A growing share of buying decisions in the US now starts with a question typed into ChatGPT, Perplexity, or Google — where an AI Overview answers before a single traditional result appears. When those answers recommend businesses, they cite sources. Being one of those cited sources is the new version of ranking #1, and most companies haven't even started competing for it.

How AI search decides who to cite

AI assistants don't crawl the web the way you might imagine. They lean on three things: their training data, live retrieval from search indexes, and structured signals that make a business easy to identify and trust. In practice, that means citations flow toward businesses that are:

  • Unambiguous: the AI can tell exactly who you are, what you do, and where you operate — with no conflicting information across the web
  • Specific: your content answers a real question directly, with numbers, steps, or firsthand detail — not 800 words of throat-clearing
  • Corroborated: other credible sources (directories, press, reviews, industry sites) say the same things about you that your own site says

AI search optimization: the concrete steps

1. Fix your entity signals first

Search engines and AI models resolve businesses as entities. Inconsistent names, addresses, service descriptions, or category labels across your site, Google Business Profile, LinkedIn, and directories make you a risky citation — so the AI simply picks a competitor it's more certain about. Audit for one consistent name, description, and service list everywhere your business appears.

2. Publish answers, not essays

Look at what actually gets cited in AI Overviews: content that resolves a specific question in the first two paragraphs, then earns depth afterward. Structure every page so a machine can lift a clean answer — a clear H2 phrased like the question, an immediate direct answer, then supporting detail. This is the same structure that wins featured snippets, and it's why pages built this way get cited disproportionately.

3. Use schema markup everywhere it applies

Structured data (Organization, Service, FAQPage, Article schema) is how you hand AI systems your facts in a format they can't misread. It's low-effort, invisible to visitors, and dramatically underused by small businesses — which makes it one of the few remaining cheap advantages in search.

4. Add an llms.txt file

A newer convention, llms.txt gives AI crawlers a plain-language summary of who you are and what your key pages contain. It takes an hour to implement and puts you ahead of the overwhelming majority of US businesses that don't know it exists.

5. Earn third-party corroboration

AI systems weight consensus. A business described the same way by its own site, two industry directories, a local news mention, and fifty Google reviews is a safe citation. Reviews matter here beyond reputation — they're independent evidence of what you actually do and where.

Traditional SEO asked "how do I rank for this keyword?" AI search optimization asks "why would a cautious machine vouch for my business?" The second question has stricter standards.

How to measure whether it's working

  • Ask the assistants directly: monthly, run the questions your customers ask ("best [your service] in [your city]", "how much does [your service] cost") through ChatGPT, Perplexity, and Google's AI mode, and log who gets cited
  • Watch referral traffic: in GA4, traffic from chatgpt.com, perplexity.ai, and similar referrers is small but converts remarkably well — these visitors arrive pre-sold by a recommendation
  • Track branded search volume: people often see you cited in an AI answer, then Google your name — rising branded search is a lagging indicator that citations are happening

Why this matters more for smaller businesses

AI answers compress ten blue links into three citations. That's brutal for whoever holds positions four through ten — but it means a well-structured small business can leapfrog bigger competitors who are still optimizing for a results page that fewer buyers ever scroll. The window where this is a first-mover advantage is open now, and like every search advantage before it, it will close once everyone catches up. Paid channels face the same squeeze from rising costs, which is why we've written about getting Google Ads budgets right — but organic AI visibility is the compounding asset.

If you want to know whether AI assistants can even see your business today — and what it would take to become the answer they give — that's exactly the kind of audit we run before recommending any strategy.