THE NEW SEARCH · A FIELD GUIDE
Will AI say your name?
When a buyer asks ChatGPT for a recommendation, the engine names two or three brands and moves on. This page explains how those brands get chosen, and gives you a 60-second way to check your own odds.
THE ALPHABET, DECODED
Three acronyms. One job.
Being the thing that gets recommended, wherever the asking happens.
SEO
Search Engine Optimization
Getting ranked in the classic list of links. Still the foundation: AI engines lean on the same quality signals, and plenty of buyers still click.
AEO
Answer Engine Optimization
Getting chosen as the direct answer: featured snippets, AI Overviews and assistant replies. Won by pages that answer the question in the first breath.
GEO
Generative Engine Optimization
Getting named and cited inside AI-generated answers on ChatGPT, Perplexity, Gemini and Claude. Won with structure, freshness, credibility and third-party proof.
THE ANSWER PATH
The answer is short. The evidence path is not.
An AI answer compresses many source signals into a few sentences. Clear structure, corroboration and direct language help the engine preserve your meaning, trace it back to you, and cite the page instead of quietly using it as an unnamed ingredient.
UNDER THE HOOD
How an AI answer gets made.
Four steps between a buyer's question and the brands that get named. You can lose at any of them. Most sites lose at the first two.
- 01Crawl
AI bots like GPTBot, ClaudeBot and PerplexityBot read your site. If robots.txt blocks them, or your content only exists after heavy scripts run, you are out before the game starts.
- 02Extract
Engines pull passages, not pages. Clean headings, direct answers and schema markup decide whether your best material survives extraction intact.
- 03Synthesize
The model blends sources into one answer, favoring content that is specific, current and consistent with what the rest of the internet says about you.
- 04Cite
A handful of sources get named. Everyone else is invisible ingredients. The whole discipline of GEO is engineering your way into that handful.
A handful of sources get named. Everyone else is invisible ingredients.
THE SIZE OF THE SHIFT
This is not a trend piece. It's the traffic report.
0%
predicted drop in traditional search volume by 2026 as buyers move to AI assistants
GARTNER PREDICTION · FEB 2024
0%
of US adults in a tracked panel hit at least one AI-summarized search in a month, and they clicked results about half as often
PEW RESEARCH CENTER · JUL 2025
0%
visibility boost measured when content cites its sources
PRINCETON GEO STUDY · KDD 2024
YOUR TURN
Score yourself before the machines do.
Six questions, sixty seconds, no email required. Honest answers get honest scores.
THE 60-SECOND VISIBILITY CHECK
1 / 6
Ask ChatGPT about your product or service category. Does it mention you?
Go ahead, we'll wait. This is the whole ballgame.
THE WORK ITSELF
The checklist we actually run.
No mystique. This is the punch list every engagement works through, tuned to what the audit finds.
A NOTE ON HONESTY
Some vendors sell llms.txt files like magic beans. Truth: it is an emerging convention, cheap to add and sensible to have, but no major AI company has committed to reading it yet. We include it, we label it as the hedge it is, and we spend your budget on the levers with measured effect: structure, freshness, schema and third-party authority.
- [01]Prompt-panel baseline: what 5 AI engines say about you today
- [02]Robots.txt access for GPTBot, ClaudeBot, PerplexityBot, Google-Extended
- [03]Answer-first rewrites of your money pages
- [04]Question-shaped headings that match how buyers actually ask
- [05]Schema markup: Organization, Service, FAQ, Product, Article
- [06]llms.txt and machine-readable service and pricing data
- [07]Freshness pass: dated updates, current stats, pruned dead weight
- [08]Third-party footprint: reviews, communities and mentions engines trust
- [ ✓ ]re-run the prompt panel. show the delta.
STRAIGHT ANSWERS
The questions everyone asks.
No, and nobody honest can: the models decide. What we guarantee is the inputs that measurably move citation rates: extractable structure, schema, crawler access, freshness and third-party credibility, plus a monthly prompt panel that shows exactly where you stand and how it changes.
Some surfaces move in weeks: Perplexity and ChatGPT's browsing pull from the live web, so structural fixes can show up fast. Others, like model training data, move over months. That is why we measure monthly against a fixed prompt panel instead of promising a date.
No. Done right, it is the same discipline Google rewards: clear answers, real expertise, structured data and content written for people. Google explicitly recommends against making special AI-only content, and so do we. One site, organized well, serves every engine.
THE NEXT MOVE
Find out what the machines say about you.
The free visibility report runs your real buyer prompts across five AI engines and hands you the gaps, ranked by payoff. It is the least salesy sales call you will take this year.
NO RETAINER REQUIRED TO START · REPLY WITHIN ONE BUSINESS DAY