Duane Forrester has spent 30 years in search and digital strategy, and his resume reads like a timeline of the industry itself. He built Bing Webmaster Tools during his time as Senior Product Manager at Microsoft and was part of the team that co-launched Schema.org, the structured data standard now used across the web to help machines understand content. His career includes leadership roles at Bruce Clay Inc. and Yext.
He now runs Unbound Answers, where he created CitationIQ, a tool for measuring how often and where brands appear in AI-generated responses. His book "The Machine Layer" outlines frameworks for becoming visible and trusted in AI-powered search.
On the podcast, Duane introduced the concept of "machine comfort bias," the tendency of LLMs to favor sources that provide complete, trustworthy information without requiring additional verification. When an AI system needs to answer a question, it gravitates toward content that feels reliable on its own, content that doesn't force the model to cross-reference or hedge. He argued that content structure matters differently for AI than it does for humans: rather than reformatting pages into 300-word chunks, the key is placing facts, figures, and bullet points at the top of sections where LLMs encounter them first.
His "Ironman vs Superman" analogy captures his broader view on AI: these systems amplify your existing skills rather than replacing them. Someone who already understands search strategy gains leverage from AI tools. Someone hoping AI will substitute for that understanding is using hope as a strategy. And hope, as Duane puts it, is not a strategy. Trust has become the new algorithm.
LLMs evaluate trust in chunks, not pages. Duane Forrester explains the machine comfort bias reshaping discovery.
