How have click-through rates changed since AI Overviews appeared? Which websites does ChatGPT cite most often? How do query fanout, RAG, and other AI search complexities actually work?
Research is how you answer these questions, and how you keep up with what’s changing in AI search. These are the 25 people publishing the most original research on AI search, and the ones worth following.
How I compiled this list
I started with a longlist of 87 names: people widely recommended as AEO and GEO experts, and researchers at AI visibility companies.
Then I counted how many original AI search studies each person has published. I pulled every page published on each person’s main publishing sites since 2023 (more than 17,000 pages), checked each one against the criteria below, and checked who was credited. Where the same study appeared in more than one place, such as a cross-post, translation or update, I counted it once.
A page counted as a study if it met four tests:
- Original data. The author or their team collected or analyzed it themselves, such as prompts run through AI assistants, AI Overview results, citation datasets, analytics, server logs, surveys or controlled experiments.
- About AI search. How ChatGPT, Perplexity, Gemini, Claude, Copilot, Grok, AI Overviews or AI Mode find, cite and recommend content, the traffic they send, or how people use them to search.
- A meaningful sample. Roughly 50+ prompts, queries or pages; analytics from 10+ websites; 100+ survey respondents; or a controlled experiment with a described setup.
- A published write-up. An article, report or paper published between January 2023 and November 2026.
People only got credit for studies carrying their byline, or that name them as a co-author or as the analyst behind the data. Company research with no named author didn’t count toward anyone.
I ranked by number of studies, and broke ties by the number of websites linking to those studies.
The top 25 AEO/GEO experts
|
Growth Memo |
Data-driven LLM citation and AI search behavior |
|||
|
AI search citations and click impact |
||||
|
AI grounding and citation mechanics |
||||
|
Data-driven AI search citation research |
||||
|
Large-scale AI citation and visibility data |
||||
|
Reverse-engineering AI search citation and retrieval behavior |
||||
|
Search Atlas |
Large-scale LLM citation behavior research |
|||
|
AI citation behavior across answer engines |
||||
|
Large-scale AI Overviews and AI citation studies |
||||
|
Data-driven AI search citation and traffic research |
||||
|
seoClarity |
AI citations vs organic rankings |
|||
|
Snippet Digital |
How AI search engines pick sources |
|||
|
Large-scale AI citation and search behavior research |
||||
|
Peec AI / ZipTie |
AI citations, fan-out queries and tracking |
|||
|
Fan-out queries and AI crawling research |
||||
|
Moz / STAT |
AI search results versus organic rankings |
|||
|
AI search citations and source reliability |
||||
|
GPT Insights |
German AI search and brand visibility research |
|||
|
Measuring AI search against traditional search |
||||
|
Previsible |
LLM referral traffic measurement |
|||
|
Data-driven AEO measurement and AI search research |
||||
|
seoClarity |
AI Overviews’ impact on organic search |
|||
|
Orbit Media |
AI search traffic, conversion, and visibility data |
|||
|
AI brand visibility measurement and research |
||||
|
Siege Media |
Data-driven generative engine optimization research |
1. Kevin Indig
Growth Memo | 36 studies
Kevin Indig is a growth advisor and author of the Growth Memo newsletter, and he previously led SEO and growth at Shopify, G2, and Atlassian. His research uses large citation and user behavior datasets to show what drives LLM citations, including retrieval rank, content depth, brand popularity, and G2 profile signals, and how AI Overviews and AI Mode change the way people search and choose.
Where their studies are published: growth-memo.com (25), searchenginejournal.com (6), sell.g2.com (3), kevin-indig.com (1), learn.g2.com (1)
Most-linked studies:
2. Xibeijia Guan
Ahrefs | 31 studies
Xibeijia Guan is a data scientist at Ahrefs and the analyst behind most of its AI search research. Her studies cover how much AI Overviews cut clicks (34.5%), how often AI-cited pages rank in Google’s top 10, and how stable AI citations are over time.
Where their studies are published: ahrefs.com (31)
Most-linked studies:
3. Dan Petrovic
DEJAN | 20 studies
Dan Petrovic is the Founder and Director of DEJAN, an SEO agency, and of its AI-focused arm DEJAN AI. His research uses large query and citation datasets to measure how Google and OpenAI ground answers, including how many words Google takes from each source, how little page content survives filtering, and how formatting and intent match affect AI visibility.
Where their studies are published: dejan.ai (20)
Most-linked studies:
4. Louise Linehan
Ahrefs | 19 studies
Louise Linehan is a content marketer at Ahrefs who leads much of the company’s AI search research writing. Her studies cover how AI Overviews and ChatGPT choose citations, including a 75,000-brand visibility analysis, a finding that 97% of llms.txt files never get read (137,210 domains), and a test showing 1,885 pages that added schema saw barely any change in AI citations.
