How do AI assistants decide which businesses to recommend?

Q: How do AI assistants decide which businesses to recommend?

AI assistants search the web, read the pages they find, and name the businesses they can identify, verify and trust most confidently. I call that confidence Machine Trust. It grows when your details are consistent, your content answers the question, and independent sources agree about you.

Where do AI assistants get their information?

Two places: what the model learned during training, and pages it fetches live. Google describes its AI features as retrieving pages from its search index (a technique called grounding) and running several related searches at once to build a fuller answer. ChatGPT search uses OpenAI's own crawler and other search providers.

Sources: Google Search Central: Optimizing your website for generative AI features · OpenAI Help Center: Publishers and developers FAQ

What signals matter most?

Ahrefs analysed 75,000 brands across ChatGPT, Google AI Mode and AI Overviews. YouTube mentions had the strongest link to AI visibility (about 0.74), followed by brand mentions across the web (0.66 to 0.71). The number of pages on a site barely mattered (about 0.19).

These are correlations, so they show what visible brands have in common rather than a guaranteed recipe.

Sources: Ahrefs: Top brand visibility factors in ChatGPT, AI Mode and AI Overviews

Do AI assistants pick businesses at random?

No. Answers vary between runs, but Ahrefs found the three assistants it studied largely name the same brands (an overlap of 0.779). The same well-documented businesses keep appearing because the AI has more confidence in them.

Sources: Ahrefs: Top brand visibility factors in ChatGPT, AI Mode and AI Overviews

What builds Machine Trust?

The AI Search Flywheel describes six stages: Be Understood, Be the Answer, Be Cited, Be Recommended, Be Chosen and Be Proven. Each one feeds the next, and real customer results feed back into the first.