ChatGPT Citation Share - March 2026
Klaas FoppenThe Most Cited Domains on
ChatGPT in March 2026
Based on Promptwatch data from Mar 1 - Mar 31, 2026. Citations tracked from real prompt responses using ChatGPT Search when triggered in the UI.
What this means for you
March is a two-horse race at the top: Wikipedia takes 3.61% of all ChatGPT Search citations with Reddit right behind at 3.21%, a gap of just 0.4 points. Encyclopedic reference and community discussion are pulling nearly equal weight in ChatGPT's answers.
The rest of the field sits below half a percent: TechRadar leads the chasing pack at 0.46%, followed by arXiv (0.34%), Forbes (0.33%), and Walmart (0.31%). Even summed, these 20 well-known domains capture only around a tenth of all citations; the long tail of specialized sites takes the rest.
How to act on it
- Cover both flanks: maintain accurate, well-cited Wikipedia content where your brand has notability, and earn authentic mentions in the subreddits your buyers frequent. In March, ChatGPT drew on both almost equally.
- Take the niche categories seriously. arXiv (0.34%) and Walmart (0.31%) cracking the top six shows that original research and structured product data get cited. Publish primary data and keep product pages machine-readable.
- Aim your own content at specific, answerable questions. With ~90% of citations going past these 20 giants to smaller sites, precise factual pages on your domain can win the prompts that matter to your business.
ChatGPT Citation Share
February vs March 2026
Side-by-side comparison of citation share for the top domains on ChatGPT Search between February and March 2026.
What this means for you
The gap at the top is closing: Reddit climbed from 2.96% in February to 3.21% in March while Wikipedia slipped from 3.8% to 3.61%, which cut Wikipedia's lead from 0.84 points to just 0.40. Further down, the movers are striking: arXiv more than doubled (0.14% to 0.34%), Walmart tripled (0.10% to 0.31%), and Business Insider doubled (0.08% to 0.16%), while LinkedIn (0.34% to 0.25%) and Trustpilot (0.23% to 0.16%) gave up ground. ChatGPT's sourcing mix is anything but static.
How to act on it
- Get ahead of the Reddit trend: it added a quarter point of share in one month, so start building genuine subreddit presence now rather than after the platform becomes even more dominant.
- Copy what the risers have in common: primary sources. Publish original research or data (the arXiv playbook) and keep product and pricing information structured and current (the Walmart playbook).
- Don't build your AI visibility on a single third-party platform. LinkedIn and Trustpilot both lost meaningful share this month; a spread across community, review, and owned channels survives these swings.
How we collect this data
We collect millions of prompt responses, citations, and click data from the actual user interfaces of major AI platforms: over 26 billion data points and growing. This gives us one of the largest datasets on how AI search engines cite sources and recommend brands.
Real UI monitoring
Data straight from the interfaces of ChatGPT, Gemini, Perplexity, Claude, AI Overviews, and more.
26B+ data points
Over 26 billion analyzed citations, prompts, and responses, one of the largest AI search datasets available.
Continuously updated
Refreshed constantly so the trends you see reflect the latest behavior of AI search engines.
Aggregated & public
Published freely for the GEO community, based on aggregated, non-identifiable trends.
Want to start tracking your own AI search data? Get started with Promptwatch
Track Your Domain's AI Citation Share
See which domains AI search engines cite most for your industry. Monitor your brand's citation share compared to competitors and get alerts when citation patterns shift.
