When asked whether to own source material on a brand website or third-party consensus, Drosos emphasizes that owning consensus is far more powerful. Similar to social media, consumers and LLMs naturally distrust direct brand claims. Because LLMs over-index on independent citations from platforms like Reddit and YouTube as validation signals, marketers should focus content strategies on influencing the third-party voices that machines and consumers trust.
Mark Drosos warns that tools capturing broad prompt surveys provide noisy, generic signals that can misguide marketers. Winning in AEO and agent selectability requires completely breaking down content by exact buyer personas, core topics, and stages of the buying journey. Marketers must evaluate prompts against those granular segments to understand how LLM embeddings interpret their content, rather than chasing vanity citation spikes from low-quality survey data.
Mark Drosos notes that AEO is mirroring the evolution of social media. Initially, marketers tried to game LLMs by churning out content just to win citations and mentions. However, as LLMs increasingly pull from diverse third-party sources like YouTube and LinkedIn, brands must shift from superficial content creation toward earning genuine participation by clearly solving customer problems and demonstrating authentic business value.
Agent selectability will eventually enable AI to choose and buy products for consumers, but AI cannot inherently judge personal fit, like trying on jeans. Because buyers do not always know what constraints to prompt, Drosos argues brands must connect first-party customer knowledge with product data to guide the journey. That ensures agents surface recommendations tailored to unique consumer preferences rather than just whatever generic web content exists.
Rachel Truair points out that Demandbase's most surfaced content in LLMs often comes from technical help documentation rather than marketing materials. While this documentation outlines product capabilities, it was originally built for technical operators rather than prospective buyers. As a result, B2B organizations need to deeply re-examine their existing technical content through the lens of prospect personas and stages of the buying journey.
As AI agents increasingly evaluate products and manage buying workflows, Drosos and Truair suggest marketing teams treat agents as distinct personas on the B2B buying committee. Because agents will become more verticalized and specialized, companies must allocate specific budget and strategic focus toward understanding what algorithmic agents want, tailoring content to satisfy their evaluation criteria alongside human buyers.
Drawing a direct comparison to the early mobile app craze where businesses built apps they could not maintain, Drosos and Truair caution that current widespread agent creation is unsustainable. While companies are rushing to build proprietary agents, Drosos predicts the market will mature around specialized, vertical players that monetize specific domains, leading companies to augment third-party specialized tools with their own proprietary data instead.