All work

GEO strategy

How I Get Businesses Recommended by AI

My generative engine optimization (GEO) process for getting a brand named, described accurately, and linked by ChatGPT, Claude, and Google's AI search.

Photo by Zetong Li on Pexels

SEO brief for this piece
Target keyword
generative engine optimization (primary); how to get recommended by AI (secondary)
Why this target
Business owners and employers searching for GEO want a working method, not another definition. This page answers with a process and first-party test results.
GEO tactics
An answer-first summary, five numbered steps AI tools can quote one at a time, dated first-party test results as proof, and Article schema.
Search intent
Informational with a hiring angle. A reader wants to know how GEO works in practice and who can do it.
Secondary terms
GEO strategy, AI search visibility, get recommended by ChatGPT, optimize for Google AI Overviews
Meta title
How I Get Businesses Recommended by AI | GEO Process
Meta description
My generative engine optimization process for getting a business named, described accurately, and linked by ChatGPT, Claude, and Google AI search.
Structured data
Article schema
The short answerAI tools recommend businesses they can identify with confidence and verify through other sources. My job is to make both easy: clear answers on the brand's own pages, consistent facts everywhere it appears, outside proof, and pages machines can read. Then I test every month to see what's working.

1. Lead with the answer

AI tools pull short, self-contained statements. Every page I write opens with a direct answer to the question it targets, in two or three sentences a model can quote as is.

2. Make the business easy to identify

Name, location, hours, and description should match on the website, the Google Business Profile, and every directory listing. Schema markup states those facts in a format machines read directly. Mixed signals make a model hesitate, and a hesitant model recommends someone else.

3. Earn outside validation

A brand's website is one voice; reviews, local press, and travel guides are many. AI tools weigh what others say, so review responses and third-party listings belong in the content plan.

4. Build pages machines can read

Real HTML tables instead of images, FAQ and product schema, clean headings, and a robots.txt that welcomes AI crawlers.

5. Measure with real prompts

Each month I run a fixed set of prompts through ChatGPT, Claude, and Google AI Mode, including searches that never mention the brand. I record whether it's named, described correctly, and linked. Google Search Console separates branded from non-branded traffic.

Proof

In October 2026, I put this process to the test for St. John Brewers. I asked Google AI Mode and ChatGPT to name the best brewery on St. John, without mentioning the brand. Both chose St. John Brewers. Next, I asked Claude for things to do on the island. The Tap Room was the only bar or restaurant in its answer. With no local competition, ranking first proves little on its own. The real result is accuracy. Every tool described the brand correctly and linked readers to its website. Read the case study.

What I don't promise

Nobody controls what an AI recommends. Models change, and answers vary from one run to the next. What I offer is a repeatable process and honest measurement.