What's in this guide
1. The Shift: Why AI Search Changes Everything
For the last twenty years, "being findable online" meant one thing: ranking on Google. You bought a domain, you tuned your title tags, you built backlinks, you waited. SEO was a discipline so well-understood that entire careers were built on its rules.
Those rules are quietly being rewritten.
Today, an increasing share of people don't open Google when they want a recommendation. They open ChatGPT. They ask Claude. They search inside Perplexity. They use the new AI summary at the top of Google itself. In every one of those interfaces, the user gets a direct answer, not a list of ten blue links to click through. The business named in the answer wins. Everyone else is invisible.
This is AI Search Optimization — also called Agent Discovery Optimization (ADO), Generative Engine Optimization (GEO), or Answer Engine Optimization (AEO). The terms vary; the practice is the same. The goal: get your business recommended by name when someone asks an AI for a suggestion in your category.
2. Why Traditional SEO Doesn't Work for AI
The first instinct of most marketers is: "OK, but my site already ranks #1 for my keywords. So I'm fine, right?"
Not quite. Here's the uncomfortable truth: a business can rank #1 on Google and still be completely invisible to ChatGPT. The two systems are evaluating you on fundamentally different signals.
| Signal | Traditional SEO (Google) | AI SEO (ChatGPT, Claude, etc.) |
|---|---|---|
| Primary input | Live crawl of your website | Training data + retrieval at query time |
| What gets weighted | Backlinks, page speed, keyword density | Citations on authoritative sources, schema markup, structured data, content formatted for quoting |
| Result format | 10 blue links the user picks from | One direct answer that names 1-3 businesses |
| How often it updates | Continuously (re-crawls weekly) | Training updates quarterly; retrieval updates daily |
| What makes you "win" | Out-ranking competitors on a SERP | Being the business the AI names first in its answer |
If those signals barely overlap, that's because they barely do. Traditional SEO maximizes your chance of being one of ten options. AI SEO maximizes your chance of being the option.
3. The 4 Pillars of AI Search Optimization
Strip the discipline down to its load-bearing components and there are exactly four. Get these right, and you're ahead of 95% of small businesses — not because the work is hard, but because nobody else is doing it yet.
Pillar 1 · llms.txt
A simple text file at the root of your site that tells AI crawlers what you are, who you serve, and what to quote. The AI equivalent of robots.txt — but instead of restricting crawlers, it guides them. Most sites don't have one. Yours should.
Pillar 2 · AI-friendly robots.txt
The traditional robots.txt rules need updating for 14+ AI crawlers (GPTBot, ClaudeBot, Google-Extended, PerplexityBot, and more). Most sites either accidentally block them or have no rules. Get this right and you're literally inviting AI in.
Pillar 3 · Schema.org structured data
JSON-LD markup that turns your unstructured webpage into machine-readable facts: who you are, what you sell, your address, your hours, your prices, your services. AI models lean on schema heavily because it's unambiguous.
Pillar 4 · Citations & authority signals
AI models pick businesses they've "heard of." That means being mentioned on Wikipedia, industry directories, local press, podcasts, and authoritative blogs. Citations are the trust currency of the AI era.
That's the whole game. There are nuances inside each pillar — and entire books being written about them — but if you address all four, you're operating on a different plane than your competitors.
4. How AI Models Choose Who to Recommend
The black box isn't as black as it looks. When you ask ChatGPT for a recommendation in any category, here's the rough mental model of what's happening behind the curtain:
- The AI parses your query. It identifies the entity ("coffee shop," "plumber," "marketing agency"), the location ("Huntington Beach"), and the intent (find/compare/learn).
- It pulls from training data. Anything indexed before the model's knowledge cutoff is fair game — including your website (if it's accessible), business directories, Wikipedia, news articles, reviews, and forum discussions.
- For real-time models (ChatGPT search, Perplexity, Gemini), it also retrieves live web results. This is where having a clean schema and llms.txt suddenly matters a lot.
- It synthesizes an answer. The businesses it names are the ones that appeared most frequently, most authoritatively, and most cleanly across its sources.
Notice what's not in that list: keyword density, meta tag optimization, page speed scores. None of that matters at this layer. What matters is: did the AI see your business mentioned by trusted sources? Was your data structured cleanly when it crawled you? Did your content answer the question someone is asking right now?
5. ADO vs SEO vs GEO vs AEO (Cutting Through the Noise)
You'll see a lot of acronyms in this space. Here's the honest breakdown:
- SEO (Search Engine Optimization) — The original. Ranking on Google. Still matters, but no longer enough.
- AEO (Answer Engine Optimization) — A term that emerged with Google's featured snippets. About being the answer in a search result.
- GEO (Generative Engine Optimization) — The trendiest term right now. Same concept as AEO but framed for the generative AI era.
- ADO (Agent Discovery Optimization) — The term we use. Emphasizes that modern AI assistants act as agents on the user's behalf — they discover, evaluate, and recommend. ADO is the practice of being recognized by those agents.
All four terms describe the same practical work — about 80% overlap. The vocabulary will probably consolidate over the next year or two. In the meantime, anyone telling you these are dramatically different things is selling you complexity that isn't there.
