Sep 14, 2025 · 1h 11m · lennys-podcast
The ultimate guide to AEO: How to get ChatGPT to recommend your product | Ethan Smith (Graphite)
0:00 / 0:00
▶ Watch on YouTube →
gold bands on the timeline = statements, start to end. Hover to read, click to jump. CC turns on captions
Host Lenny Rachitsky interviews Graphite CEO Ethan Smith to deliver an actionable masterclass on Answer Engine Optimization (AEO), detailing how companies can systematically optimize for LLM citations, RAG search retrieval, and high-converting conversational queries across platforms like ChatGPT and Perplexity.
How this conversation actually went
Every chapter scored 0–10 on four independent dynamics. Hover any point for the reasoning behind the score. Lenny holds 25.7% of the talking time here. How this is scored →
Lenny as informed peer 2.3 Guest teaching 6.5 Guest disagreement 1.4 Lenny pushing back 1.1
speaking balance: gold is Lenny, purple is the guest (3 minute bins)
Sharpest disagreement ▶ 47:18 Debunking Industry Panics Around Google Dying
Ethan firmly rejects the pervasive industry narrative that Google search is declining, pointing out that historical panics over TikTok and Instagram search never actually shrunk Google's overall share.
Hardest push from Lenny ▶ 8:56 Host Challenges Advice to Ignore SEO/AEOLenny cites OpenAI leadership advising founders to ignore optimization and just create great content, pushing Ethan to explain why active optimization is necessary.
Biggest teaching moment ▶ 53:00 Graphite's Large-Scale AI Content StudyEthan presents findings from their 100,000 URL Common Crawl study, demonstrating that 90% of top-ranking search and chat citations remain human-created while 100% automated AI content consistently fails.
Lenny holds their own ▶ 42:08 Clarifying Robots.txt Strategy for Training vs IndexingLenny synthesizes and highlights the distinction between allowing LLM indexing for citation traffic while blocking foundation model training scraping.
the scores for every segment, with the reasoning behind each
| Chapter | Topic | Lenny as informed peer | Guest teaching | Guest disagreement | Lenny pushing back | Why |
|---|---|---|---|---|---|---|
| The Shift from Traditional Search to Generative AI | 0 | 0 | 0 | 0 | This introductory segment consists of the host monologue introducing Ethan Smith and reading sponsor messages for Orcus and Vanta. | |
| Historical Evolution of SEO and Major Algorithmic Shifts | 2 | 6 | 1 | 1 | Ethan educates Lenny on how historical algorithmic shifts like Panda compare to today's shift toward AEO, noting current changes are the second largest in SEO history. | |
| Optimizing for LLM Citations and Early Stage Opportunities | 3 | 7 | 2 | 2 | Lenny brings up OpenAI leadership advice to just write good content, but Ethan reframes this, explaining how LLM citations work and why startups can win citations immediately without domain authority. | |
| The Conversational Long Tail and High Converting Traffic | 2 | 7 | 1 | 1 | Ethan explains the expansion of the conversational long tail in AI queries and shares Webflow data showing a sixfold increase in conversion rates from LLMs over Google search. | |
| On-Site Content Structure and Authentic Reddit Engagement | 3 | 7 | 2 | 1 | Ethan details why growth hacking fake accounts on Reddit fails due to community moderation and explains how authentic brand participation is rewarded by LLM RAG pipelines. | |
| Understanding RAG Mechanics Versus Static Core Training | 2 | 8 | 1 | 1 | Ethan breaks down the technical distinction between training core foundation models and live RAG retrieval, clarifying that AEO operates almost entirely at the retrieval layer. | |
| Combating Generic Content and Measuring Information Gain | 3 | 7 | 2 | 2 | Lenny asks if LLMs will deteriorate into SEO spam farms, prompting Ethan to explain how algorithms measure information gain and typicality to suppress derivative content. | |
| Sponsor Message: Customer UX Research with Great Question | 1 | 8 | 1 | 0 | Following a sponsor break for Great Question, Ethan provides a structured tactical playbook covering question research, citation targeting, and randomized experimentation. | |
| Measuring Answer Share of Voice and Tracking Methodology | 2 | 7 | 1 | 1 | Ethan educates Lenny on why keyword tracking is replaced by statistical distribution sampling and share-of-voice metrics due to the non-deterministic nature of LLMs. | |
| Analyzing Citation Differences Across Major AI Search Platforms | 2 | 7 | 2 | 2 | When Lenny asks if optimizing for ChatGPT is sufficient, Ethan provides empirical overlap numbers between platforms and uses an early search engine historical analogy to argue for multi-platform optimization. | |
| Tailoring AEO Across B2B SaaS, E-Commerce, and Startups | 3 | 7 | 1 | 1 | Ethan clarifies how AEO strategies diverge across B2B SaaS, commerce, and early-stage companies, pointing out distinct citation sources and attribution tracking hurdles. | |