Where their studies are published: ahrefs.com (19)
Most-linked studies:
5. Johannes Beus
SISTRIX | 14 studies
Johannes Beus is the founder and managing director of SISTRIX, the SEO and visibility analytics software company. SISTRIX research under his leadership is known for large-sample studies of how AI Overviews, AI Mode, and ChatGPT choose and change their cited sources, and what that does to click-through rates and brand visibility.
Where their studies are published: sistrix.com (7), sistrix.de (6), sistrix.fr (1)
Most-linked studies:
6. Metehan Yeşilyurt
Peec AI | 14 studies
Metehan Yeşilyurt is a GEO Researcher at Peec AI who publishes experiments and large data analyses on how ChatGPT, Perplexity, and Google AI Overviews retrieve and cite sources. His work is known for specific findings such as GPTBot crawling a new 60,000-page site 29,000+ times in 12 hours, Perplexity’s reranking patterns, and the drop in Reddit citations in ChatGPT shopping answers across millions of conversations.
Where their studies are published: metehan.ai (9), peec.ai (2), metehanai.substack.com (2), speakerdeck.com (1)
Most-linked studies:
7. Manick Bhan
Search Atlas | 13 studies
Manick Bhan is the founder and CTO of Search Atlas and the founder of LinkGraph. His research, built on large citation datasets such as 120M+ AI citations across 6.4M domains, shows that LLM citations are only loosely tied to Google rankings and domain authority, and that schema markup and /llms.txt show no measurable effect on visibility.
Where their studies are published: searchatlas.com (12), signalgenesys.com (1)
Most-linked studies:
8. Josh Blyskal
Profound | 11 studies
Josh Blyskal is the AI Strategy & Research Lead at Profound, where he runs large-scale studies of how AI answer engines choose which sources to cite. His research draws on datasets of up to billions of citations and shows that cited domains change heavily over time, differ sharply between ChatGPT, Perplexity and Claude, and are only weakly explained by traditional SEO metrics.
Where their studies are published: tryprofound.com (8), joshblyskal.com (2), speakerdeck.com (1)
Most-linked studies:
9. Ryan Law
Ahrefs | 9 studies
Ryan Law is Director of Content Marketing at Ahrefs. His studies analyze AI search at large scale, including one finding that AI Overviews reduce clicks by 34.5% across 300,000 keywords and another examining 146 million SERPs to see what triggers AI Overviews.
Where their studies are published: ahrefs.com (9)
Most-linked studies:
10. Aleyda Solís
Orainti | 9 studies
Aleyda Solís is an international SEO consultant and founder of Orainti, and she also created the SEOFOMO newsletter and LearningAIsearch. Her research tests how AI search works and measures where citations and traffic come from, including findings that third-party sources drive 84% to 93% of AI citation weight for SaaS brands and that AI search is only 0.19% of visits across 40 sites.
Where their studies are published: aleydasolis.com (8), hub.seofomo.co (1)
Most-linked studies:
11. Mitul Gandhi
seoClarity | 9 studies
Mitul Gandhi is co-founder and Chief Architect of seoClarity, the enterprise SEO platform. His studies use seoClarity’s search data to measure how AI Overviews, AI Mode and ChatGPT citations line up with Google rankings, including a 1,000-query AI Mode study (12,011 citations) and a report on 77 million AI search visits across 300+ brands.
Where their studies are published: seoclarity.net (9)
Most-linked studies:
12. Suganthan Mohanadasan
Snippet Digital | 7 studies
Suganthan Mohanadasan is a co-founder of Snippet Digital and Keyword Insights and an AI SEO researcher. His research inspects network traffic, API responses, and server logs to show how ChatGPT, Perplexity, and DeepSeek find, label, and choose the sources they cite.
Where their studies are published: suganthan.com (7)
Most-linked studies:
13. Davis McCain
Profound | 7 studies
Davis McCain is a Research lead at Profound and one of its early research hires. His studies are known for large-scale AI search data, including a 1.19 million page-engine lifecycle analysis across 7 AI engines that found the half-life of an AI citation is 11 days, and a 1.4M citation analysis showing LinkedIn is the most-cited domain for professional queries.
Where their studies are published: tryprofound.com (7)
Most-linked studies:
14. Tomek Rudzki
Peec AI / ZipTie | 6 studies
Tomek Rudzki is a GEO researcher at Peec AI and co-founder of ZipTie.dev, and previously led R&D at Onely. His recent studies map which domains AI search cites (30M sources), how ChatGPT builds fan-out queries, and why AI Overviews are the most under-tracked AI surface.
Where their studies are published: peec.ai (5), ziptie.dev (1)
Most-linked studies:
15. Chris Long
Nectiv | 6 studies
Chris Long is a co-founder at Nectiv and an SEO and AEO consultant who focuses on B2B and technology brands. His research is known for analyzing how ChatGPT and Google build fan-out queries (such as ChatGPT’s jump to 7.61 per prompt and Google’s average of 9.06), as well as OpenAI’s crawl activity across 7B+ log files and a CHEQ case study showing a 5,600% rise in AI citations.
