How AI assistants decide which businesses to recommend
AI assistants combine three signal categories when choosing which businesses to recommend: (1) the entity data they have indexed about your business — your website's schema, Wikipedia/Wikidata entries, and structured listings on review platforms; (2) the natural-language descriptions, FAQs and direct-answer content on your own site that they can quote with confidence; and (3) third-party signals like review volume, mention frequency in cited articles, and presence in trusted directories. The businesses recommended most often are the ones strong on all three.
Signal 1: Entity data
AI assistants build internal knowledge graphs of businesses, services and locations. The cleaner your entity data, the more confidently they can place you in those graphs. The big inputs:
- Schema.org markup on your website — Organization, LocalBusiness, Service — with name, URL, sameAs, contactPoint, areaServed, services and price range explicit.
- Wikipedia / Wikidata entries (where applicable) for your brand or founders.
- Consistent NAP (Name, Address, Phone) across Google Business Profile, Bing Places, Apple Maps, Yelp, Trustpilot and industry directories.
Signal 2: Quotable on-site content
AI assistants prefer to cite specific, verifiable, attributable text. That means:
- Direct-answer paragraphs — 2–3 sentences answering the page's core question, immediately after the H1.
- FAQ blocks in conversational language, schema-marked as FAQPage.
- Specific numbers — delivery times, prices, success rates, years in business. AI prefers '14-day delivery, from £300' over 'fast and affordable'.
- Clear entity mentions — your business name, services and locations stated naturally in body copy, not hidden behind generic pronouns.
Signal 3: Third-party validation
Even an immaculately optimised site benefits from external corroboration:
- Reviews — both volume and recency. AI assistants weight Google and industry-specific reviews (Trustpilot, Checkatrade, etc).
- Citations in trusted articles — being mentioned in industry publications, blogs or guides indexed by major search engines.
- Directory presence — sector-relevant listings (Law Society directory, Gas Safe register, BUPA finder, etc) that AI assistants treat as authoritative.
How to engineer your way in
Start with what you control: your website. Implement the GEO/AEO foundations first — they're the highest-leverage move. Then progressively address third-party signals: reclaim and optimise your Google Business Profile, automate review collection, get listed in 5–10 sector directories you respect, and pursue 2–3 mentions in industry publications per quarter.