Marketing Speak® 529. What AI Really Thinks of Your Brand with Dan Petrovic
10 snips
Nov 26, 2025 
Dan Petrovic, a pioneering machine learning specialist and managing director of Dajon, dives into the evolution of SEO and AI's role in brand visibility. He argues that traditional link building is transitioning to content seeding with digital PR being critical for model training. Dan shares insights on measuring brand visibility through his tools, AI Rank and Treewalker, and emphasizes the importance of grounding in reducing model biases. He explores the future of websites and agent-based commerce, warning that without adapting to AI perceptions, brands risk being left behind.
AI Snips
Chapters
Books
Transcript
Episode notes
Optimize For Model Perception, Not Only Ranks
- AI visibility requires optimizing how models perceive your brand, not just traditional rankings.
- Dan Petrovic frames AI as a presentation layer that interprets retrieved search data and model memory.
Use Bi-Directional Probing To Measure Visibility
- Ask the model bi-directional questions: brand→entities and entity→brands to map associations.
- Track mention frequency and average position over time to create a brand visibility score.
Grounding Is Essential For Up-To-Date Answers
- Models can't hold fresh world knowledge forever, so they rely on search grounding for up-to-date facts.
- Grounding dramatically changes model outputs versus memory-only responses.