| Robots.txt Indexation Protocols and Scientific Experimentation Design | 4 | 7 | 2 | 2 | Lenny expresses concern about AI scraping newsletter content, leading Ethan to explain the nuance between blocking training bots versus indexing bots via robots.txt and designing controlled A/B experiments. | |
| Debunking Search Myths and Understanding LLM Market Inflection | 2 | 8 | 3 | 1 | Ethan debunks pervasive industry narratives claiming Google search traffic is collapsing, citing VP of search publisher data and contextualizing historical channel scares. | |
| Research Findings on AI Content Efficacy and Model Collapse | 3 | 8 | 3 | 1 | Ethan reveals findings from Graphite's rigorous study of 100,000 URLs, proving that pure AI-generated content fails to rank and explaining the threat of model collapse from derivative loops. | |
| Search Engine Convergence and the Rise of Autonomous Agents | 2 | 6 | 1 | 1 | Ethan outlines the convergence of classic search and chat into multimodal autonomous agents that execute high-context complex workflows on behalf of users. | |
| Unlocking High-Impact Long Tail Traffic Through Help Centers | 2 | 7 | 1 | 1 | Ethan introduces an overlooked AEO tactic: moving help desks from subdomains to subdirectories and populating granular, unaddressed customer support edge cases. | |
| Lightning Round: Essential Books, Mindset, and Remote Hardware | 3 | 4 | 0 | 0 | In the lightning round, Ethan highlights psychology books, his passion for intense sports balanced by Zen climbing, remote hardware setup, and actionable LinkedIn growth tactics. |
Statements from this episode (31)
Ethan Smith: AI is only the second-biggest shift in SEO history
“That was probably the biggest shift, which is Google introduced a bunch of algorithms, Panda and similar things to prevent you from doing spam. So essentially you went from SEO being spam to not spam. That was probably the biggest change. And then this is prob…”
Rachitsky: ChatGPT now drives more newsletter referral traffic than Twitter
“I was looking at my referral traffic and I found that ChatGPT is driving more traffic to my newsletter than Twitter.”
Smith: AI referral traffic surged in January from adoption and clickable UI
“Companies that we work with started in January and it started one because of more adoption, but two is because the answers became a bit more clickable. You have maps, you have shopping carousels, you have clickable cards. So I think the clickability of the ans…”
Smith: Everything that works in SEO works in AEO
“Everything that works in SEO works in AEO, but there are additional things beyond SEO that also work in AEO.”
Smith: Top LLM answers are determined by citation frequency, not ranking
“Usually when you ask something like what's the best tool for X, the first answer will be mentioned at the most in the citations. So that's very different from Google.”
Smith: Startups should not do SEO until Series A or B
“When startups come to me and ask me for SEO help, my first response is don't do it at all. Spend your time on something else. Cause you're not going to be able to grow SEO early on as a in search because you don't have enough domain authority and it takes a wh…”
Smith: Early-stage startups can rank in AEO immediately through citations
“That's not the case for answer engine optimization because you can get mentioned by a citation tomorrow and start showing up immediately. You can have a Reddit thread. You can have a YouTube video. You can be mentioned on a blog, like a, you know, a brand new …”
Smith: Chat prompts average 25 words versus six words on Google
“So the average number of words, I think perplexity said this, or somebody else said it was around 25 words versus Google words, around six words.”
Smith: Webflow LLM traffic converts 6x higher than Google search
“So Webflow, we saw a six X conversion rate in difference between LLM traffic and Google search traffic.”
Smith: Authentic, transparent Reddit participation beats automated spam for LLM citations
“So the strategy is find a thread that is a part of a citation that you want to show up in, say who you are, say where you work, and then give a useful piece of information, and that works really well. And that sounds simple if you're not in the growth mindset …”
Rachitsky: Deel drove its early growth by answering questions on Reddit
“We had the early growth leader from Deel, D-E-E-L on the podcast a while ago, and this is how they grew up and how they grew initially before AI even came around, just going big on Reddit and answering people's questions and like, Hey, it happens to be, Deel c…”
Smith: Google intentionally tuned search algorithms to rank Reddit, Twitter, and Quora
“Google has specifically configured their search algorithm to rank Reddit and Twitter and Quora because they want user generated content. And if it wasn't good content, then they would change the algorithm and they wouldn't rank it.”
