How AI assistants choose what to recommend, and what moves it

Research and guides for the people who have to get a brand into the answers. Every claim names its source, and every post carries the date that source was last read.

6 postsRSS feed

Every post here was re-checked on 9 September 2026.

A magnifying glass over a brand silhouette with four labeled diagnostic paths

A brand that should be recommended for a category query is often absent from AI answers for one of five checkable reasons. Here's how to find which one.

Absence has five checkable causes, in this order: crawlers are blocked at robots.txt or the CDN, you are missing from the sources models cite, your pages never say plainly what you sell or its price, you name the category differently than buyers, or the model names you but gets a fact wrong. Most brands find it in the first two.

Read the post

  1. Education

    Why Review Sites Shape AI Recommendations More Than Your Website

    G2, Capterra and Trustpilot often carry more weight than brand websites in AI answers. Here's why, and what to do about it.

    7 min read

  2. Guide

    How to audit your brand's AI visibility in an hour

    A repeatable one-hour audit of your brand's AI visibility across ChatGPT, Gemini, Perplexity and Claude, with a prompt set and a scoring sheet.

    7 min read

  3. Education

    Training Data vs. Live Retrieval: How AI Models Answer

    Some AI models answer from training data, others retrieve live pages first. See how each works and what it means for AI visibility.

    7 min read

  4. Education

    Why Reddit Carries So Much Weight in AI Recommendations

    Reddit shapes what AI models say about brands more than most sources do. Here's why, how that's shifting, and how to participate without astroturfing.

    7 min read

  5. Guide

    What Makes AI Search Engines Cite a Page

    A practical checklist for what makes a page more likely to be cited by ChatGPT, Perplexity, and Gemini, grouped by five checkable attributes.

    7 min read