active users, as of the end of 2025," nearly one in six AI Mode queries "non-text, using voice or images," AI Mode queries "3x longer than traditional searches," and Google Lens at "over 25 billion visual searches every month." It adds that 71% of shoppers coming to Google Search "agree they are open to trying new brands or products," and that Google or YouTube was present in 82% of journeys where consumers said they discovered a new brand, product or retailer.

Two figures compare Google directly with OpenAI's product. Shoppers are "2.3 times more likely to say they use Google Search vs Chat GPT for purchase decisions," based on 37,261 online shoppers who use search engines or AI platforms. And 40% of consumers who use AI Mode for shopping "say they're using ChatGPT less now," based on 9,347 people who use both. Both come from the December 2025 Ipsos wave Google commissioned. An independent May 2026 survey of 1,295 US consumers by Smarty Marketing found Google was the starting point for 56.68% of shopping searches, against 7.26% for ChatGPT.

The headline advertising claim is that "Google Search drives an incremental ROAS of $6 for every dollar advertisers spend on paid campaigns globally." The footnote bases the iROAS figure on 143 Conversion Lift studies run between January 2021 and July 2024, a window that closes before AI Mode existed. The footnote still carries a drafting label, "Optional methodology line," before its description of Conversion Lift.

The report also previews commerce features Google had announced in January, when it launched the Universal Commerce Protocol alongside Business Agent and Direct Offers. According to the report, "later in 2026, advertisers can deliver tailored Direct Offers to shoppers who are ready to buy in AI Mode - like a special 20% off discount." Four advertiser case studies, Hilton, Rare Beauty, Aritzia and Toast, report results such as "a 7X return on ad spend from Search" for Rare Beauty and "a 42% lift in Q4 e-commerce net revenue" for Aritzia, without methodology.

Why the argument matters for marketers

The essay answers a question Kraham says he hears often from marketing leaders: whether AI answers erode the value of appearing in search. His answer is that the journey is bigger, the bar for content is higher, and the old fundamentals still hold. For advertisers, it is a case for continued investment in Search. For a publisher or a brand that depends on organic traffic, it offers no data on clicks.

The two sides of the argument rest on different kinds of evidence. The case for expansion rests on a survey Google commissioned, reported through footnotes that leave the central term undefined. The case for contraction rests on independent click studies that Google has disputed but not answered with its own published click data. The reporting Google offers in Search Console shows how often a site appears in AI features, not how often those appearances turn into visits, leads or sales.

That leaves a gap between the instruction and the available tools. Measuring "real business goals like leads or sales" is standard practice for paid campaigns, where Google controls conversion tracking. For unpaid visibility in AI Overviews and AI Mode, the link between an impression and a sale remains something marketers have to infer.

Timeline

Summary

Who: Brendon Kraham, vice president of Search and Commerce for Global Ads Solutions at Google, writing for chief marketing officers. The piece quotes Expedia Group's Daniel Shin Un Kang and IKEA's deputy CMO Linda Ha, and draws on Ipsos research Google commissioned and on a February 2026 report with a foreword by Vidhya Srinivasan.

What: An essay arguing that AI in Search expands the consumer journey rather than narrowing it, based on survey data that AI users visit 12 touchpoints, nearly three times as many as non-AI journeys, and that three in four users of Google's AI features decide faster and more confidently. It sets three priorities for brands: follow the "super-empowered consumer," build authority through specific expertise, and measure leads and sales rather than "unproven proxy metrics."

When: Published October 14, 2026. The underlying survey was fielded in December 2025; the companion report is dated February 2026.

Where: On Think with Google, Google's publication for marketers, with survey data averaged across 25 markets in the Americas, Europe and Asia-Pacific.

Why: Marketers face pressure from finance teams and boards to show what AI search does to their business, while independent studies show AI summaries cutting clicks to websites. Google's case for a wider journey rests on its own commissioned data, and the Search Console reports it offers for AI features count impressions, not the leads or sales Kraham tells CMOs to track.