6. The DIY Path: What You Can Do This Week
Most of AI SEO is not hard. It just requires knowing what to do. If you have a few hours and you're comfortable editing your website, you can knock out the foundation yourself:
- Run a baseline check. Ask ChatGPT, Claude, Gemini, and Perplexity to recommend "[your business type] in [your city]." Note whether you appear. Save screenshots. You'll want this comparison data later.
- Write and upload your llms.txt. Plain text file at the root of your domain. Should describe your business, your services, your service area, and the URLs of your most important pages.
- Update your robots.txt. Explicitly allow GPTBot, ClaudeBot, Google-Extended, PerplexityBot, anthropic-ai, OAI-SearchBot, ChatGPT-User, CCBot, and Applebot-Extended. Many sites accidentally block these.
- Add schema.org markup. At minimum: Organization or LocalBusiness schema. Better: add Service, Offer, and FAQ schema for your key pages.
- Submit to citation sources. Get listed on Google Business Profile, Yelp, Bing Places, and 10-20 industry-specific directories. The goal is to appear consistently across the open web so AI models notice you exist.
- Rewrite your homepage and About page in a Q&A format. AI models quote content that's structured as direct answers. "What does [your business] do?" → clear answer. "Where are you located?" → clear answer. Don't make the AI work to extract your facts.
That's roughly 6-10 hours of focused work. It will put you ahead of most of your competitors. If you'd rather have us do it for you in 48 hours and hand you a finished package, that's what our $20 audit is for.
See where you stand in 30 seconds — free
Run a free AI visibility check on your URL. We grade your discoverability across ChatGPT, Claude, Gemini, and Perplexity instantly.
Run Free AI Check → $20 Full Audit →7. The Done-For-You Path: How We Help
If you don't have the time, or you'd rather skip the learning curve, Discovery ADO offers three tiers of done-for-you service:
$20 · AI Visibility Audit + Custom Fix Kit
A full audit across ChatGPT, Claude, Gemini, and Perplexity delivered as a PDF in 48 hours. Includes ready-to-paste code for your llms.txt, robots.txt, and schema markup — plus AI-optimized rewrites for your homepage and About page, a citation submission list, and a 5-step action plan. Start here →
$349 · ADO Setup (Done-For-You Install)
We install the complete Fix Kit on your site, submit you to 10 essential citation sources, and fully optimize your Google Business Profile. Completed in 5-7 days. No developer required on your end. Learn more →
$199/mo · ADO Growth (Ongoing Optimization)
Monthly AI visibility tracking, 2 new AI-optimized content pages per month, 5-10 new citations per month, schema updates as new standards emerge, quarterly competitor analysis, and direct support. Get details →
For multi-location businesses, brands wanting Wikipedia citations, or businesses that want a full content engine, we also offer ADO Authority at $499/mo. Talk to us →
8. Frequently Asked Questions
What is AI Search Optimization?
AI Search Optimization (also called Agent Discovery Optimization, Generative Engine Optimization, or Answer Engine Optimization) is the practice of optimizing a website so AI assistants like ChatGPT, Claude, Gemini, and Perplexity recognize, categorize, and recommend the business when relevant. It uses different signals from traditional SEO — including llms.txt, schema.org markup, robots.txt rules for AI crawlers, and citations across authoritative sources.
How is AI SEO different from regular SEO?
Regular SEO targets Google's ranking algorithm and focuses on backlinks, page speed, and keyword targeting. AI SEO targets large language models that generate answers from a fundamentally different process. A business can rank #1 on Google and still be completely invisible to ChatGPT, because the two systems use different signals. AI SEO emphasizes structured data, citations on trusted sources, llms.txt, and content that LLMs find quotable.
What is llms.txt?
llms.txt is a simple text file placed at the root of a website that tells AI crawlers what the site is about, who it serves, and how to interpret its content. It's the AI equivalent of robots.txt — but instead of restricting crawlers, it guides them. Most websites don't have one, which means a well-written llms.txt is currently a major competitive advantage. Look at our own llms.txt for a working example.
What's the difference between GEO, AEO, and ADO?
GEO (Generative Engine Optimization), AEO (Answer Engine Optimization), and ADO (Agent Discovery Optimization) all refer to roughly the same discipline. GEO emphasizes the generative nature of the AI. AEO emphasizes the answer-first experience. ADO is the term we use at Discovery ADO and emphasizes that AI assistants act as agents that "discover" and recommend businesses. The tactics overlap about 80%.
How long does AI SEO take to show results?
Technical changes like schema markup, llms.txt, and robots.txt updates can be picked up within days as AI models re-crawl. Content changes typically take 2-4 weeks to reflect in AI training data. Citation-building takes 4-8 weeks to show measurable impact. Most businesses see clear improvement in AI visibility within 4-8 weeks of implementing a proper AI SEO strategy.
Do I still need traditional SEO?
Yes. Google is still where most search happens. AI SEO complements traditional SEO — it doesn't replace it. The good news: about 60-70% of the work for AI SEO (clean schema, good content, fast load times, proper citations) also helps your traditional SEO. You're not picking one or the other.
Can I do this myself?
Absolutely. Most of AI SEO is straightforward once you know what to do. The DIY section above walks you through the foundation. If you'd rather have us do the diagnosis and hand you ready-to-paste code, that's what our $20 audit delivers.
Ready to be findable by AI?
Start with the free check. Most businesses score under 40/100. We'll show you exactly what to fix.
Run Free AI Check → Get the $20 Audit →