The Aperiam Podcast What LLMs Think Of Your Brand
Mar 11, 2026
Todd Sawicki, founder and CEO of Gumshoe.ai, helps brands see how AI models perceive and recommend them. He explains why traditional ranking fails with LLMs. He compares AI to a digital salesperson that brands must train. He outlines measuring AI visibility, persona-driven testing, and building authoritative content to influence recommendations.
AI Snips
Chapters
Transcript
Episode notes
LLMs Break Traditional Search Ranking
- AI search removes fixed ranking because LLMs are probabilistic and produce varied answers across repeated queries.
- Todd Sawicki explains models recommend 3–8 items from a large set and personalization plus stochasticity make visibility about varied recommendations, not rank.
Train AI Like A Store Salesperson
- Treat AI as a digital sales rep and invest in training it the way brands train in‑store salespeople.
- Todd recommends regular, authoritative content and training analogous to Nike presenting to store reps to influence the virtual salesperson.
Publish Authoritative Category Content
- Publish large volumes of authoritative, factual content across your category to become the "Wikipedia" for your product area.
- Todd advises bylines with bios/credentials and deep technical detail because LLMs favor credible source signals.