Do "best of" lists get you recommended by AI?
Ask ChatGPT for "the best roofer in Tampa" or "the best CRM for a small team" and the sources behind the answer are mostly lists. Here is how often AI engines cite "best of" pages, what happens when a company ranks itself first, the experiment that showed how self-promotion backfires, what Google asks of any "best" claim, and how to get onto lists honestly.
Yes, for commercial questions. In Ahrefs' study of 750 ChatGPT prompts, "best X" lists made up 43.8% of cited pages. Wix measured listicles at 40.86% of citations for commercial prompts. But a later Ahrefs experiment found 43% of answers that cited a self-promoting page still recommended someone else. Third-party lists did most of the work.
What are the key facts about "best of" lists and AI?
| Fact | Detail | Source |
|---|---|---|
| Lists dominate ChatGPT's sources | "Best X" blog lists were 43.8% of all page types cited across 750 top-of-funnel prompts in software, products and agencies (26,283 source URLs) | Ahrefs (third-party) |
| Same pattern at scale | Blog lists were 43.83% of ChatGPT's sources on average, more than three times the next format, in a study of 1.4 million prompts | Ahrefs (third-party) |
| Listicles lead across engines | Listicles 21.9%, articles 16.7%, product pages 13.7% of 1,056,727 citations from 75,000 answers in ChatGPT, Google AI Mode and Perplexity | Wix Studio AI Search Lab (third-party) |
| Commercial questions | Listicles took 40.86% of citations for commercial-intent prompts, 86.7% above their average | Wix Studio AI Search Lab (third-party) |
| Self-ranking is common | In 169 of 250 "best X software" results checked (67.6%), a list ranked its own publisher number one | Ahrefs (third-party) |
| Self-ranking can surface | When a software publisher was recommended, its own self-promotional list appeared in more than a third of those ChatGPT responses | Ahrefs (third-party) |
| Authority isn't required | 35% of cited best lists sat on low-authority domains | Ahrefs (third-party) |
| Lists are fresh | 79.1% of 1,100 cited lists had been updated in 2025; 26% in the previous two months | Ahrefs (third-party) |
| Self-promotion backfires | Across 9,886 answers from ChatGPT, Gemini, Perplexity and Copilot, 43% of answers citing Ahrefs' conference-promoting pages never named the conference | Ahrefs experiment (third-party) |
| Third parties do the work | Of new mentions in slots where Ahrefs had been absent, 94% came from answers citing third-party pages | Ahrefs experiment (third-party) |
| Google on "best" claims | Explain why you consider a product the best, with first-hand evidence |
How often do AI engines cite "best of" lists?
More than any other kind of page when someone is choosing what to buy. Three studies, from two companies, agree.
- Ahrefs, 750 prompts. Ahrefs, an SEO software company, ran 750 top-of-the-funnel ChatGPT prompts in three categories: software, products and agencies. Across the 26,283 source URLs behind the answers, "best X" blog lists were 43.8% of all page types. Ahrefs
- Ahrefs, 1.4 million prompts. In a much larger study of why ChatGPT cites one page over another, blog lists such as "The 10 Best Web Design Agencies" were 43.83% of ChatGPT's sources on average, more than three times the next format. Ahrefs
- Wix, 1 million citations. Wix Studio's AI Search Lab analysed 75,000 AI answers and 1,056,727 citations from ChatGPT, Google AI Mode and Perplexity across five industries, including home repair and professional services. Listicles were the most-cited page type at 21.9%, ahead of articles at 16.7% and product pages at 13.7%. For commercial-intent prompts, listicles rose to 40.86% of citations, 86.7% above their average. Wix
Wix found that the intent of the question predicted the page type better than the industry or the engine did. Informational questions pulled articles. Commercial questions, "best", "top", "which should I hire", pulled lists.
Does it work to rank your own business first on your own list?
Sometimes, and it is very common. Ahrefs checked 250 "best X software" searches and found that 169 (67.6%) included a list where the company that wrote it ranked itself number one. When ChatGPT recommended a software company, that company's own self-promotional list appeared in more than a third of those responses. Ahrefs
Authority was not a gate. 35% of the cited best lists were on low-authority domains, many of which Ahrefs described as "highly questionable" and built mainly for link building or citations. Freshness was: of 1,100 cited lists, 79.1% had been updated during 2025 and 26% in the previous two months.
