Methodology

Engines and exact versions

A study that does not name the model version it ran against is not reproducible, because the answer changes when the model does.

Engine	Model version
Claude	rankxa-warehouse
Gemini	rankxa-warehouse
ChatGPT	rankxa-warehouse

Run window, UTC: 2026-08-04T16:00:01Z to 2026-08-31T20:45:00Z.

Method

The design goal was to approximate what a real prospective buyer does: ask a plain-language question and see what an assistant volunteers, without logging into an API playground or crafting a prompt engineered to surface a particular brand.

The cohort. 100,966 companies were tracked across categories ranging from narrow ones like Legal Digital Marketing (10 companies) to broad ones like Mobile App Development (6,048 companies). Category size varies by two orders of magnitude in this dataset, which matters when reading any category-level cut: a percentage computed over companies in a category like Tourism Digital Marketing (3 companies) is a statement about a much smaller sample than the same percentage computed over Education IT Services (4,049 companies), and should carry proportionally less weight.

The prompts. 14,951 distinct prompts were built to mimic buyer-style questions across informational, navigational, commercial-investigation, and transactional intents, the same four-way split search researchers have long used to classify query intent. Each prompt was run against each engine inside the study window, August 4 through August 31, 2026, producing 1,013,625 total rows.

The engines. Three assistants, 3, each queried on a single model version held constant for the duration of the run (1 version for ChatGPT, 1 for Gemini, 1 for Claude). Holding the version constant matters because these products update frequently and silently; a study that spans a version change cannot tell you whether a shift in results is real or an artifact of the vendor shipping a new model mid-study.

What counts as usable. Of 1,013,625 rows produced, 1,013,625 returned a valid, parseable answer, a status rate of 1,013,625 usable answers out of 1,013,625 attempted. A row fails parsing if the engine declines to answer, times out, or returns a format the extraction pipeline cannot read as a ranked list of company mentions. The clean status rate here means the usable-answer figures in this study are not quietly built on a smaller, unrepresentative subset.

This design cannot see what the assistant would have said to a different phrasing of the same question, or to a follow-up question in a multi-turn conversation. It measures first-answer behavior to a single well-formed prompt, which is a meaningful but partial slice of how people actually use these tools.

Which models answered. Each engine was queried through its vendor's developer API, not through the consumer ChatGPT, Claude or Gemini apps, which add account history, personalisation and their own defaults. Every answer came from the vendor's small, fast model tier responding from the model's own knowledge, with web search switched off. Where the model name was recorded, it was OpenAI's GPT-4o mini, Anthropic's Claude Haiku and Google's Gemini Flash Lite. Rankxa began stamping the model name on each answer partway through August, so earlier answers in this window carry no stamp, although the configuration did not otherwise change. Read these results as what each vendor's fast tier recommends when asked cold, which is the tier most third party tools build on, not as what a logged in consumer sees.

Limitations we volunteer

Written by us, before anyone else found them.

Terms used in this study

Appearance rate

The share of answers to a given prompt type in which a company was mentioned at all, regardless of position.

Average mention position

Where in a ranked or listed answer a company tends to show up when it is mentioned; a lower number means it tends to appear earlier in the response.

Pair agreement

A measure of how often different engines, when both discuss the same prompt, name the same company; low agreement means engines are drawing on different source material or weighting it differently.

Prompt type

A category describing the intent behind a query, such as informational (seeking an explanation), navigational (seeking a specific known company), commercial-investigation (comparing options before buying), or transactional (ready to buy or sign up).

Visibility

Whether a company was mentioned by a given engine at least once across all prompts run against it in the study window.

Run window

The start and end timestamps, in UTC, over which the study's prompts were sent and answers collected.

Status ok

A row of data where the engine returned a usable, parseable answer, as opposed to an error, a refusal, or an empty response.

References

Sources this study reads against. Every link was fetched and confirmed reachable at publication.

1. Top domains cited by AI search: Analysis based on 30M sources Peec AI, 2026 Primary source for the top-ten most-cited domain ranking across AI search engines.
2. GEO: Generative Engine Optimization arXiv, 2023 Foundational paper defining generative engine optimization and proposing the method our industry context (GEO practice) responds to.
3. Position: Generative Engine Optimization Creates Underexamined Risks, Governance Must Target Concentration, Disclosure, and Academic Blind Spots arXiv, 2026 Contrasts academic and industry GEO practices, informing our discussion of what our passive measurement approach does not attempt to influence.
4. AI Visibility Statistics (2026): How Often Brands Appear in ChatGPT, Perplexity, Claude & Gemini Boring Marketing, 2026 Comparable cross-engine audit methodology (16,201 platform checks) measuring brand citation rates on the same four assistants, used here as the closest methodological peer to our study.
5. GEO: Generative Engine Optimization Princeton University, 2024 Publication record reporting the headline visibility lift GEO techniques achieved in the authors' own evaluation.
6. Web search OpenAI Developer Docs, 2026 Technical specification of the url_citation annotation object ChatGPT's API returns for web search results.
7. What Influences Brand Visibility in AI Search? A Practical Guide for 2026 Onely, 2026 Synthesizes third-party research, including a University of Toronto finding on how rarely AI answers cite brand-owned domains versus third-party content.
8. New G2 Research: Half of B2B Software Buyers Now Start Their Research With AI Chatbots PR Newswire / G2, 2026 Survey finding on how often chatbot guidance changes buyers' vendor choice, used to motivate why citation visibility matters commercially.
9. The Most-Cited Domains in AI: A 3-Month Study Semrush, 2026 Documents volatility in a single domain's citation share within ChatGPT over a few months, relevant to how we frame the stability of our own single-month window.
10. AI search engines cite Reddit, YouTube, and LinkedIn most: Study Search Engine Land, 2026 Reports Peec AI's cross-engine analysis of 30 million cited sources, used to characterize what kind of domains assistants cite most.
11. The Most Cited Domains by LLMs Similarweb AI Search, 2026 Quantifies Wikipedia and Reddit's share of ChatGPT citations in the US, used as a comparison point for engine-level citation concentration.
12. Searching the web with ChatGPT OpenAI Help Center, 2026 Vendor documentation on when ChatGPT includes citations and that search results can be incomplete or incorrect.
13. Web search tool Claude Platform Docs, Anthropic, 2026 Vendor documentation stating citations are always enabled for Claude's web search tool.
14. Half of B2B Software Buyers Now Start Their Research with AI Chatbots: G2 Demand Gen Report, 2026 Reports survey sample size and share of buyers starting research with AI chatbots, giving scale context for the G2 finding.
15. Grounding with Google Search Google AI for Developers, 2026 Vendor documentation describing how Gemini attaches inline citation annotations to grounded responses.
16. Enable and use web search Anthropic Support, 2026 Confirms Claude invokes search for time-sensitive topics and attaches citations to every such response.
17. State of AI Citations 2026: Who AI Engines Cite 5W Research, 2026 Reports the gap between brand mention and brand citation-with-link in ChatGPT, and a correlation between brand search volume and citation likelihood.

Data

The complete row-level dataset is published open and ungated under CC BY 4.0. Every number in this study can be recomputed from it.