You’ll be able to simply repeat this evaluation your self.
Merely run an open database search in Model Radar, navigate to the “Quotation Pages” report on your desired AI assistant, and export the highest 1,000 cited pages.
Subsequent, run the cited pages by way of a batch evaluation software to get extra natural information for every URL.
Claude also can aid you analyze that information and create scripts for Google Colab to seize and parse content material freshness indicators for every cited URL.
That is precisely what occurred. That is what I discovered…
ChatGPT citations should not restricted to encyclopedias or editorial pages. Nonetheless, Wikipedia stands out as probably the most cited content material kind by a large margin over different content material sorts.
The remaining citations are unfold throughout academic content material, homepages, app listings, blogs, and different codecs.
I used Claude to categorize this information. Though it will not be 100% sure, it may give you a directionally understanding of probably the most cited pages on ChatGPT.
The highest 1,000 cited URLs, damaged down by content material kind, seem like this:
Wikipedia performs a central position in ChatGPT’s quotation conduct. The assistant is clearly on the lookout for a structured reference-style supply that summarizes the subject comprehensively and predictably.
Academic pages and homepage pages are additionally displayed recurrently.
However probably the most fascinating factor is that there are few alternatives to really do it. talked about With these excessive profile quotes.
Our Director of Content material Advertising and marketing, Ryan Regulation, calls Wikipedia, homepages, and app retailer pages “lifeless” citations in that they cannot be simply influenced.
In accordance with the above classification, solely 32.3% of ChatGPT’s high 1,000 citations are pitch/outreach worthy, or a minimum of impactful offsite.
In different phrases, roughly two-thirds of ChatGPT’s high citations are successfully outdoors the scope of conventional outreach methods. These are group pages, reference websites, and different “lifeless” citations that may’t realistically have an effect.
Pages that ChatGPT cites are sometimes newer and fewer established in search, suggesting that recency impacts quotation visibility, however it’s not the one issue.
For pages with detectable dates (roughly 40% of the full), the median age was 55 days (just below 2 months).
Essentially the most cited pages by 12 months of publication are…
And this is the breakdown of pages with detectable modification dates…
| up to date | % |
|---|---|
| 30 days | 42.7% |
| 1-6 months | 20.0% |
| 6-12 months | 11.9% |
| Greater than 1 12 months | 25.4% |
This information exhibits a transparent recency bias and is per the next findings. Metehan Yesilurt’s recent research.
Yesyurt is URL_freshness_score ChatGPT prioritizes new content material and cited analysis exhibiting that artificially updating publication dates can enhance AI rating positions by as a lot as 95 locations.
That stated, my information exhibits that almost 1 in 4 cited pages have been up to date greater than a 12 months in the past, suggesting that older content material that’s helpful and well-structured can keep visibility in AI.
Practically one-third of ChatGPT citations level to pages with out conventional search visibility.
Potential explanations for this are:
freshness: Some pages with zero key phrases could also be new content material that has not but been ranked by search engines like google and yahoo. ChatGPT could uncover and cite new content material earlier than accumulating search rankings.
Area of interest subjects: Citations could cowl particular long-tail subjects whereas minimizing search demand. You reply the query precisely, however you do not get a lot search site visitors. This additionally applies to fan-out queries. Even when somebody asks ChatGPT a couple of widespread subject, the dialog usually branches out into extra particular subtopics, with few pages immediately focusing on these angles.
Completely different findings: ChatGPT tends to entry and consider content material in a different way than search engines like google and yahoo, and seems to prioritize accuracy, freshness, and relevance over reputation indicators corresponding to backlinks.
In actuality, it is probably a mix of all three. Contemporary content material, area of interest subjects, and different high quality indicators all contribute.
web page known as do The ranks show some fascinating patterns. The important thing factors are:
- Web site permissions actually matter: 65.3% DR 81+, DR median 90
- However web page permissions should not: 67.3% have a UR between 0 and 10, with a median UR of solely 6. ChatGPT cites pages from authoritative domains, however not essentially probably the most linked pages on these domains.
- Pages are keyword-rich: Median 279 key phrases
- Most have robust backlink profiles: Median referring area 70
- Half have excessive search visibility: 52.1% ranked within the high 3 for key phrases
- A low privilege exception exists: 11.7% have DR 0-20
ChatGPT extremely prioritizes pages from high-authority domains which have robust backlink profiles and rank properly in search, however cites lower-authority sources for sure related content material.
abstract
ChatGPT’s high citations are skewed towards new content material, with almost a 3rd of pages with no natural visibility, and a majority of referring websites that may’t be simply influenced.
Information exhibits that your finest guess for growing visibility on ChatGPT is to get talked about in recent, particular content material that falls into the influential class.
You should utilize Model Radar to seek out outreach alternatives. Cease by your market or area of interest to entry experiences on the quotation web page and discover high blogs, publications, and evaluate websites.
