The Open Index · Sources
Every study we use, graded and linked.
19 sources from 13 publishers. Each one was read on its own page. Here is what each is, how much to trust it and what we take from it.
Headline number
19
studies, each graded and dated
Edition 1
Written byKris GjorgjioskiKeyword research and industry benchmarks, Webtec Research
Checked byHristijan NajcheskiHead of SEO and AEO (technical and AI SEO), data verification
Updated October 15, 2026 · Edition 1Refreshed quarterly · next January 2027
Built from research on
Google
OpenAI
Perplexity
Gemini
Anthropic
Cloudflare
Semrush
AT A GLANCE
What kind of evidence is in the Index.
- 9Vendor study
- 1Agency study
- 2Independent research
- 3Platform data
- 2Public dataset
- 1Platform documentation
- 1Academic paper
Grade A
Independent, official or peer-reviewed: Pew Research, HTTP Archive, Cloudflare, Vercel, Google’s documentation and an academic paper.
Grade B
Vendor or agency studies with a disclosed method. They have an interest in the topic, so we label them and explain their limits.
SOURCES
The full library
19 sources on this page: 8 independent or official (grade A) and 11 vendor or agency studies with a disclosed method (grade B). Numbers in brackets in the text match the numbers below. We read every figure on the source’s own page. How we grade sources.
- What we use:
- AI Overview prevalence and AI referral share by industry; ChatGPT’s share of AI referrals; most-cited domains by industry.
- Period:
- May to Sept 2025 (AI Overviews: Sept 15 to Oct 12, 2025)
- Sample:
- 13,770 enterprise domains in 10 industries; 17M AI responses and 100M+ citations; AI referral data from 1,215 domains (3.3B sessions, 35.7M AI sessions); 21.9M US Google searches for AI Overviews.
- Read with care:
- US data. AI Mode traffic is excluded because Google Analytics groups it with organic Google traffic. Some figures appear twice with different rounding; we use the more precise one.
- Vendor studyGrade BPublished 2025-11-13Read the source
- What we use:
- Sub-industry AI referral share and AI Overview prevalence for technology.
- Period:
- May to Sept 2025
- Sample:
- The IT industry and three sub-industries, from the same dataset as the main report.
- Vendor studyGrade BPublished 2025-12-15Read the source
- What we use:
- AI Overview prevalence over time and by category.
- Period:
- January to November 2025
- Sample:
- 10M+ keywords for prevalence over time; industry figures are the share of each industry’s keywords that show an AI Overview.
- Read with care:
- The page gives exact figures for only some categories and contradicts itself on Health.
- Vendor studyGrade BPublished 2025-11-10Read the source
- What we use:
- Which domains AI assistants cite most, and how fast that mix changed.
- Period:
- July 14 to Oct 12, 2025 (13 weekly snapshots)
- Sample:
- 230,000+ prompts and 100M+ AI citations across ChatGPT search, Google AI Mode and Perplexity.
- Read with care:
- Most shares appear only in charts; we use the figures given in the text.
- What we use:
- The first measurement of click loss for the top result when an AI Overview appears.
- Period:
- March 2024 vs March 2025
- Sample:
- 300,000 keywords: 150,000 with an AI Overview and 150,000 informational keywords without one.
- Read with care:
- Desktop CTR, informational keywords, from Google Search Console data. Correlation, not proven cause.
- What we use:
- The re-run: 58% lower top-result CTR, and the loss by position from 1 to 10.
- Period:
- December 2023 vs December 2025
- Sample:
- 300,000 keywords (150,000 with an AI Overview, 150,000 without), Google Search Console data.
- Read with care:
- Desktop only; keywords grouped by their December 2025 AI Overview status; correlation, not proven cause.
- Vendor studyGrade BPublished 2026-03-02Read the source
- What we use:
- How many AI Overview citations come from pages ranking in the top 10.
- Period:
- AI Overviews as of January 2026
- Sample:
- 863,000 keyword SERPs and 4M AI Overview URLs (Ahrefs Brand Radar).
- Read with care:
- Parsing method changed since the 2025 study, so the comparison is not like-for-like.
