How to Get on the “Best Of” Lists That ChatGPT and Google AI Cite

To get on the “best of” lists that AI assistants cite, first find out which lists they cite for your customers’ questions (ask the questions, record the sources), then read each list to see what its author values, and send the author a short, checkable case for including you. Fix the basics that list authors check: a clear description, published prices where you can, and independent reviews. It’s outreach, it’s slow, and it works better than anything you can do on your own site, because when an assistant recommends three businesses it is usually repeating pages like these.

Why lists decide so many AI recommendations

A recommendation question needs a page that compares options and says which suits whom. That is what a “best X” list is. Ahrefs studied 750 top-of-funnel prompts across software, products and agencies and found lists were the largest category of cited page, at 43.8% (vendor study). Separately, Ahrefs found ChatGPT cited 88% of the pages it retrieved through web search, against about 2% of Reddit pages (1.4M prompts, vendor study), so lists that rank are lists that get read. The broader pattern holds too: in its 75,000-brand study, being mentioned across the web correlated with AI visibility far more than backlinks did.

Step 1: find the lists that matter to you

1. Write ten questions your customers would ask an assistant: “best [category] for [customer]”, “top [service] in [city]”, “[competitor] alternatives”.
2. Ask each in ChatGPT (with search), Perplexity, Google (AI Overviews or AI Mode) and Copilot, three times. Answers vary from run to run; in SparkToro’s test, the same brand list almost never came back twice.
3. Copy every cited URL into a sheet. Mark the ones that are lists, and count how often each appears.
4. Add the lists that rank on page one of Google and Bing for the same questions; assistants that search will find them.

You’ll usually end up with 10–30 lists, of which five or six come up again and again. Those are your targets. The competitor guide has a tracking sheet for this step.

Step 2: read each list before you pitch it

Step 3: pitch with a checkable case

List authors get a lot of email. The pitches that work are short, specific and easy to verify. A template:

Subject: A candidate for “Best [category] for [customer]”

Hi [name],

I read your [list title] (updated [month]). One option you may not have seen:
[Business], [one sentence: what, for whom, where].

Why it might fit your criteria:
- [Hard fact: price, a feature, a number, a specialism]
- [Independent proof: review rating with link, award, press coverage]

Happy to set you up with [a demo / a sample / a test account], no strings.
If it doesn’t fit, no reply needed.

[Name], [role], [business]

Then follow up once, a week later, and stop. If a list already includes you but gets a fact wrong (an old price, a discontinued product), ask for a correction; that is the easiest yes you’ll get, and wrong facts on cited lists are a common source of wrong AI answers (fixing wrong AI information).

Step 4: give authors a reason to include you

• Publish your prices, or a starting price. Authors comparing options skip the ones they can’t price.
• Have independent reviews on the platforms your category uses (reviews and AI recommendations).
• Make one thing clearly yours: the best option for a specific customer, budget or use. Lists are organised by “best for”; give them a slot to put you in.
• Keep a press page with your logo, a one-paragraph description, screenshots or photos, and a contact.

Pay-to-play lists and your own lists

Some lists sell placements. If you buy one, it should be labelled as paid; an undisclosed paid endorsement can break advertising rules in many countries, and the FTC’s 2024 rule on fake reviews and testimonials targets deceptive endorsements in the US. Ahrefs also found 35% of cited lists sat on low-authority domains, many apparently built for links, and expected that share to fall as assistants get better at judging quality. A paid spot on a page assistants learn to ignore is money spent twice.

Publishing your own list is a separate question, and the evidence splits. Ahrefs saw no clear downside for brands that list themselves; in Lily Ray’s test of 100 B2B queries, 69% of the self-promotional lists AI Overviews cited (224 of 323) belonged to brands the answer didn’t recommend. Google’s guide is blunt about the wider tactic: “seeking inauthentic ‘mentions’ across the web isn’t as helpful as it might seem”. If you publish comparisons, make them honest: name competitors accurately, say who each suits, and don’t rank yourself first because you can.

Measure it

Re-run your ten questions monthly and count two things: how many cited lists now include you, and how often the assistants name you. The first moves before the second. Our AI visibility checker, rolling out engine by engine, shows who is recommended instead of you and which sites each answer cites, which is your target list refreshed.