The old model had one box
For twenty years, organic growth meant one thing. A person typed words into a search box, got ten blue links, and clicked one. Win the ranking and you won the visit.
That model still exists. It just isn't the only one anymore. The same buyer now asks the same question in three different places, and each place decides differently what to show.
Three places the question gets answered
- SEO, search engine optimization
- Earning a ranked link in a classic results page. The prize is the click.
- AEO, answer engine optimization
- Earning the direct answer: the featured snippet, the AI overview, the voice reply. The prize is being quoted.
- GEO, generative engine optimization
- Earning a citation when ChatGPT, Gemini, Claude or Perplexity writes the answer from scratch. The prize is being named as the source.
Most companies staff these as three projects. One team chases rankings, another chases snippets, and nobody owns the chatbots. That's how you end up with three versions of your own price on the internet.
Why it has to be one system
All three engines read the same pages. They reward the same things: a clear question, a direct answer, consistent facts, and a source that looks like it knows what it's talking about. A page that wins in one usually helps in the others.
There's a second-order effect most people miss. Generative engines learn which sources to trust partly from what ranks and what gets cited elsewhere. So SEO feeds GEO. And a brand that gets named in AI answers gets searched for by name, which feeds SEO. The loops run both ways.
One question map. One claim library. One scorecard. Three engines.
How we run it
- Map the questions. Every question a buyer asks, grouped by stage. See the question map.
- Lock the facts. One canonical version of every price, definition and claim. See claim lock.
- Write answer pages. Each page answers one question in the first two sentences, then earns the right to go deeper.
- Mark it up. Schema for the organization, the people, the offers and the FAQs, so machines don't have to guess.
- Treat video as search. YouTube is the second-biggest search engine, and its chapters show up in AI answers. See YouTube is a search engine.
- Measure citation share. For a fixed set of prompts, how often are we the cited source? Track it weekly, next to rankings and conversions.
What counts as winning
Visibility is necessary but not sufficient. A citation that never turns into a customer is a nice screenshot. We report citation share and rankings as leading indicators, and we report paid customers as the outcome. Both go on the same page so nobody can celebrate one without seeing the other.
Scope conditions
This works when buyers ask questions before they buy. That covers almost every legacy industry we target: law, accounting, compliance, consulting, financial services. People research those decisions hard.
It works less well for impulse purchases, or for categories so new that nobody searches for them yet. There, you have to create the question before you can answer it, and distribution starts with outbound and creators instead.
Building in a legacy industry?
We bring the agent crew and the distribution. You bring what you know about the industry.