Ahrefs' conclusion from that study was that software companies and agencies should publish self-promotional "best" lists, pointing out that brands such as Shopify, Slack, Salesforce, HubSpot and Superside already do, with no apparent downside for rankings or AI citations.
When does a self-promotional list backfire?
When the AI reads your page and recommends someone else. Ahrefs tested this directly. Between 7 February and 31 May 2026 it published 34 self-promotional pages on 5 domains, some promoting its Brand Radar tool and some promoting its Ahrefs Evolve conference, then tracked 9,886 answers from ChatGPT, Gemini, Perplexity and Copilot. Ahrefs
- The conference pages often backfired. Among answers that cited one of the conference-promoting pages, 43% never mentioned Ahrefs Evolve. The engines used the page as research and recommended other conferences.
- The tool pages mostly held. Among answers citing one of the Brand Radar pages, only 11% failed to mention Ahrefs. Ahrefs was already well known for that product.
- Third parties did the breaking in. Where Ahrefs had been absent from an answer before, only 6% of its new mentions came from answers citing its own pages. The other 94% came from answers citing third-party content.
- Citations were fleeting. In about a quarter of cases, a page was cited once for a query on one engine and never again.
The lesson for a small business: a list on your own site can help an AI engine understand your category, but it rarely convinces the engine that you are the answer. If you are not already known for something, other people's pages are what get you named.
What does Google say about "best" lists?
Google treats a ranked list as a review, and asks for evidence. Its reviews system "aims to better reward high quality reviews": content with insightful analysis and original research, written by experts or enthusiasts who know the topic well, rather than thin content that simply summarises a set of products or services. Google
Its guide to writing reviews is direct about "best" claims: if you recommend something as the best overall or the best for a certain purpose, share with the reader why you consider it the best, and include supporting first-hand evidence. It also says the guidance applies to ranked lists, and that because lists are shorter you may want to show expertise more concisely, for example by citing results and including original images from your own tests. Google
Google's AI Overviews and AI Mode run on its core Search ranking and quality systems, so a list that the reviews system rates poorly is unlikely to be the one Google's AI quotes. Google
Google's spam policies also define site reputation abuse: publishing third-party pages on a host site to take advantage of the host's ranking signals. Google Paying to have your business slotted into a "best of" article on a big site that does not otherwise review your trade is the pattern that policy targets.
How do I get my business onto the lists AI cites?
This is our advice, built on the studies above and Google's review guidance. The goal is to be named on honest lists that other people wrote, and to publish your own comparison only when it is genuinely useful.
- Find the lists AI already cites. Ask ChatGPT, Gemini, Perplexity and Google "best [your service] in [your city]" and "top [your service] for [a customer type]". Write down every list behind the answers. That is your target list.
- Check whether you belong on each one. Read the criteria. If a list ranks roofers by warranty length and review count, you know what the writer needs from you.
- Pitch the writer with facts, not money. Send your years in business, licence numbers, review count and rating, service area and one result a customer agreed to share. Never pay to be added; on a large site that pattern is what Google's site reputation abuse policy targets.
- Get onto local and trade lists. Local news "best of" polls, chamber of commerce directories, trade association member lists and supplier "certified installer" pages are lists too, and they are written by third parties.
- Earn the reviews lists are built from. Many lists rank by rating and review count. Ask every happy customer for a review on Google and on your trade's main review site.
- Publish your own comparison only if it helps. A page comparing the options a customer actually weighs, such as "repair vs replace", "metal vs shingle" or "us vs a franchise", with your real prices and honest trade-offs, is useful. A list of competitors with you at number one and no evidence is not.
- If you do rank yourself, say so. Put "We wrote this and we are on it" at the top, explain your criteria, and include first-hand evidence for every "best" claim, as Google asks.
- Keep lists fresh. 79.1% of cited lists had been updated within the year. Re-check yours, and the ones you appear on, every quarter.
