Blog

What is GEO, and how to check if ChatGPT recommends your brand

Jisung Yoon · Yunju Company · 15 November 2026

Short answer: GEO (generative engine optimization) is the work of making your business easy for AI assistants and AI search (ChatGPT search, Perplexity, Google AI Overviews, Copilot) to find, read and recommend. You can check where you stand in about 30 minutes: ask 10 real buyer questions, note whether you are named and which pages are cited, then check four technical files on your site.

GEO in one paragraph

Classic SEO tries to rank a page in a list of links. GEO tries to get your business named and cited inside the answer. The answer engine reads pages it is allowed to crawl, prefers sources that answer the question directly, and cross-checks facts across your site and third-party profiles. So GEO combines three things: technical access (can the crawler read you?), answer-ready content (does a page answer the buyer's question in the first lines?), and consistent entity facts (is your name, location, offer and price the same everywhere?).

The 30-minute self-check

Step 1: write 10 buyer questions (5 minutes)

Write the questions a buyer types before they know your name. Mix three kinds:

• Category: "best [category] for [audience]"
• Problem: "how do I [problem your product solves]"
• Comparison: "[competitor] alternatives" or "[you] vs [competitor]"

Step 2: ask them in a clean session (10 minutes)

Use a logged-out or temporary chat so your history doesn't bias the answer. For each question, record: were you named (yes/no), your position in the list, which competitors were named, and which URLs were cited as sources. The cited URLs matter most: they tell you which pages the engine trusts for that question.

Step 3: check four technical files (10 minutes)

Check	Where	What good looks like
AI crawler access	/robots.txt	No Disallow: / for OAI-SearchBot, GPTBot, ClaudeBot, PerplexityBot or Google-Extended. Explicit Allow rules are clearer.
Site summary for AI	/llms.txt	A plain-text file: who you are, what you sell, prices, key URLs.
Structured data	View source, search application/ld+json	Organization plus Service or Product with Offer (price) and FAQPage where you have FAQs.
Sitemap	/sitemap.xml	Lists every page you want found, with recent lastmod dates.

Step 4: read your own pages like an answer engine (5 minutes)

Open your main service page. Does the first paragraph say what it is, who it's for, the price and the delivery time? If the answer is buried under a slogan, an engine has to guess. Rewrite the first two sentences so they can be quoted on their own.

What we found when we checked 10 brand websites

On 15 November 2026 we ran the same checks on 10 public brand websites (six B2B software companies and four direct-to-consumer product brands). Using a fixed 100-point readiness rubric:

• Scores ranged from 52 to 97, with a median of 80.
• 5 of 10 had an llms.txt with a real brand summary; 3 more had one that only described a shopping protocol; 2 had none.
• Only 3 of 10 named AI crawlers explicitly in robots.txt. None blocked them.
• 4 of 10 had both FAQ content and FAQPage markup.
• The site with the lowest technical score (52) was still named first for both of its buyer questions, because its own comparison page was the top cited source.

The lesson: technical files remove friction, but a page that directly answers the buyer's question is what gets cited. The full method and one sample report are on our samples page.

The fixes that usually matter most

1. An answer-first paragraph at the top of each service or product page.
2. A dedicated pricing page (engines get asked "how much does X cost" constantly).
3. An FAQ page with FAQPage markup, written in the buyer's words.
4. Organization, Service/Product and Offer JSON-LD with the same facts as the visible page.
5. llms.txt and llms-full.txt summarizing the business.
6. Consistent name, description and links on every third-party profile.
7. Comparison and "how to choose" pages for your category.

When to get help

If you run the self-check and want it done properly (25+ questions, competitor share-of-answer, ready-to-paste fixes), our GEO audit is $149 and delivered in 48 hours, in writing.