The Most Cited Domains in AI Search: Full Breakdown [58M Citations Analyzed]

We analyzed 58.6 million citations in LLMs like ChatGPT and Gemini from October 2025 through March 2026. See how citation share shapes AI visibility and brand authority in search.

Not every website gets cited by AI equally. Across 58.6 million citations analyzed from October 2025 through March 2026, a small group of domains captures a disproportionate share of the references that LLMs attach to their answers. Those citations shape what people read, which brands they encounter, and whose authority gets reinforced every time someone asks an AI a question.

Wikipedia leads the overall ranking at 3.4% citation share. YouTube and Reddit follow. After the top three, the list becomes more instructive: Alibaba sits at #6 tied with TechRadar, gaming and tech review outlets claim five spots in the top 20, and social platforms, including LinkedIn, Instagram, Facebook, and X cluster tightly in positions 5 through 14.

Citation share is the percentage of AI answers that reference a given domain. It determines who shows up when an AI constructs a response. Here’s the full top 20. 

Key Takeaways

• Wikipedia leads all LLMs with 3.4% citation share across 58.6 million citations, up from 2.2% in our previous study.
• 13 of the top 20 most-cited domains turned over between our February 2025 and October 2025 studies; the biggest shift we have recorded.
• News publishers have lost significant ground. Eight major outlets that held top 20 positions 6 months ago no longer appear. Gaming and tech review sites have taken their place.
• Social and community platforms remain the highest-influence tier for brands, with Reddit, YouTube, LinkedIn, Instagram, Facebook, and X all in the top 15.
• Roughly 74% of the most-cited domains are ones where marketing activity can directly affect citation share.
• Industry rankings diverge sharply from the overall list. TechRadar commands 8.86% citation share in B2B SaaS CRM. Kaja Beauty holds second place in Skincare and Cosmetics ahead of Alibaba and Instagram.
• Citation share compounds. Brands in the top quartile receive more than 10 times the AI citations of those in the next quartile down.

Top 20 Most Cited Domains in AI

Source: Goodie AI analysis of 58,642,927 citations. Oct 2025 – Mar 2026. Models: ChatGPT, Gemini, Claude, Perplexity. Influence Score (0-100) reflects marketing influenceability weighted by citation volume and domain type.

These are the top 20. Download the full study to see the full list of 50 (the PDF also includes a sector-by-sector breakdown).

What Major Trends in Citations Are We Seeing? 

Our previous analysis covered 5.7 million citations from February through June 2025. This update draws on 58.6 million citations from October 2025 through March 2026. The scale of the dataset grew significantly, but the more notable finding is how much the rankings themselves shifted.

News publishers dominated the previous top 20. Reuters, AP News, CNN, The New York Times, CNBC, Bloomberg, Fox News, and The Guardian all held positions in the top 20 six months ago. In this dataset, only Reuters and Forbes remain, and both have dropped in rank. 13 of the top 20 slots turned over between studies.

What moved in to fill them tells the real story. Google.com, Alibaba, Tom’s Guide, NerdWallet, Instagram, PCGamer, GamesRadar, Tom’s Hardware, and Vertu are all new entrants. Five of the 10 new additions are gaming or consumer electronics outlets. The top 20 has shifted from news-heavy to platform and review-heavy in roughly six months.

A few constants are worth noting. Wikipedia held #1 and actually grew its citation share from 2.2% to 3.4%. YouTube climbed from #4 to #2. Reddit, LinkedIn, X, GitHub, and Forbes all stayed in the top 20, though most moved around in rank.

Goodie’s view: as AI models have matured and query volumes have shifted toward product research and commercial intent, the sources they reach for most often have followed. Publishers built for breaking news have lost ground. Platforms and review sites built for decision-making have gained it.

Where Do Brands Have the Most Room to Influence Citations?

Many of the domains in the high-influence tier share one thing: their content is made by people. Social platforms, forums, commerce sites, and community-driven reference pages all reflect what real users, brands, and creators are saying, which is exactly the kind of signal AI models weigh heavily when building their knowledge of a topic.