Smith: Influencing LLM core pre-trained models takes a year
“Most of what I'm describing is about the RAG piece, not the core model piece to influence the core model is probably extremely hard. And maybe you will see the impact a year later.”
Smith: LLMs will not recommend products absent from RAG search results
“And I think also the LLM is, is probably not going to say your product if it didn't show up anywhere on the RAG. So I think that's where most of the interesting stuff is for, from an optimization perspective.”
Smith: Answering unaddressed subtopic questions drives LLM search retrieval
“The more you answer all the questions, the better. If you don't answer a question, then you're probably not going to show up, and if you answer a question, this follow-up question and subtopic that somebody else is not answering, you're gonna be more likely to…”
Smith: Dotdash Meredith is probably the most cited source in LLMs
“Dot dash Meredith is a large media conglomerate with good housekeeping, all recipes, Investopedia. It's the, it's probably the most successful SEO company of all time. And it's also one of the most cited, probably the most cited in LMs as well.”
Smith: 1 in 20 landing pages drives roughly 85% of traffic
“One out of 20 landing pages drive roughly 85% of all your traffic. So 19 out of 20 landing pages drive little to no traffic”
Smith: Google's E-E-A-T signals show no observable effect on rankings
“Potentially Google has eat expertise, authority, trustworthiness, which actually I don't see having an effect unfortunately, but it could.”
Smith: High-LTV B2B video content offers major AEO opportunity
“YouTube, Vimeo, other video sites, the kinds of things people make videos for are food, traveling, fun, beauty. There's not that many Videos about AI powered payment processing APIs as interesting as that is, but it's a great money turn. So if you make a video…”
Smith: Majority of SEO and AEO best practices are incorrect
“SEO and AEO are both interesting in that the majority of the information and best practices are not correct. And the reason why is because people don't do analysis. Somebody will say something and then it will get repeated and everyone, and then it becomes bes…”
Smith: Buy the cheapest answer or keyword tracking tool available
“My general suggestion is pick the one that, pick the cheapest one that does what you need. Just like keyword tracking, you can only, you know, there's not a premium version of keyword tracking. You rank number three or you don't. So pick the keyword tracker th…”
Smith: ChatGPT citation overlap with Google is only 35%, Perplexity is 70%
“We did a study looking at thousands of questions and saw the citation overlap with Google search results was around
35% for ChatGPT and Google. So not that much perplexity was around 70%”
Smith: AEO traffic is falsely attributed to branded search and direct visits
“What they will do is they will open a new tab and they will type in the brand name and they will go to Google and then they'll click on your domain and you will think that it was branded Google search when it wasn't. Or they'll open up a new tab and they will …”
Smith: LLM answer volatility requires control groups in AEO testing
“We are seeing a fair amount of variance in answers just without doing anything at all. So you definitely want a control group.”
Smith: Webflow gets 8% of signups from LLMs
“Webflow, they get eight percent of those signups from LMS.”
Smith: Google VP confirmed publisher referral traffic is up slightly
“Google published something recently. Their VP of search explicitly said, I looked at our, the traffic that we're sending to publishers and it is not down. It's up slightly. So it is not true that Google search is going down.”
Smith: Only 10 to 12% of Google and ChatGPT results are AI-generated
“It was around 10 to 12% of content in Google search and in ChatGPT are AI generated. 90% or not.”
Smith: There is more AI-generated content online than human-created content
“We found that there's more AI generated content on the internet than human generated content. So back to the common crawl study, we looked at a 100,000 different URLs over the past five years. And then you can see this curve where AI generated is now higher th…”
Smith: Feeding AI derivatives into RAG collapses output diversity into single opinions
“If you feed in derivatives of derivatives into the model, you will basically take the wisdom of the crowd and that will shrink and you'll have a single opinion on everything, which is really bad.”
Smith: Help center subdomains perform worse than subdirectories
“It's frequently on a subdomain for whatever reason, subdomains don't work well to subdirectories. So move it to a subdirectory.”
Smith: Community-generated help content captures long-tail LLM citations
“I might even open up to the community. Anyone can ask anything because the community will then fill in the tail and then answer those. And again, in many cases, there might be nobody talking about this at all. So you could be the only citation for this and the…”