Nobody in revenue publishes research with nothing to sell
Sixty publishers who write about AI, GTM, or both, assessed against four tests. Three pass all four. None of the three cover revenue.
The short answer
What does the research show about Nobody in revenue publishes research with nothing to sell?
Evidence
- The near misses fail on commercial independence, not on rigour. Read them, and read the disclosure page first.
- Where a benchmark exists to support a consulting practice, the data can still be the best available. Cite it and name the incentive in the same sentence.
- The fix for a research economy built on lead magnets is not more skepticism. It is opening the methodology appendix: find the n, find the field dates, find the success definition.
Supporting pages
- Dictionary plain-language definitions
Four tests. Synthesizes many independent sources rather than running one survey. Links every number to the publisher that produced it. Publicly critiques the methodology of research it cites, including research that agrees with it. Has no software product and no research subscription to sell.
Three of sixty pass all four, and all three sit on the AI side. The near misses fail on the fourth test, which is the one almost nobody clears. Import AI is written by an Anthropic co-founder. Interconnects discloses paid advisory relationships. Epoch AI sells commissioned research. On the revenue side the conflict is structural rather than incidental: Wynter states plainly that it publishes research for buyers, press, and AI engines to cite, and it sells the panel platform the research runs on. Benchmarkit research is partner-sponsored. The Bridge Group benchmark supports a consulting practice, and the data is good and the incentive is still there. GTMnow is the media arm of a venture fund.
The analyst firms carry the same problem in a suit. Gartner's median contract is roughly $75,300 a year and Forrester's roughly $65,755. The research is the product, which means the research has to keep producing reasons to renew.
Editorial disclosure: the editor runs marketing at an AI-native GTM company. That is a conflict on any question about whether AI works in revenue teams, and it is disclosed inside every chart footer that touches vendor claims.
What this page is
Sixty publishers who write about AI, GTM, or both, assessed against four independence tests, with the failures named.
The argument
Revenue has no research publisher with nothing to sell. The conflict in GTM research is structural rather than incidental, which means the correct response is disclosure and method reading rather than blanket skepticism.
How to read it
- The four tests are about independence and method transparency, not about quality. Conflicted research can still be the best available data.
- Failing the fourth test is the norm, not a scandal. Almost every publisher in the set sells something.
- This assessment is made by a publisher inside the same market, and the editor's disclosure sits with the assessment.
Three of sixty pass all four tests. None of them cover revenue.
Publishers assessed August 2026. Pass is a filled mark, fail is an open mark. Copper marks the three that clear every test.
| Publisher | Synthesizes many sources | Links every number | Critiques methodology | Nothing to sell | Tests passed |
|---|---|---|---|---|---|
| Zvi MowshowitzAI coverage, no product, no subscription | 4/4 | ||||
| AI Snake OilAI coverage, academic authors | 4/4 | ||||
| Gary MarcusAI coverage, no commercial research line | 4/4 | ||||
| Import AIWritten by an Anthropic co-founder | 3/4 | ||||
| InterconnectsDiscloses paid advisory relationships | 3/4 | ||||
| Epoch AISells commissioned research | 3/4 | ||||
| KellblogOnly methodology critic on the GTM side, does not link primary sources on most claims | 2/4 | ||||
| WynterStates its research exists to be cited, sells the panel platform | 1/4 | ||||
| BenchmarkitPartner-sponsored research | 1/4 | ||||
| The Bridge Group434-company benchmark supporting a consulting practice | 1/4 | ||||
| GTMnowMedia arm of a venture fund | 0/4 | ||||
| Gartner and ForresterMedian contracts of roughly $75,300 and $65,755 a year | 1/4 | ||||
| 48 other publishers assessedNone passed all four | 0/4 |
What the fourth test costs
Every publisher above the line is credible on method. The fourth test is not about quality. It is about whether the research has to keep producing reasons to renew.
What this does not say
Failing a test is not a judgment on research quality. The tests measure disclosure and incentive, not accuracy.
