How to Rank in ChatGPT, Claude & Perplexity (Generative Engine Optimization)

A year ago, "ranking in ChatGPT" sounded like a buzzword. Today, it's the most important shift in search since Google launched the algorithm. When a customer in Mississauga asks ChatGPT "best web designer near me", only one company gets recommended. The 9 below it never get clicked. There is no page 2.

This guide covers Generative Engine Optimization (GEO) — the practical work of making your business the one that AI answer engines recommend. It's the playbook we use for our clients on our SEO program.

What Generative Engine Optimization actually is

GEO is the discipline of optimizing for AI answer engines — ChatGPT, Claude, Perplexity, Google's AI Overviews, Microsoft Copilot, and Gemini. Where traditional SEO targets a list of ranked links, GEO targets inclusion in the generated answer itself.

The shift is structural. A user who used to skim 10 blue links and pick one now reads a single 4-paragraph answer with 2-4 source citations. That answer either mentions you, or it doesn't.

How AI engines pick the sources they cite

Different engines use different stacks, but the selection signals are remarkably consistent. AI engines favour pages that are:

Crawlable. The bot can fetch the page without being blocked.
Parseable. Clean HTML, semantic headings, structured data — no JS-only walls of text.
Answer-ready. The page explicitly answers a clear question, with the answer near the top.
Authoritative. Cited and linked to from sources the engine already trusts (news, .edu, .gov, established brands).
Recent. Freshness matters more in AI search than in classic SEO.
Specific. Concrete facts, numbers, examples beat generic prose every time. Vagueness is invisible to LLMs.

Step 1 — Make your site crawlable by AI bots

Most sites silently block half the AI crawlers without realizing it. Open your robots.txt and explicitly allow:

User-agent: GPTBot
Allow: /

User-agent: OAI-SearchBot
Allow: /

User-agent: ChatGPT-User
Allow: /

User-agent: PerplexityBot
Allow: /

User-agent: ClaudeBot
Allow: /

User-agent: Google-Extended
Allow: /

User-agent: Applebot-Extended
Allow: /

If your site is a JavaScript SPA, make sure your meta and structured data are present in the initial HTML — not injected client-side only. Most LLM crawlers don't render JavaScript.

Step 2 — Publish an llms.txt file

llms.txt is to AI engines what robots.txt is to search engines, and what sitemap.xml is to crawlers — a plain-text, structured map of your site optimized for LLM consumption.

Put it at /llms.txt and include: a one-paragraph summary of who you are, your services with canonical URLs, your service area, and a list of high-value pages. Keep it short and factual — no marketing fluff. LLMs reward density, not length.

Step 3 — Structured data is your AI passport

Schema.org JSON-LD is the single highest-leverage GEO tactic. It's the only way to give an AI structured, unambiguous facts about your business: who you are, where you operate, what you sell, what it costs, what your reviews look like.

The non-negotiable schema types for a local business:

LocalBusiness (or a more specific subtype like ProfessionalService or Plumber)
Person for the founder or key team members
Service for each service line, with Offer for pricing
FAQPage for FAQ sections — these get pulled directly into AI answers
BreadcrumbList for navigational context
Article / BlogPosting for content pages

Step 4 — Write in answer-ready format

LLMs extract direct, specific answers. The pages that get cited follow a predictable pattern:

Question-shaped H2s. "How long does local SEO take?" beats "Timeline considerations for SEO".
Answer in the first sentence. Then expand. Don't bury the lead.
Specific numbers, dates, and prices. "$89/mo" and "7-10 days" and "60 days" are the kind of concrete facts LLMs love to cite.
Comparison tables. When you compare options, format it as a table — LLMs extract them cleanly.
Inline citations. Link to your sources. LLMs check link graphs to verify trust.

Step 5 — Build third-party authority

On-page work is necessary but not sufficient. AI engines weigh cross-source agreement heavily. If only your own site says you're the top web designer in Mississauga, the LLM doesn't believe you. If three news sites, two industry directories, and your Google Business Profile all back it up, the LLM cites you with confidence.

The fastest authority moves for a small business:

Get listed and verified on Google Business Profile, Apple Business Connect, and Bing Places.
Earn at least 25 reviews on Google with keyword-rich text.
Get cited in 2-3 local news articles, blogs, or podcasts (a small business mention in Mississauga News is gold).
Build out 2-3 high-quality directory profiles (BBB, industry-specific).
Publish on at least one third-party site (guest post, expert quote, industry publication).

How to measure GEO success

Classic analytics misses most of this. Three signals to watch:

Branded queries in ChatGPT, Claude, Perplexity. Ask the engines directly: "Who is the best [your service] in [your city]?" — and track when you start being mentioned.
Referral traffic from chat.openai.com, perplexity.ai, claude.ai, copilot.microsoft.com. Tag and track them as a separate channel in GA4.
Inbound calls and form fills citing AI. Ask new leads "how did you find us?" — the answer "ChatGPT recommended you" is becoming surprisingly common.

FAQ

Is GEO the same as SEO?

No, but they overlap. Classic SEO optimizes for ranked lists of links. GEO (Generative Engine Optimization) optimizes for being cited inside an AI-generated answer. The technical foundations (crawlability, structured data, authority) are similar; the on-page tactics (answer-ready phrasing, llms.txt, evidence density) are different.

Should I block AI crawlers like GPTBot?

Only if you have a strong reason to (paywalled or proprietary content). For most small businesses, blocking AI crawlers is the equivalent of de-listing yourself from Google in 2008 — you become invisible to the place customers will increasingly start their search.

How fast can I rank in ChatGPT or Perplexity?

Faster than classic SEO. ChatGPT's web search and Perplexity index in days, not months. We've seen well-structured pages get cited within 2-4 weeks of publishing. The catch: you need both crawlability AND authority — neither alone is enough.

Does an llms.txt file actually do anything?

Today, llms.txt is a soft signal — not enforced like robots.txt. But it's a low-cost, future-proof way to give LLMs a clean, structured map of your site. Anthropic's Claude, Perplexity, and several other providers have indicated they read it. Even if today's lift is small, it costs an hour and ages well.

Want this implemented for your business? GEO is built into every SEO engagement we run, and into every monthly website plan we build. Book a free 15-minute audit and we'll show you where you currently stand in AI search.