Why “best X” answers cite lists, and what that leaves for your product pages

Ask ChatGPT for the best treadmill for a small apartment and look at the sources under the answer. Most of them won't be treadmill brands. They'll be list posts, "the 9 best treadmills for small spaces", written by someone else.

That isn't an accident, and it changes what a store should expect from its own website.

What the assistants cite

Ahrefs analyzed 26,283 source URLs that ChatGPT cited across 750 top-of-funnel prompts, covering software, services, and physical products such as cameras and treadmills. "Best X" list posts made up 43.8% of all cited pages, the largest single type. 35% of those lists sat on low-authority sites.

BuzzStream looked at 12,000 answers across Google AI Mode, AI Overviews, Gemini, and GPT-5.4-mini. When a brand was named in a list-style answer, only 7.2% of those mentions came with a citation. When the question was about a single brand ("is X any good?"), 39% did. Across everything, 80% of citations pointed at other people's websites.

Why lists win the broad question

A broad question asks the assistant to compare. A list post has already done the comparing, in a format that's easy to lift: a name, a reason, a verdict. The assistant can repeat it with a citation attached. A single product page, however good, only speaks for itself.

What your product data still decides

So the shortlist for a broad question is largely decided off your site. Three things are still decided on it.

1. Whether you're eligible. A list can name your product, but the assistant still has to find your page or feed, read it, and match it to the shopper's requirements. Missing attributes, blocked crawlers, or a price that disagrees with your feed all cost you here.
2. Whether the answer about you is right. When a shopper asks about you specifically, the answer leans much more on sources tied to you. If your own pages are vague about shipping, returns, or sizing, the answer will be vague or wrong.
3. Whether the click converts. The shopper who clicks through lands on your product page to confirm what they were told. That page closes the sale, or doesn't.

Off-site mentions build on this rather than replacing it. A list post that names your product sends the shopper, and the assistant, to check your page. A specific page confirms the mention; a vague one wastes it.

The order of work

That's why we start with product data and work outward. It's the part a store fully controls, it has to be right before any mention can pay off, and it's where the sale happens. Lists, reviews, forums, and video come next, and they work better on a catalog that can back them up.

What you can't control

Which lists exist, who writes them, and which ones an assistant trusts on a given day. The same question asked twice can return different sources. Anyone promising a fixed spot in a "best X" answer is guessing.

Try it on your category

1. Ask three assistants the broad question your best product should win.
2. Open every source cited, and count how many are lists, how many are retailers, and how many are brand sites.
3. Then ask about your own brand by name, and check whether the answer gets your shipping, returns, and sizing right.