Breaking
AEO ≠ SEO — being mentioned most beats ranking first 6x — Webflow's conversion edge, LLM vs Google traffic 8% — of Webflow signups now arrive from LLMs 25 words — the average chat query vs Google's six 1 in 20 — landing pages drive 85% of all traffic Butter lettuce — Ethan Smith's proudest ranking
The Story · Answer Engine Optimization

The Answer Men: How to Get a Robot to Recommend You

For eighteen years the game was ranking first on Google. Then the machines started summarizing — and the whole board flipped.

Ethan Smith, CEO of Graphite
Ethan Smith spent 18 years teaching machines to notice him. Now he teaches the machines' machines. Also: he really does like butter lettuce.
The Feature

A Man, a Machine, and the Vanishing Blue Link

Here is a fact that ought to trouble anyone who has ever paid for a Google ad. When Ethan Smith's clients ask ChatGPT to name the best website builder, the winner is not the company whose link appears first in the citations. The winner is the company that gets named the most times across all of them. Ranking, that old religion, has been quietly replaced by something closer to gossip: whoever is talked about most, wins.

Smith is the CEO of Graphite and, by his own accounting, has been optimizing for search engines since 2007 — back when the sport involved auto-generating a hundred million landing pages of scraped product snippets and waiting for Google to notice. It did notice. It launched an algorithm named Panda, and an entire industry of clever people watched their traffic evaporate overnight. Smith calls that the biggest change he has ever witnessed. This new one — the migration of curious humans from the search box to the chat box — he ranks second. Which is itself the point: everyone else is calling it the apocalypse.

"A lot of people are seeing this as everything is different," Smith says. "Nothing we've done before is going to work. We have to rethink everything." He does not believe this. What has arrived is not a new universe but a new layer — search, plus summarization, plus a few novel inputs. The engineers call it RAG, for retrieval-augmented generation: the model runs a search, gathers citations, and then does the genuinely new thing, which is to blend them into a single confident paragraph. Everything that worked in the old SEO still works. It's just that the old SEO no longer describes the whole board.

You need to get mentioned as many times as possible. Usually the first answer will be the one mentioned the most in the citations.— Ethan Smith

To understand why this matters, consider the two ends of the animal. The head — the popular questions, the "what's the best X" — behaves differently now. In Google, if your blue link showed up first, you won, full stop. In an answer engine, a first-place citation buys you nothing if a rival is name-checked five times and you're name-checked once. The machine is not choosing a winner; it is taking a poll and reading you the average.

The tail, meanwhile, has grown a second life. The average chat query runs to roughly twenty-five words, Smith notes, against Google's spare six. People ask, then follow up, then follow up again, trailing intent behind them like a lit fuse. Whole categories of hyper-specific questions — the ones about whether some obscure payment-processing API integrates with some obscure analytics tool — have never been typed into a search bar in the history of the internet, because search could never support that much specificity. Chat was built for exactly it. And a company that answers a question nobody else has bothered to answer can become, delightfully, the only citation in the room.

The startup that couldn't win Google can win ChatGPT tomorrow

This is the part that upends the usual advice. When early-stage founders came to Smith asking for SEO help, his standard reply was blunt: don't. Not yet. Google rewards domain authority, and domain authority is a slow, expensive fruit — a Series A or B luxury, not a launch-week one. But answer engines have no such patience tax. "You can get mentioned by a citation tomorrow and start showing up immediately," Smith says. A Reddit thread, a YouTube video, a blog post — a brand-new YC company that everyone is suddenly talking about can appear in an answer the same afternoon. The long tail, pronounced dead in the SEO era, has been resurrected in chat, and this time the little companies can reach it.

You can get mentioned by a citation tomorrow. Early stage companies can win. They can win quickly.— Ethan Smith

And the traffic, it turns out, is not merely early — it is better. At Webflow, Smith's team measured a sixfold difference in conversion rate between LLM referrals and ordinary Google search visitors. Sixfold. His theory is that a person arriving from a chat has already been primed by a long, meandering conversation, has narrowed their want to a fine point, and shows up not as a browser but as a buyer. Eight percent of Webflow's signups now come through this channel. It is not the largest channel. Paid still is. But it is, unmistakably, a top one — and, as everyone agrees, growing.

Then there is Reddit, which occupies a strange and central role in this story. It is among the most-cited sources inside the large language models, and it is the single thing Smith's clients ask about most. The instinct of every growth hustler is identical and identically doomed: spin up a hundred fake accounts, auto-post, upvote yourself, manufacture a trust score, and declare your product the best in every thread. It does not work. The accounts get banned; the comments get deleted; the community — which is very, very good at this — swats them like flies.

What works is almost insultingly simple. At Webflow, a couple of real employees find the right thread, say who they are, say where they work, and offer one genuinely useful piece of information. Five such comments can do more than ten thousand fabricated ones. The machines lean on Reddit precisely because humans police it, and the moment it stopped being trustworthy, the engineers tuning the citation algorithms would stop citing it. Authenticity, in other words, is not a virtue here. It is the exploit.

Smith is careful to locate all of this on the correct side of the machine. None of it touches the training data — the frozen, next-word-predicting core that knows the capital of California is Sacramento. It lives in the retrieval layer, the live search the model runs before it answers. That is the surface a mortal can move. Influencing the core model is, he suspects, nearly impossible and profoundly not worth your afternoon.

The Core Ideas of Answer Engine Optimization

Six moving parts. Master them and you don't rank the machine — you become the thing it can't help but recommend.

01

LLM + RAG

An answer is a search that got summarized. Optimize the retrieval layer, not the frozen training core.

02

Mentions, Not Ranks

First place buys nothing. The most-cited brand wins the answer. Get named again, and again, and again.