That has a direct consequence for how brand authority gets established in AI answers. When an LLM has seen your brand discussed on Reddit, referenced in LinkedIn posts, and covered across YouTube, it has already formed a view of your credibility before any user asks about you. Organic social presence and community engagement used to be measured in awareness and reach. Now they are inputs into how AI represents your brand.

Moderate & Low Influence Sources

The low-influence tier is mostly academic, scientific, and medical sources. These earn citations through peer review and institutional credibility. Brands in regulated categories should not expect to crack this tier through content campaigns.

News outlets sit in the moderate-influence tier. Their content can be shaped through press coverage, contributed articles, and earned placements, though the editorial gate presents a challenge for many brands. Reuters at #18, Forbes at #8, and finance.yahoo.com at #15 all represent domains where consistent PR work can move citation share over time with the right earned strategy in place.

Add up the high and moderate categories, though, and roughly 74% of the most-cited domains in LLMs are ones where marketing activity can have a direct effect. Most brands have barely started working that window.

How Do Citation Rankings Change by Industry?

Citation behavior varies considerably by category. The domains that surface for banking queries are not the same ones that show up for skincare or software comparisons. Below are the top 10 citation leaders across three categories. The full breakdown across all 31 industries is available to download below.

Retail Banking & Fintech

NerdWallet is the clear leader in retail banking and personal finance, appearing consistently across all four AI models and across a wide range of query types. GF Magazine, MoneyWeek, and Forbes dominate the middle of the table for analysis and market commentary. Wise’s position at #8 is the most brand-specific finding: it shows that fintech companies with strong editorial coverage can earn citation authority alongside legacy financial media.

Source: Goodie AI analysis, Retail Banking vertical. Oct 2025 – Mar 2026. Top 10 by citation share.

B2B SaaS: CRM & Sales Software

TechRadar’s citation share of 8.86% in CRM and Sales Software is the highest single-category figure we observed across any subcategory in this study. It reflects how thoroughly technology review platforms have taken over the trusted evaluator role for software buyers. 

LaGrowthMachine at #2 is a notable case of a SaaS brand earning top-tier citation share through its own content. Furthermore, Monday.com and LinkedIn both appear as citation sources, confirming that SaaS brands and professional networks are treated as reference-worthy alongside traditional tech media. The X social platform is particularly popular among SaaS founders, but it doesn’t crack the top 10 rankings in our measurement. 

Source: Goodie AI analysis, B2B SaaS – CRM & Sales Software vertical. Oct 2025 – Mar 2026. Top 10 by citation share.

Skincare & Cosmetics

Wikipedia anchors the top at 4.39%, but Kaja Beauty is right behind it at 4.20%, which is one of the stronger brand-level citation shares in the full dataset. Most brands of that size do not reach the second position in their category. Alibaba at #3 reflects how heavily product discovery in beauty runs through global commerce platforms.Instagram and YouTube together make a clear case that social content is a direct input into how LLMs understand beauty trends. Vogue, Business of Fashion, and Who What Wear round out the top 10 as the editorial authorities AI consistently turns to for style and product context.

Source: Goodie AI analysis, Skincare & Cosmetics vertical. Oct 2025 – Mar 2026. Top 10 by citation share.

Which Domains Lead AI Answers in Your Industry?

The three industries above are a sample. Our full dataset covers 31 industries, from Wireless Carriers and Streaming Services to Pharmaceuticals, Oil and Gas, and Gaming Equipment. A few findings from across the full set:

• In Consumer Electronics – Wearables, vertu.com holds 15.62% citation share, the highest single-domain figure across any category in this study.
• In Gaming Equipment and Peripherals, the top 4 domains, YouTube, GamesRadar, RTINGS, and Tom’s Hardware, collectively account for over 25% of all citations.
• In Hotels and Resorts, Wikipedia alone commands 10.4% citation share, more than double the second-place domain.
• In Pharmaceuticals, essfeed.com holds 4.9% citation share, ranking second behind Wikipedia, ahead of LinkedIn, industry trade publications, and Statista.
• In Wireless Carriers, Tom’s Guide leads the category at 6.1%, ranking above Wikipedia, one of the only categories where a review outlet outranks the UGC giant.
• In eCommerce Platforms, Alibaba leads at 7.2%, with Shopify as the only brand-built platform in the top 3.

