AI summaries are reshaping search traffic
Brian opened with a report from Vox describing a significant shift in how search works. AI-generated summaries, now appearing at the top of Google search results, are reducing the number of clicks to publisher websites. The declines range from 15% to 64% depending on the industry and query type. As one example, the New York Times saw organic referrals drop from 44% to 36% in April 2025.
The underlying dynamic is straightforward: when an AI answers the question directly, users have less reason to click through to a website. Andy noted that Google is essentially competing with itself, since its traditional business model depends on driving traffic to advertisers and publishers.
Google is like eating itself. Their value to media companies and advertisers and retailers and B2B companies is driving traffic to websites. What if that traffic never goes to a website? Their business model is upside down.
Brian Beck, Master B2B
Google’s response signals the severity
Andy pointed to recent news that Google is offering voluntary exit programs to employees in its Knowledge and Information unit, the group that houses search, ads, and commerce. The company had already laid off around 12,000 people in 2023, according to public reporting. The hosts read this as evidence that Google recognizes how serious the threat is to its core business.
What GEO means for marketers
The main topic of the episode was what this shift means for B2B marketers. Traditional SEO involved optimizing pages for keywords, building backlinks, and ensuring fast load times and good user experience. GEO, or generative engine optimization, is the emerging discipline of getting content surfaced by AI answer engines like ChatGPT, Google Gemini, Perplexity, and others.
The hosts noted that when they asked practitioners in recent roundtables how to optimize for answer engines, nobody had a clear answer. Some agencies claim to have a formula, but the specifics remain vague.
Everybody said it is a priority, and when we asked how do you do it, nobody knew how to do it.
Andy Hoar, Master B2B
IDC’s predictions for the next few years
Brian cited IDC analyst Heather Hershey, who spoke at the Commerce Tools Elevate conference. Her projection was that by 2028, gen AI enabled search platforms will become the primary user interface for shopping and product discovery. She called this a profound shift in corporate marketing strategy. IDC also projected that traditional search volume could drop by 25% by 2026 as buyers turn to AI-powered tools.
Andy suggested these numbers might be conservative, given how quickly personal adoption of AI tools has spread among professionals.
What works in a GEO world
The hosts outlined several principles that appear to matter for GEO. Semantic relevance, meaning the topical depth and context of content, matters more than keyword density. IDC uses the framework EAT, standing for Expertise, Authoritativeness, and Trustworthiness, to describe what answer engines value.
Content that is structured and scannable, with clear facts like price, availability, and specifications, is more useful to AI systems than emotionally-driven marketing copy. Third-party validation, such as backlinks from authoritative sources, continues to matter because it signals credibility.
The content explosion and its limits
Andy predicted that companies will respond by producing large volumes of AI-generated content, hoping to be cited by answer engines. He compared this to the keyword stuffing tactics of early SEO and suggested it will become a similar arms race. However, he also expected answer engines to get better at identifying and deprioritizing low-quality content.
The hosts agreed that authoritative third-party content, such as research from recognized analysts or coverage from established publications, is likely to retain value because it carries more credibility than self-promotional material.
Authentication and product liability
Brian shared a quote from Jason Hein raising a concern specific to B2B: the chain of custody for product information. If AI hallucinates or misrepresents product specifications, the consequences in B2B can be serious, potentially affecting safety, compliance, and liability.
In B2B the consequences can be pretty dire, and so there is probably going to be an industry that is going to arise around validating whether things are real or not.
Andy Hoar, Master B2B
Andy noted that this is part of a broader societal challenge around distinguishing authentic content from AI-generated material, but the stakes are higher in B2B where incorrect information can lead to equipment failure or safety issues.
Practitioners expect rapid change
The hosts shared results from a LinkedIn poll asking when GEO would become more important than traditional SEO. Of those who responded, 76% said within one year, 21% said three years, and only 3% said five years. Nobody said never.
Andy attributed the speed of this shift to personal adoption. Unlike some enterprise technology trends that take years to filter down, most white-collar professionals are already using AI tools daily and seeing the change in their own behavior. That firsthand experience is accelerating expectations about how quickly the market will shift.