Get Cited in AI Best-Of Lists (Stop Publishing Your Own)
Ahrefs studied 26,283 source URLs cited by ChatGPT. Best-of lists made up roughly 43.8% of them, the single biggest content type in the data. Self-published “best X” pages often backfire: they earn the citation while a rival named inside earns the recommendation. Getting placed in an already-cited list beats writing another one. It is a free method any business can run today.
We place links for a living, so we tested this ourselves before writing about it. We checked ten published “best GEO agency UK” listicles that ChatGPT and Perplexity already draw from. GeoLinks appeared in zero of them, the exact gap this post exists to close.
Key takeaways
- Best-of lists are roughly 43.8% of ChatGPT’s cited sources, the largest single content type (Ahrefs, 26,283 URLs).
- Self-published “best X” pages often get cited, then lose the recommendation to a rival named inside the same list.
- Placement inside an already-cited list beats publishing a new one, because the host page is already trusted and crawled.
- Finding the right lists is free: ask the engine your buyer’s question, then open every source it cites.
- We audited ten UK “best GEO agency” listicles already cited by ChatGPT and Perplexity. GeoLinks was in none of them.
Why best-of lists dominate AI citations
Ahrefs pulled 26,283 URLs cited across ChatGPT answers and sorted them by content type. Best-of and comparison listicles came out on top at about 43.8%, ahead of product pages, guides and news. When someone asks ChatGPT to recommend a product, an agency or a tool, the model reaches for a ready-made comparison. That page already ranks several options side by side.
That makes sense from the model’s side. A ranked list does the comparison work the engine would otherwise have to assemble from scratch. It is a shortcut source, and AI engines take shortcuts constantly.
The self-promotion trap
Here is the part most content plans miss. Ahrefs also tracked what happens when a brand publishes its own “best conferences” or “best tools” page and gets cited. In a meaningful share of those citations, the AI used the page purely as a research source. It then recommended a different name mentioned inside it. The self-promoter earned the citation and handed the recommendation to a rival on its own list.
Peec AI ran a parallel analysis across 232,000 citations and reached the same conclusion from a different data set. Self-promotional lists get read, not necessarily believed. Writing your own best-of page is not wasted effort, but it is not the highest-leverage move either.
The free method: find lists AI already cites
You do not need a paid tool for the first pass. You need the engines themselves, the same method that works for any AI citation gap.
- Write down the exact question a buyer would ask, such as “best CRM for a small agency” or “best GEO agency UK”.
- Ask it in ChatGPT, Perplexity, Gemini and Microsoft Copilot. Source pools differ by engine, so check more than one.
- Open every source cited and note which ones are ranked lists rather than single-product pages.
- Check whether you are named. If not, that list is your pitch target.
- Reach out to the list owner with a genuinely relevant addition, not a generic outreach template.
That shortlist beats a dozen cold pitches to random blogs. The engine has already told you which pages it treats as trustworthy sources for your exact topic.
Placement vs publishing your own: the comparison
| Approach | Time to citation | Who gets the recommendation | Effort |
|---|---|---|---|
| Get placed in an already-cited list | 4 to 8 weeks | You, if the placement is genuine | Low to medium |
| Publish your own “best X” list | 2 to 3 months | Often a rival named inside your own list | Medium to high |
| Do both | Ongoing | You, across more source pages | Medium |
Neither route is a silver bullet alone. A tracker that only tells you where you are missing will not fix the gap. We cover that theme in GEO that doesn’t work. The fix is the placement itself, on a page the engine already trusts.
Our own audit: zero of ten UK “best GEO agency” listicles
We asked ChatGPT and Perplexity variations of “best GEO agency UK” and pulled every listicle source cited across both engines. We then checked ten separately published UK “best GEO agency” articles that showed up as sources. GeoLinks was named in none of them. That is despite running placements and a 12-month link guarantee competing agencies in those same lists do not offer.
That gap is now our own to-do list, and it is the same gap most businesses have without knowing it. The listicle placement method described above applies whether you sell software, garden ornaments or GEO services. If nobody has pitched the list owner yet, the slot is open.
Placement is a form of niche edit
Listicle placement is not a separate tactic on its own. It is the same principle behind a niche edit, an addition to a page that already exists. It applies to the one content type AI engines cite most. The host page is indexed, trusted and already inside the engine’s source pool. A genuine addition inherits that trust rather than starting from zero.
This is also why the placement has to be real. A padded, irrelevant entry gets removed at the next update and earns nothing. A useful, specific entry the list owner wants to keep tends to survive refreshes. It stays in the citation pool for months.
Multi-engine reality check
ChatGPT and Perplexity lean on listicles hardest for commercial and comparison queries, but the pattern is not ChatGPT-only. Claude, Gemini, Microsoft Copilot, Grok, Google AI Mode and AI Overviews all pull from ranked comparison pages. The exact source mix shifts by engine. We break down which sources each engine favours in the per-engine citation sources playbook. Treat this as a starting method, not a ChatGPT-only checklist.
Common questions
What share of ChatGPT citations are best-of lists?
Ahrefs studied 26,283 source URLs cited by ChatGPT. Listicles, mostly best-of and comparison lists, make up about 43.8% of citations, the single largest content type ChatGPT pulls from.
Should I write my own best-of list to get cited by ChatGPT?
Ahrefs found self-published best lists often backfire. AI engines cite the page, then recommend a competitor named inside it instead of the list’s own publisher.
How do I find best-of lists that AI engines already cite?
Ask your buyer’s exact question in ChatGPT, Perplexity, Gemini and Copilot, then open every cited source. The listicles among them are your pitch shortlist for a placement.
Does getting placed in a listicle guarantee an AI recommendation?
No, but it removes the biggest barrier. The engine already trusts and crawls that page. A placement inside it earns a citation far faster than a brand-new article would.
How is listicle placement different from a niche edit?
A niche edit adds a link inside an existing article. Listicle placement is the same principle applied to one content type, a ranked comparison page an engine already trusts.
Which AI engines cite best-of lists the most?
ChatGPT and Perplexity lean on listicles heavily for commercial queries. Claude, Gemini, Copilot, Grok and Google AI Mode all pull from comparison pages too, though source mixes vary by engine.
How long does listicle placement take to earn a citation?
Often four to eight weeks once the list owner adds your entry and refreshes the page. That is faster than most brand-new content, because the host page is already indexed and trusted.
Related reading
- Niche edits for GEO covers the wider principle behind placement on already-trusted pages.
- Which sources each AI engine actually cites breaks the source mix down by engine.
- Why isn’t my brand in ChatGPT covers the other reasons a business gets skipped entirely.
Want placements on the lists AI already cites, not a new one nobody reads? Our niche edits service finds and pitches the pages your buyers’ AI answers already trust, backed by a 12-month guarantee. Not sure where you stand yet? Run the free AI visibility check.