- Agency studyGrade BPublished 2025-11-04Read the source
- What we use:
- Organic and paid CTR with and without AI Overviews, and the effect of being cited.
- Period:
- June 2024 to September 2025
- Sample:
- 3,119 informational and educational search terms across 42 organizations; 25.1M organic and 1.1M paid impressions.
- Read with care:
- Retrospective cohort; paid sample is small; the page has small internal inconsistencies, so we use the Q3 2025 averages and endpoint values.
- Independent researchGrade APublished 2025-07-22Read the source
- What we use:
- Click and session behaviour with and without AI summaries; share of searches with a summary; which sources summaries cite.
- Period:
- March 2025
- Sample:
- 900 US adults tracked in KnowledgePanel Digital; 68,879 Google searches.
- Read with care:
- A US panel; does not measure publisher revenue.
- Independent researchGrade APublished 2025-06-25Read the source
- What we use:
- How many people use ChatGPT, by age and education.
- Period:
- Feb 24 to March 2, 2025
- Sample:
- 5,123 US adults, American Trends Panel, weighted to the adult population.
- What we use:
- The share of Google searches that end without a click, and clicks to the open web per 1,000 searches.
- Period:
- Sept 2022 to May 2024 (mobile from Jan 2024)
- Sample:
- A US and EU clickstream panel; no sample size stated.
- Read with care:
- The authors say voice and the Google app are not captured, so zero-click is likely under-counted. The post links to a newer study.
- Platform dataGrade APublished 2025-08-29Read the source
- What we use:
- Crawl-to-referral ratios by platform and the purpose of AI crawling.
- Period:
- January to July 2025 (purposes: past 12 months)
- Sample:
- Cloudflare Radar crawl and referral data; referral data comes from news-related customers.
- Read with care:
- Purpose classification relies on operator disclosures; referral data may not represent the wider web.
- What we use:
- None of the major AI crawlers it studied render JavaScript.
- Period:
- 2024
- Sample:
- Crawler request logs across Vercel’s network.
- Vendor studyGrade BPublished 2025-11-07Read the source
- What we use:
- llms.txt adoption among a sample of domains, and the lack of a measurable citation effect.
- Period:
- 2025
- Sample:
- About 300,000 domains; Spearman correlation, XGBoost and SHAP analysis.
- Read with care:
- The authors say results are context-dependent and may be outdated within weeks.
- Platform dataGrade BPublished 2025-03-17Read the source
- What we use:
- Growth of AI referral traffic by sector, and how AI-referred visits behave.
- Period:
- Nov 2024 to Feb 2025
- Sample:
- More than 1 trillion visits to US retail sites; a survey of 5,000+ US respondents.
- Read with care:
- Measures clicks on AI links only. Engagement and bounce are not defined on the page.
- What we use:
- How common titles, descriptions, canonicals, structured data, robots.txt and llms.txt are across the web.
- Period:
- 2025 crawl
- Sample:
- A crawl of millions of public websites; figures are for desktop and mobile pages.
- Read with care:
- Last updated 2026-06-09.
- What we use:
- Core Web Vitals pass rates, 2021 to 2025, by device and page type.
- Period:
- CrUX July 2025
- Sample:
- Chrome UX Report field data plus HTTP Archive lab data across public websites.
- Read with care:
- Last updated 2026-05-05.
- What we use:
- What Google says is and is not required to appear in AI features.
- Period:
- Documentation, last updated Dec 10, 2025
- Sample:
- Google’s own guidance for AI Overviews and AI Mode.
- What we use:
- The first formal study of optimizing for generative engines; reports visibility gains of up to 40%.
- Period:
- v3 revised June 28, 2024
- Sample:
- GEO-bench, a benchmark of diverse queries across domains.
- Read with care:
- The authors note efficacy varies across domains.
Preferred sources on Google
Make Webtec a preferred source on Google.
Pick us once, and Google shows you more of our research and guides in Top Stories when we publish. It takes one click and you can change it at any time.
A LIVING REFERENCE
Refreshed every quarter.
We re-check every source in January, April, July and October, add new studies and retire old ones. This is Edition 1 (October 15, 2026); the next is due January 2027. Changes are logged on the methodology page.
