How to Get Your Products Mentioned by ChatGPT, Gemini and Perplexity

"What''s the best [product] for [use case]?" used to be a Google query with ten blue links. Now it is a conversation that ends with three or four named products. If your product is not one of them, the comparison never happens.

How the recommendation is actually formed

Assistants build product answers from three inputs:

1. Retrieval — live search results fetched at answer time, weighted toward sources the system already trusts
2. Model knowledge — patterns learned from published text, where repetition across independent sources creates confidence
3. Structured data — product feeds, markup and merchant data where available

Notice what is missing: your ad spend, your social following, and how good your product actually is. The answer reflects what has been written about you, not what is true about you.

The three requirements

1. Be described consistently. One brand name, one product name, one spelling, everywhere. Inconsistent naming is the most common reason a product simply is not recognised as an entity.

2. Be described by others. Three or more independent published sources naming your product in the relevant category is the practical threshold for appearing as a candidate. One mention is noise.

3. Be machine-readable. Complete product structured data, crawlable specs, real availability and pricing, shipping and returns in HTML.

What to publish yourself

Your own content will not make an assistant recommend you, but it does make you easier to describe:

• A specification page per product with plain, factual attributes
• Comparison pages that honestly position you against alternatives
• A buying guide per category, structured with clear headings and direct answers
• An llms.txt at your root summarising what you sell

Direct-answer formatting matters: state the answer in the first sentence under each heading, then support it. Retrieval systems extract passages, not pages.

What to earn elsewhere

This is the part most stores skip. The inventory that assistants draw on:

• Category roundups ("best X for Y") on independent publications
• Product reviews with named reviewers
• Business and sourcing coverage that establishes the brand as a real company

Place these deliberately. Our marketplace shows each publisher''s AI Authority Score — an estimate of how much influence its content carries in AI answers — so budget goes to sources that are actually cited.

Measuring it

Run your priority questions against each major assistant monthly, record whether you are named and in which position, and track the share over time. Full method in how to measure AI search visibility.

Continue with Shopify SEO in the AI era and LLM SEO: the complete guide.

Or tell us your category and we will map which publishers are shaping the answers in it.