How AI assistants choose what to recommend
Ask ChatGPT or Perplexity for the best mattress for back pain, and it usually searches the web, reads a handful of pages and writes an answer from them. Google's AI Mode and AI Overviews do the same with Google's own search results and shopping data. The brands that get named tend to have three things: pages that answer the question clearly, product data the assistant can trust, and other sites that recommend them for that need.
Every assistant your buyers might ask
Shoppers don't all use the same assistant. ChatGPT, Claude, Perplexity and Grok search the web and cite the pages they used. Google's AI Overviews, AI Mode and Gemini draw on Google Search and its shopping data, and Copilot draws on Bing. Meta AI answers inside Facebook, Instagram and WhatsApp. On Amazon, Alexa for Shopping answers from listings and reviews, which Amazon listing optimization covers. A brand can show up in one and be missing from another, so the monthly check covers all of them.
What we work on
- Your own pages: clear answers near the top, with the key facts in plain text. Some AI crawlers only read a page's HTML, so reviews or details that an app loads later can go unseen.
- Product data: complete titles, attributes, prices and stock in Shopify and your product feeds. Assistants use this data to show products, and more of them now let people buy right in the chat.
- Mentions elsewhere: assistants look for a second opinion in reviews, comparison articles, forums and directories. We help you earn a place there.
Measured every month
Nobody controls what an AI assistant says. Answers change with the question, the person asking and the week. So each month we ask the assistants a fixed set of buyer questions and record who's mentioned and which sources they cite. We also add AI assistants to your GA4 tracking, so you can see the visits and sales they send. Most of this work also helps your Google rankings, so it pairs well with SEO.