Lutz Finger | From SEO to GEO: How AI is Changing How Car Buyers Find You
About this episode
Sanjeev Yajnik welcomes AI technologist Loot’s finger to unpack how AI search is reshaping dealership discovery. The conversation contrasts SEO with generative search optimization (GEO), showing how people search differently and how chatbots and new algorithms change planning and recommendations. A data scientist shares a story of scraping dealer sites to choose and buy quickly. The episode then pivots to practical dealer steps: make interactions public, rely on contextual, authentic content, anonymize data, and structure it in Q&A so AI agents can answer with current pricing and inventory.
Sanjiv and Lutz discuss how authenticity, thinking creatively and the tulips outside your dealership plant the right seeds to grow your SEO (Search Engine Optimization) and GEO (Generative Engine Optimization) performance.
Lutz Finger is an AI and data science leader with over 20 years of experience building real-world products at Google, LinkedIn and Snap. He currently teaches at Cornell University, focusing on how AI can improve decision-making and the customer experience.
ABOUT THREE KEY INSIGHTS The auto industry is being reimagined in real time. To win in today’s landscape, dealers need actionable insights and high-performance tools to take the lead. Hosted by the President of Capital One Auto, Sanjiv Yajnik, this series convenes elite dealers and industry innovators to decode the shifts shaping the future. Known for his "Customer-first" philosophy, Sanjiv’s perspective is rooted in the belief that when dealers succeed, the entire industry thrives. His conversations prioritize transparency, pulling back the curtain on the backend infrastructure that powers modern customer confidence. Now’s the time to take the lead.
The views and opinions expressed by guests on this podcast are solely their own and do not necessarily reflect the official policy or position of Capital One Auto, Capital One, N.A., or its affiliates. This content is provided for informational and educational purposes only and is not intended to serve as financial, legal, or tax advice. Viewers should consult their own independent professionals regarding their specific individual or business needs.
CONNECT WITH SANJIV YAJNIK LinkedIn: https://www.linkedin.com/in/sanjivyajnik/ X: https://x.com/sanjiv_yajnik
Honda Odyssey
"I want to buy a car. I type in Honda Odyssey and then I get a list of Honda Odyssey's nearby. ..."
The Honda Odyssey is a minivan, which is a family-focused vehicle with room for passengers. In the podcast, it’s used as an example of what happens when you search for a specific model online and see nearby options. The key point is that the search results can quickly show cars you can buy locally.
The Honda Odyssey is a minivan designed for families, with space and seating meant for everyday passenger use. It’s brought up in the context of searching online—typing “Honda Odyssey” and getting a list of nearby vehicles—highlighting how popular models are surfaced by digital shopping tools. That makes it relevant to discussions about improving the car-buying experience.
Ford F150
"They go and talk to a chatbot and the chatbot completes the sentence with it's the F-150, because everybody is talking to it. Meaning, how do now your dealers who have a personalized experience with their customers,"
The Ford F-150 is a large pickup truck made by Ford. People often search for it when they want to buy one, so a chatbot may suggest nearby F-150 listings. It’s being used as an example of how online tools respond to common car searches.
The Ford F-150 is a full-size pickup truck known for being widely sold and frequently discussed, including in dealer and customer experiences. In a podcast context about chatbots and personalized shopping, it’s mentioned because people commonly search for the F-150 and expect the system to quickly surface nearby options. That makes it a good example of how popular models drive what customers see in digital tools.
GEO
"Episode: Lutz Finger | From SEO to GEO: How AI is Changing How Car Buyers Find You ... be authentic. It has to be powerful. It has to be unique."
GEO is about making it easy for AI tools to find and use your dealership info when someone asks a question. It’s like SEO, but for AI answers instead of just web search rankings.
GEO (generative engine optimization) is the idea of structuring your dealership’s information so AI systems can retrieve it and generate answers for shoppers. Instead of optimizing only for search results, you optimize for how models respond to questions about your inventory, services, and policies.
de-anonymize data
"Now you could try to de-anonymize data. It's a very tricky thing to do. So I always would anonymize and ask just to be on the safe side."
De-anonymizing is when data that was supposed to be anonymous can be traced back to a real person. The speaker is saying you should be careful because that can create privacy problems.
De-anonymizing data means trying to re-identify information that was stripped of personal identifiers. In dealership AI use-cases, it matters because customer interactions can be sensitive, and privacy laws may treat re-identification as a risk.
Q&A format
"So the best way you package is up, package up your content is actually in a Q&A format because the tool is doing auto-completion. If somebody asks, how many wheels does a car have?"
Q&A format means writing content like: question first, then a clear answer. The point is that AI can more reliably copy the right answer when someone asks the same question.
Packaging dealership content in a Q&A format makes it easier for a generative model to match a user’s question to a direct answer. The transcript argues that this reduces the model’s “work” (extra reasoning) and increases the chance it outputs the intended response.
auto-completion
"So the best way you package is up, package up your content is actually in a Q&A format because the tool is doing auto-completion."
Auto-completion is when an AI tries to finish what you’re asking or writing by predicting the next words. Here, it’s why clear question-and-answer pages can work well.
Auto-completion is the behavior where a model predicts likely next text based on the prompt. In the transcript, it’s used to justify why Q&A content can be “picked up” quickly and answered consistently.
cognitive load
"One is read the sentence, the car has four wheels, and then think about it, I should answer four. That's a cognitive load for model costs money."
Cognitive load here means how much extra “thinking” the AI has to do. The speaker’s point is that direct questions with direct answers are easier for the AI to handle.
Cognitive load (in this context) is the amount of extra reasoning effort the model must do to produce an answer. The speaker contrasts answering a simple fact embedded in a sentence versus answering a direct question, claiming the latter reduces the model’s reasoning burden.
AI-generated data
"Do not use SLOP, AI-generated data, and package it up in Q&A because that is what the model will use most."
AI-generated data is information written by an AI instead of coming from real dealership experience. The speaker is saying you should prefer your own real customer-based info.
AI-generated data refers to content produced by AI rather than sourced from the dealership’s own real customer interactions or records. The transcript warns against relying on it because models will preferentially use whatever content is most available and well-structured.
Request an Explanation
Heard something you'd like explained? We'll add it to this episode.
Sign in to request explanations for terms you heard.
Want to learn more?
Browse our glossary for plain-English explanations of automotive terms, jargon, and concepts.
Help improve this episode
See something that's not quite right? Our annotations are AI-generated and can sometimes miss the mark. Click the flag icon on any annotation to suggest a correction.