- Publisher
- Assessment by The Revenue AI Report
- Sample and method
- 60 named publishers assessed against four stated tests. Each row links to the publisher's own about or disclosure page
- Field dates
- August 2026
- Source
- No primary URL reachable at research time.
This is our own assessment against our own criteria, not a survey. The criteria are published above the chart so the scoring can be argued with.
One research asset already exists and nobody else in revenue publishes it.
Four candidate first-party series, placed on incremental effort against differentiation value. Concept view. There is no dataset behind this chart and the page says so.
What this does not say
This does not rate the findings those publishers produced. It records how the research is funded and framed.
- Publisher
- Assessment by The Revenue AI Report
- Sample and method
- August 2026. Do not do: sponsored benchmarks, partner funding on any research asset, or a percentage published without its n
- Field dates
- August 2026
- Source
- No primary URL reachable at research time.
Concept chart. Positions are editorial judgement, not measurement.
Also in the record
Figures that sit alongside these charts.
- The near misses fail on commercial independence, not on rigour. Read them, and read the disclosure page first.
- Where a benchmark exists to support a consulting practice, the data can still be the best available. Cite it and name the incentive in the same sentence.
- The fix for a research economy built on lead magnets is not more skepticism. It is opening the methodology appendix: find the n, find the field dates, find the success definition.
The brief
What is going on here, and why it matters.
The charts above are the evidence. This is the read: what the data describes, the mechanism behind it, where the argument could be wrong, and what a revenue team does about it.
The four tests and who clears them
The tests: synthesizes many independent sources rather than running one survey, links every number to the publisher that produced it, publicly critiques the methodology of research it cites including research that agrees with it, and has no software product or research subscription to sell. Three of sixty pass all four, and all three sit on the AI side rather than the revenue side.
The near misses fail on the fourth test only. Import AI is written by an Anthropic co-founder. Interconnects discloses paid advisory relationships. Epoch AI sells commissioned research. These are good sources with named incentives, which is the best available combination.
Why revenue is worse
On the revenue side the conflict is built into the business model. Wynter states plainly that it publishes research for buyers, press, and AI engines to cite, and it sells the panel platform the research runs on. Benchmarkit research is partner-sponsored. The Bridge Group benchmark supports a consulting practice, and the data is good and the incentive is still there. GTMnow is the media arm of a venture fund.
The analyst firms carry the same structure with more formality. Gartner's median contract runs roughly $75,300 a year and Forrester's roughly $65,755. The research is the product, which means it has to keep producing reasons to renew. That does not make it false. It makes the direction of any ambiguity predictable.
What to do instead of distrusting everything
Blanket skepticism throws away the only data that exists. The workable practice is to cite the source and name the incentive in the same sentence, then go one level deeper: find the n, find the field dates, find the success definition. Research that hides those three is the research to drop.
That practice is cheap and almost nobody runs it, which is why a 153-person conference survey became a Fortune 500 footnote. The gap in the market is not more opinions about AI in revenue. It is a methodology appendix somebody actually opened.
What to do with it
The move, by seat.
- Analyst relations
- Buy the data and read the appendix. Sample size, field dates, and success definition decide whether a number is usable.
- Content and comms
- Name the incentive in the same sentence as the citation. It costs one clause and it survives scrutiny.
- Executives
- Discount ambiguity in the direction of the publisher's business model. That is where ambiguity resolves.
Questions this page answers
What the data says, in plain language.
- What does the research show about Nobody in revenue publishes research with nothing to sell?
- Sixty publishers who write about AI, GTM, or both, assessed against four tests. Three pass all four. None of the three cover revenue. Four tests. Synthesizes many independent sources rather than running one survey. Links every number to the publisher that produced it. Publicly critiques the methodology of research it cites, including research that agrees with it. Has no software product and no research subscription to sell.
- What does the figure "Three of sixty pass all four tests. None of them cover revenue" show?