Where their studies are published: nectivdigital.com (5), botify.com (1)
Most-linked studies:
16. Tom Capper
Moz / STAT | 6 studies
Tom Capper is a Senior Search Scientist at Moz and STAT, where he has also worked on search product strategy. His research uses large keyword samples, often 40,000 or more, to measure how AI Overviews, AI Mode and SGE citations compare with traditional top 10 organic rankings, including where AI Overviews appear on the page.
Where their studies are published: moz.com (3), getstat.com (2), webtan.impress.co.jp (1)
Most-linked studies:
17. Lily Ray
Amsive | 5 studies
Lily Ray is VP of SEO & AI Search at Amsive, where she studies how AI search tools choose and cite sources. Her recent studies cover self-promotional listicles in Google AI Overviews (69% of 100 B2B queries cited them while recommending competitors), the link between Google organic visibility losses and lost AI citations across 11 subfolders, and how quickly different LLMs repeated fictional claims.
Where their studies are published: lilyraynyc.substack.com (3), lilyray.nyc (1), amsive.com (1)
Most-linked studies:
18. Hanns Kronenberg
GPT Insights | 5 studies
Hanns Kronenberg is an AI search advisor and the founder of GPT Insights in Germany. His research covers German-language AI search, including a 1,000,000 prompt dataset (DEVAL-25) with about 14,000 Netflix-related prompts, an analysis of 3,000 keywords with AI Overviews using Sistrix data, and a Brand-AIO study of 100 brands across 800 SERP observations.
Where their studies are published: gpt-insights.de (4), groundingpage.com (1)
Most-linked studies:
19. Rand Fishkin
SparkToro | 4 studies
Rand Fishkin is the cofounder and CEO of SparkToro and the author of Zero Click Marketing. His research uses large device panels and volunteer prompt runs to measure how people search and how consistently AI tools recommend brands, including a finding that Google handled about 373 times more searches than ChatGPT and a 2026 study showing AI brand lists rarely repeat.
Where their studies are published: sparktoro.com (4)
Most-linked studies:
20. David Bell
Previsible | 4 studies
David Bell is Chief Product Officer and co-founder of Previsible, a search consultancy. His annual State of AI Discovery reports use GA4 data from dozens of sites (1.96M LLM sessions in the 2025 edition) to measure how much traffic AI assistants send and which platforms send it.
Where their studies are published: previsible.com (2), searchengineland.com (2)
Most-linked studies:
21. Ethan Smith
Graphite | 4 studies
Ethan Smith is the Founder and CEO of Graphite, an SEO and growth-marketing agency focused on SEO and answer engine optimization. His research uses large datasets to test claims about AI and search, including that AI-generated content is a small share of Google results and AI citations, that SEO traffic fell only 2.5% year over year, and that sampling about 10 AI responses is enough to estimate brand visibility and position.
Where their studies are published: graphite.io (4)
Most-linked studies:
22. Mark Traphagen
seoClarity | 4 studies
Mark Traphagen is VP of Product Marketing and Training at seoClarity. His research uses seoClarity’s keyword datasets to measure how AI Overviews affect SEO, including a 432,000-keyword impact study and an analysis of 120 million US keywords.
Where their studies are published: seoclarity.net (4)
Most-linked studies:
23. Andy Crestodina
Orbit Media | 4 studies
Andy Crestodina is the CMO and co-founder of Orbit Media Studios, a web design and digital marketing agency. His research draws on surveys and site data, including a 1,110-person AI search adoption survey, conversion rates across 97 B2B websites and 29 million visits, and 560,000+ AI crawler requests, and it shows that AI visitors convert to leads at 3x the rate of other organic traffic, that AI crawlers visit homepages 15 times more than other page types, and that free AI visibility audits give arbitrary scores.
Where their studies are published: orbitmedia.com (4)
Most-linked studies:
24. Jan Ehrlinspiel
Peec AI | 4 studies
Jan Ehrlinspiel is a GEO Researcher at Peec AI who publishes data studies on how AI engines mention and rank brands. His research covers how listicle position affects brand visibility across nearly 200,000 AI responses, why tracking more prompts once gives more accurate visibility scores than repeating fewer prompts, and how a company’s name shapes the way AI models describe it.
Where their studies are published: peec.ai (3), searchenginejournal.com (1)
Most-linked studies:
25. Ross Hudgens
Siege Media | 4 studies
Ross Hudgens is the founder and CEO of Siege Media and the author of Generative Engine Optimization. His research is known for client and multi-site studies on how AI Overviews and LLMs affect traffic, including a 50-site study showing homepage clicks up 10.7%, a 116-site study naming versus pages as the strongest predictor of AI search traffic, and a 585-brand study where five or more homepage credibility statements raised AI visibility 62% for mid-authority brands.
Where their studies are published: siegemedia.com (4)
Most-linked studies:
Bonus: The 10 most popular AEO/GEO studies
These are the 10 studies by the experts on this list that have earned links from the most websites.