Published on December 24, 2025
The old model of SEO in PR was a technical game of keywords and backlinks. The new model of Generative Engine Optimization (GEO) is driven by artificial intelligence capabilities and Google’s E-E-A-T grading system for Experience, Expertise, Authority and Trust. Brand and content marketers who understand GEO can provide the clear, expert point of view that AI-powered search favors.
Google’s AI Overviews in search engine page results built on Google past practices of building “knowledge graphs” of word relationships. AI opens up these indexes beyond keywords. For more complex queries, the Gemini-powered AI Mode puts research on automatic pilot, making a bigger and more ominous leap from AI-assisted search to search-assisted AI.
Both AI-curated search and search-enhanced AI are growing sources of website traffic. ChatGPT and Perplexity web referrals now rival the volume of traffic from search engines like Microsoft Bing, Duck Duck Go and Yahoo. Brand marketers can track this growth, in the acquisition tab of Google Analytics, as the number and proportion of AI referrers over time. Choose a time frame and comparison period, then sort user acquisition metrics by first-user source.
Google is making its biggest AI impact in three products:
The growth of AI in search likely is being underestimated, because AI bots do not transmit referral information consistently. When analytics programs cannot determine AI as a traffic source, content marketers will see an increase in “direct” traffic instead.
The rules of search optimization change when AI does all the searching. Keyword density isn’t dead, but its pulse is weak. Instead, GEO helps GenAI make semantic distinctions among topical keywords. The breakouts in Google Trends suggest how AI builds a knowledge graph of related concepts, filtering them by source, activity and region. This “Who, What and Where” analysis maps content to a query’s intent.
Backlinks aren’t exactly gone either, but the links themselves are optional. AI models are better at recording mentions across the internet–from name-dropping on message boards to citations in academic journals–and inferring the strength of each testimonial. GEO backs up copywriting claims with third-party endorsements and hard numbers on ROI metrics.
Open AI’s ChatGPT, Microsoft Copilot, Google AI Mode and other generative engines are thoroughly trained from web sources. But AI models do not “read” for inspiration; they parse for relationships between entities (for brand marketers, their clients), attributes (product features or personal expertise), and solutions (what they fix). Like product displays, sports scores and other past enhancements to search results, AI research draws from web source material but presents results directly rather than simply providing links.
By clicking through the web and briefing its users, Gen AI pulls them deep down the marketing funnel. They summarize a topic, then for purchase-intent queries, they go further to make product and service recommendations. As users shift from search queries to AI prompts, overall web views are likely to fall. Users most often will click directly from AI results to a recommended item description (often an affiliate link) or to a company contact or landing page. AI might also direct users to a LinkedIn profile or conference speaker’s bio to demonstrate a service provider’s authority.
The users of generative engines increasingly ask their chatbots to solve problems and make purchase decisions. In essence, AI does the shopping for them, researching prices and features across web sources. A unique selling proposition must be not only intuitive to website visitors, but also clearly stated, structured and coded for computers.
AI builds the same kind of matrix a project manager might keep on a spreadsheet to compare vendors. The models infer product and service capabilities from multiple providers and give their users “warm” introductions of companies that align with not only the user’s stated needs but also their implied interests.
Viewed this way, the opportunity to optimize for AI to present a differentiated solution with real-world results–a familiar branding assignment. Three steps can make this marketing message stick with both purchasers and their AI agents:
Content marketers should review their copy for concrete, AI-ready cues–and AI can help them do it. Here are prompts to test content for purpose-driven differentiation.
GEO work may prove harder to predict and attribute than algorithmic SEO. Still, it opens new opportunities for brand marketing specialists to raise the profile of companies and influencers than in ranked results. Winning this game takes a purpose-driven, inherently differentiated content strategy–the single best way to demonstrate value to both humans and AI crawlers.
As soon as you submit the form, we will send to you a link to download our report.