- Publishers assessed August 2026. Pass is a filled mark, fail is an open mark. Copper marks the three that clear every test. Source: Assessment by The Revenue AI Report. 60 named publishers assessed against four stated tests. Each row links to the publisher's own about or disclosure page Fielded August 2026. Confidence: Medium. Caveat: This is our own assessment against our own criteria, not a survey. The criteria are published above the chart so the scoring can be argued with.
- What does the figure "One research asset already exists and nobody else in revenue publishes it" show?
- Four candidate first-party series, placed on incremental effort against differentiation value. Concept view. There is no dataset behind this chart and the page says so. Source: Assessment by The Revenue AI Report. August 2026. Do not do: sponsored benchmarks, partner funding on any research asset, or a percentage published without its n Fielded August 2026. Confidence: Low. Caveat: Concept chart. Positions are editorial judgement, not measurement.
- What else sits alongside these figures?
- The near misses fail on commercial independence, not on rigour. Read them, and read the disclosure page first. Where a benchmark exists to support a consulting practice, the data can still be the best available. Cite it and name the incentive in the same sentence. The fix for a research economy built on lead magnets is not more skepticism. It is opening the methodology appendix: find the n, find the field dates, find the success definition.
- Where does this data come from?
- Every figure is reproduced from a named publisher: Wynter, research positioning, The Bridge Group, 434-company benchmark, Epoch AI. Sample, field date, and confidence are shown on each chart. Sources marked as vendor research are labelled on the page.
- What could not be confirmed?
- Per-publisher pass and fail reasoning for the 48 publishers summarised in the final row is held in the working file and is not charted individually.
Cite this page
Permanent URL and suggested citation.
https://www.therevenueaireport.com/research/who-publishes-the-receipts
Kvarfordt, Jonathan. "Nobody in revenue publishes research with nothing to sell." The Revenue AI Report, Research Library. https://www.therevenueaireport.com/research/who-publishes-the-receipts
Figures on this page are reproduced from the publishers listed below. Cite the original publisher for the underlying data, and this page for the compilation and framing.
Sources
Every publisher used on this page.
If a metric, model term, or method on this page is unfamiliar, every one of them is defined in The AI and Revenue Dictionary. Sample size, field date, and confidence tags are explained there too.
Wynter, research positioningVendor research
States that its research is published for buyers, press, and AI engines to cite. Sells the panel platform the research runs on
https://wynter.com/The Bridge Group, 434-company benchmark
Benchmark research supporting a consulting practice
https://www.bridgegroupinc.com/researchEpoch AI
Sells commissioned research alongside its published analysis
https://epoch.ai/
Could not confirm
What we looked for and did not find.
Claims found during research and not charted
- Per-publisher pass and fail reasoning for the 48 publishers summarised in the final row is held in the working file and is not charted individually.
How to cite this research
Written by Jonathan Kvarfordt, Founder and Principal Analyst, The Revenue AI Report. Published under CC BY 4.0.
APA
Kvarfordt, J. (2026). Nobody in revenue publishes research with nothing to sell. The Revenue AI Report. Retrieved from https://www.therevenueaireport.com/research/who-publishes-the-receipts
MLA
Kvarfordt, Jonathan. "Nobody in revenue publishes research with nothing to sell." The Revenue AI Report, 31 Aug. 2026, www.therevenueaireport.com/research/who-publishes-the-receipts.
BibTeX
@misc{kvarfordt2026whopublishesthereceipts,
author = {Kvarfordt, Jonathan},
title = {Nobody in revenue publishes research with nothing to sell},
year = {2026},
publisher = {The Revenue AI Report},
url = {https://www.therevenueaireport.com/research/who-publishes-the-receipts}
}Next theme
The most-quoted AI numbers do not survive contact with their own methodologyHow a 153-person conference survey became a Fortune 500 footnote pointing at example.com, and the twelve failure modes every number now gets run against.
Subscribe