The full breakdown of all 31 industries, including top 10 cited domains, citation share, and influence scores for each, is available as a downloadable PDF. Fill out the form below to get access.

The full report covers citation data across 31 industries:

What Is Citation Share & How Is It Measured?

When AI models generate answers, they attach source links in footnotes, inline citations, sidebar references, or carousel cards. Citation share tracks how often a given domain appears in those positions across a large volume of AI responses.

Two versions of this metric tell different parts of the story. Mention share counts how often a domain gets cited across all AI responses. Impression share adjusts for context, giving more weight to citations that appear in high-visibility positions or in response to high-intent queries. A domain with strong mention share but weak impression share has volume without placement quality. Both are worth tracking because they reveal different things about where you stand and what to fix.

One important distinction from organic SEO: a site can have millions of monthly visitors and almost no LLM citations if AI systems do not treat it as a credible reference. The inverse is also true. Modest-traffic sites with well-structured, citable content can earn meaningful citation share in specific categories. Citation authority and search traffic are related but not the same signal.

Does Citation Share Build on Itself Over Time?

Brands in the top quartile for web mentions receive more than 10 times the citations in AI Overviews compared to those in the next quartile down. That gap tends to widen as more citations signal more authority to the AI model, which then produces more citations in future responses.

Adoption of AI-generated summaries is growing fast. A March 2025 study found that 58% of respondents had visited a search results page that included an AI summary, and that number has climbed as AI Mode, AI Overviews, and agentic browsing tools have become defaults across major platforms.

The budget case for earned media, data partnerships, and owned research has shifted. These are no longer primarily brand-building plays. They are direct inputs into citation share, and citation share has a compounding effect that traditional media coverage does not.

How Do You Improve Your Domain’s Citation Share?

Before jumping to tactics, the gap between citation share and influence score is worth understanding. It tells you where to focus.

Some domains have high citation share and low influence scores. Wikipedia, academic publishers, and research databases are consistently cited, but there is no real marketing path in. Those citations accrue through institutional credibility and research built over time, not through campaigns or PR.

The more actionable finding is the opposite: domains with high influence scores and low citation share represent untapped potential. Social platforms are the clearest example. Instagram scores 76 on influence but sits at 0.29% citation share overall. That gap exists because most brands are not treating their social presence as a citation strategy. They are measuring it in likes and reach while their competitors are quietly building the kind of consistent, citable brand presence that AI models learn from.

The same logic applies at the industry level. If your category’s top cited domain is a review outlet or affiliate site, that is both a threat and an opening. Getting onto that platform through coverage, partnerships, or product placements directly increases your citation share, unlike a press release in a low-influence trade pub.

Beyond that, we’ve observed that a few tactics have the clearest return:

• Get cited in high-quality content across a range of relevant domains. Breadth of coverage across source types matters, not just volume on a single site.
• Structure your content so AI systems can parse it easily. Clear headers, direct answers near the top of the page, and FAQ sections all improve citation rate.
• Publish original data. Studies, surveys, and proprietary research are among the most-cited content types in AI responses because they give other publishers and AI models something concrete to reference.
• Confirm that AI crawlers can reach your important pages. Check access settings and make sure key content is not sitting behind a login or paywall.
• Track citation share on a regular cadence. Weekly monitoring lets you catch drops early and find new opportunities before competitors do.

Platforms like Goodie pull all of this together in one place: citation monitoring across all major AI models, optimization recommendations, and the ability to act on what the data shows. Get in touch with us today to get started.