NextFin News - Microsoft Advertising has officially pulled back the curtain on the opaque algorithms governing AI brand recommendations, releasing an updated comprehensive guide titled "Understanding AI search: A guide for modern marketers." The announcement, made on February 11, 2026, by Paul Longo, General Manager of AI in Ads at Microsoft, provides the industry’s most detailed blueprint to date on how Large Language Models (LLMs) select which brands to recommend and which to ignore. This release follows a period of intense industry speculation and a 10% slowdown in Microsoft’s search ad growth, prompting the tech giant to provide clearer pathways for brands to maintain visibility in an era where traditional search results are increasingly bypassed by conversational assistants.
The guide outlines a sophisticated three-stage brand surfacing process that moves beyond the binary of "paid versus organic." According to Longo, AI systems now assemble responses through a sequence of baseline understanding, grounded refinement, and precision signals. The process begins with the LLM’s pre-trained knowledge, which is then refined through Retrieval Augmented Generation (RAG) to verify facts against the live web. Finally, the system looks for "precision signals"—structured, first-party data provided directly by brands—to deliver definitive answers on pricing, availability, and specific product attributes. This technical transparency marks a departure from the "black box" approach of previous years, reflecting a market where U.S. President Trump’s administration has emphasized technological competitiveness and transparency in digital markets.
The shift from Search Engine Optimization (SEO) to Generative Engine Optimization (GEO) represents a fundamental change in the labor of marketing. Microsoft’s data suggests that the goal is no longer merely driving traffic, but establishing "contextual authority." For instance, while traditional SEO might focus on the phrase "waterproof rain jacket," GEO requires what Microsoft calls "Answer Engine Optimization" (AEO). This involves providing specific, machine-readable details such as "ventilated seams," "reflective piping," and "180-day return policies." The guide highlights that products with more comprehensive data fields rank higher by default, as AI reasoning engines prioritize completeness over clever copywriting.
The economic stakes of this transition are substantial. Microsoft’s advertising business recently crossed the $20 billion annual revenue threshold, fueled by the integration of Copilot across Bing and Edge. Internal performance metrics cited in the guide reveal that when Copilot participates in a user journey, click-through rates (CTR) double compared to traditional search ads. Furthermore, customer journeys involving AI assistants are 33% shorter, with a 53% increase in purchase probability. These figures explain why 54% of marketers now plan to increase investment in AI-driven media, surpassing planned growth for traditional search for the first time in history.
However, the industry remains divided over the terminology and tactics of this new era. While Microsoft promotes GEO and AEO as essential disciplines, others in the field have expressed skepticism. According to Mueller, a representative at Google, the aggressive promotion of new AI-specific acronyms can sometimes signal spam tactics rather than genuine innovation. Similarly, Fishkin, co-founder of SparkToro, has argued against the proliferation of new terms, suggesting that "Search Everywhere Optimization" better captures the reality of a fragmented discovery landscape. Despite these debates, the technical reality remains: AI systems like ChatGPT—which opened its platform to advertisers in January 2026—and Copilot are fundamentally changing how content is consumed by "chunking" information rather than reading full pages.
Looking forward, the rise of "agentic commerce" suggests that the next frontier will be optimizing for autonomous AI agents capable of executing purchases end-to-end. Microsoft’s recent partnerships with PayPal and Shopify to enable checkout directly within the Copilot interface indicate a future where the brand’s website serves primarily as a data source for AI agents rather than a destination for human shoppers. As McKinsey research projects that agentic systems could orchestrate up to $1 trillion in U.S. retail revenue by 2030, the ability to provide "precision signals" will likely become the primary differentiator between market leaders and those who disappear from the AI recommendation loop.
Explore more exclusive insights at nextfin.ai.