03

The Long Tail Returns

25-word questions nobody ever searched. Answer the obscure and be the only citation in the room.

04

Off-Site Citations

YouTube, Vimeo, Reddit, Quora, affiliates like Dotdash Meredith. Show up everywhere the machine looks.

05

Help-Center Tail

Chat asks "does it do X?" Answer it in a cross-linked help center on a subdirectory — not a subdomain.

06

Track & Experiment

Measure share of voice. Run control vs. test groups. Trust no best practice you didn't reproduce.

The Receipts

Numbers That Reorganize Your Priorities

6x
LLM vs Google conversion (Webflow)
8%
of Webflow signups from LLMs
~25
words in the average chat query
1/20
pages drive 85% of traffic

Citation Overlap With Google Search Results

From Graphite's study of thousands of questions — how often each engine cites what Google ranks
Perplexity
~70%
ChatGPT
~35%
AI content cited
~10–12%
The Distinction

Same Machine, Different Rules

Everything that works in SEO still works in AEO — Smith is emphatic on this. But two ends of the board behave differently, and that's where the new money is.

Traditional SEO

  • Rank one blue link first and you win the click
  • Needs domain authority — a Series A/B luxury
  • Head query ~6 words; tail is thin
  • Landing pages target thousands of keywords
  • Google tells you search volume via its ads API

Answer Engine Optimization

  • Being mentioned most across citations wins the answer
  • New companies can get cited tomorrow — win fast
  • Head query ~25 words; a fat, specific long tail
  • Pages target thousands of follow-up questions
  • No truth set yet — you mine sales calls & support
The Playbook

Seven Moves to Become the Answer

Smith's tactical plan, unwrapped. Half of it will work for you; half won't. The last step tells you how to know which.

Find your questions

Take your (and your competitors') paid-search money terms and ask ChatGPT to turn them into questions.

Track them

Load them into an answer tracker. Measure share of voice and average rank across surfaces. Pick the cheapest tool that works.

Map the citations

See who's being cited and where the clicks go, then build a separate strategy for each citation group.

Build the landing pages

Match the page type that already ranks — and answer every follow-up question a curious human might ask.

Win off-site

Affiliates (easy, paid), YouTube/Vimeo (uncontested for B2B), and honest Reddit comments (trickier, but real).

Run the experiment

Control group untouched, test group intervened. Two weeks before, two weeks after. Reproduce before you believe.

Assemble the team

Your SEO people do the on-site work; a community/marketing generalist handles YouTube and Reddit.

?

"It's not your choice whether to play the game."

Block the training bots if you must — but let the index bots in. Say no to everything and your competitor simply becomes the answer instead of you. Google's slice of the pie stays the same; the pie just gets bigger.

The Cautionary Tale

The Vanilla Apocalypse

There is a temptation, natural and enormous, to let the machines write the answers the machines will read. Smith saw it coming the day ChatGPT launched — his one-word forecast was "spam" — and he had reason. In 2007 he and his rivals scraped one another's reviews, chopped them up, and spun out landing pages by the hundred million. It worked beautifully, right up until it destroyed the companies that did it.

So Graphite ran the study rather than guessing. They pushed thousands of queries through Google and ChatGPT, ran the citations through an AI detector they'd first validated against pre-ChatGPT content from Common Crawl, and found that only ten to twelve percent of cited pages were machine-written. Ninety percent were human. Purely AI-generated content, in other words, does not work. The future is AI-assisted and human-edited — a person with a good tool, not a tool with no person.

Most work is wasted in SEO. Most work is wasted in AEO. How do you know what's not wasted? You do an experiment.— Ethan Smith

The deeper danger has a name: model collapse. Imagine the answer engine citing its own summaries, then summarizing those, then summarizing the summaries — an infinite loop of derivatives. The wisdom of the crowd, which is the whole reason the average of many opinions beats the best single one, quietly narrows. Ask for the best flavor of ice cream and, generation after generation, the diversity drains away until the machine declares — with total confidence — that the only flavor is vanilla. There is no other. There never was.

Which is why there are humans at these companies whose entire job is to keep 2007-Ethan from winning. When the head of ChatGPT told Lenny not to worry about AEO — just write great content and the model will find it — Smith half-agrees. Anything can be optimized, he insists; you only need the rules of the game. But spam is not a robust strategy, and the person on the other side will see it and change the algorithm to stop you. The alignment, happily, points the right way: what the machine rewards — original research, real expertise, information nobody else supplied — is exactly what makes content worth reading in the first place.

Smith's proudest trophy is not a Fortune 500 logo. It's butter lettuce. Handed a MasterClass with none of the domain authority of All Recipes or Martha Stewart, he wasn't sure he should even take the job. He took it. It was hard. And today you can search "butter lettuce" and find a MasterClass recipe near the top — a small, green monument to the idea that the underdog can still win, if it knows exactly which questions no one else has answered. He likes butter lettuce, for the record. That helps.

The Voices

Who's Talking

Guest

Ethan Smith

CEO of Graphite and Lenny's go-to SEO expert. In the game since 2007, from programmatic commerce SEO to the answer-engine era. Runs a research team publishing original AEO studies; teaches at Reforge. Loves UFC and climbing documentaries in equal measure.

Host · @lennysan

Lenny Rachitsky

Host of Lenny's Podcast and author of the leading product-and-growth newsletter. Former Airbnb product lead — and, to his own surprise, now getting more newsletter traffic from ChatGPT than from Twitter.

The Margins

Facts for the Footnotes

The Whole Conversation, Unabridged

Ethan Smith and Lenny Rachitsky, on the new rules of showing up inside the machine.

Share the Answer

in · LinkedIn X · Twitter f · Facebook ◎ · Instagram