What should be on my "best of" list checklist?
| Check | Done when | Source |
|---|---|---|
| Cited lists found | You have a written list of the "best of" pages behind AI answers for your core questions | Our advice |
| On third-party lists | You appear on at least three lists you did not write | Ahrefs experiment (third-party) |
| No paid placements | You have not paid to be added to any list on a site that does not otherwise review your trade | |
| Review count growing | New reviews arrive monthly on Google and your trade's review site | Our advice |
| Own comparison is useful | Your comparison page compares real options with real prices and trade-offs | |
| Self-ranking disclosed | Any list ranking you first says who wrote it, on the first line | Our advice |
| Evidence for "best" | Every "best" claim carries a reason and first-hand evidence | |
| Updated this quarter | Your own list and its facts were checked in the last 90 days | Ahrefs (third-party) |
| Recommendation, not just citation | When AI cites your page, the answer also names you | Ahrefs experiment (third-party) |
How do I know if lists are working for me?
Run your customer questions every month and check two things separately: which lists are cited, and whether you are named in the answer. A citation without a mention means the engine read your page and recommended someone else, the backfire Ahrefs measured.
The Prompt Simulator runs one question across ChatGPT, Gemini, Perplexity and Google's AI answers and lists every source each one cites, so you can see which lists are doing the work. Our free AI visibility check shows whether you are named today. For the reviews those lists are built from, see Do reviews affect AI recommendations?.
Sources
- Do Self-Promotional "Best" Lists Boost ChatGPT Visibility? Study of 26,283 Source URLs — Ahrefs. Read Oct 5, 2026.
- Self-Promotional Content Works—Until It Backfires (AI SEO Experiment) — Ahrefs. Read Oct 5, 2026.
- Why ChatGPT Cites One Page Over Another (Study of 1.4M Prompts) — Ahrefs. Read Oct 5, 2026.
- The content types most cited by LLMs — Wix Studio AI Search Lab. Read Oct 5, 2026.
- Google Search's Reviews System — Google Search Central. Read Oct 5, 2026.
- How To Write Reviews — Google Search Central. Read Oct 5, 2026.
- Google's Guide to Optimizing for Generative AI Features on Google Search — Google Search Central. Read Oct 5, 2026.
- Spam policies for Google web search — Google Search Central. Read Oct 5, 2026.
Common questions
Does ChatGPT cite listicles?
More than any other page type for buying questions. In Ahrefs' study of 750 top-of-funnel ChatGPT prompts, "best X" lists were 43.8% of the 26,283 cited pages. A separate Ahrefs study of 1.4 million prompts measured blog lists at 43.83% of ChatGPT's sources on average. Ahrefs Ahrefs
Are listicles cited by Google AI Mode and Perplexity too?
Yes. Wix Studio's AI Search Lab analysed 1,056,727 citations from ChatGPT, Google AI Mode and Perplexity and found listicles were the most-cited page type at 21.9%, rising to 40.86% for commercial-intent prompts. Perplexity also drew 17% of its citations from discussions such as Reddit and forums. Wix
Should I publish a list that ranks my own business first?
Only with disclosure and evidence. Ahrefs found self-ranking lists in 67.6% of 250 software searches and saw them cited. But in a 2026 experiment, 43% of answers citing Ahrefs' own conference-promoting pages recommended other conferences, and 94% of new mentions where it had been absent came from third-party pages. Ahrefs Ahrefs
Do I need a high-authority website for my list to be cited?
No. In Ahrefs' study, 35% of the best lists ChatGPT cited were on low-authority domains. Freshness mattered more visibly: 79.1% of 1,100 cited lists had been updated during 2025. Ahrefs
What does Google require from a "best" list?
Evidence. Google's review guidance says that if you call something the best overall or best for a purpose, explain why and include first-hand evidence, and that this applies to ranked lists. Its reviews system rewards original research over thin summaries of products or services. Google Google
Is it OK to pay to be included in a "best of" article?
It is risky. Google's spam policies define site reputation abuse as publishing third-party pages on a host site to exploit its ranking signals. A paid slot on a big site that does not otherwise review your trade fits that pattern, and Google's AI features run on the same ranking and spam systems as Search. Google
Related pages.
See which lists AI cites for your trade.
Enter your domain. The free AI visibility check shows where ChatGPT, Gemini, Perplexity and Google's AI answers mention you, and which pages they cite instead.