Summary: On B2B buyer questions, the editorial publications AI search engines cite most are TechRadar (2.62% of 11,971 monitored answer runs), Forbes (2.30%) and TechTarget (1.95%). Editorial publications are a minority of what the engines cite: Reddit, LinkedIn, YouTube, arXiv and Medium are each cited more often than any of them. TechCrunch ranks 13th among editorial publications on B2B questions at 0.55%, while Business Insider, Fortune and the Wall Street Journal are cited mainly on consumer questions.
Published March 30, 2026. Updated October 13, 2026. Measured against Machine Relations Index release mri_score_v2.1+2026-10-13+6fc79cb02bbf. For the live index, see the Machine Relations Index.
A publication list for AI search has to be measured on the questions your buyers ask. The Index monitors answer runs across B2B and consumer categories, and the publications engines cite on one set are largely not the ones they cite on the other. Forbes is cited in 8.38% of consumer runs and 2.30% of B2B runs. TechTarget is cited in 1.95% of B2B runs and 0.04% of consumer runs. A single blended ranking hides both facts.
How this was measured #
Every figure below is read from the public Index file at https://machinerelations.ai/data/machine-relations-index.json and the per-domain profiles at https://machinerelations.ai/index/domains/<domain>, release mri_score_v2.1+2026-10-13+6fc79cb02bbf, observation window 2026-05-10 to 2026-10-13.
- Corpus: 24,496 monitored answer runs across six answer engines (ChatGPT, Claude, Gemini, Google AI Mode, Google AI Overviews and Perplexity). Google's two answer surfaces are counted as two engines because Google documents AI Overviews and AI Mode as separate products.
- B2B cut: 11,971 runs in the published segments of eleven B2B categories: AI infrastructure, AI security and privacy, AI visibility and GEO, cybersecurity, deep tech and hardware, enterprise software, fintech, HR and talent, industrial, logistics and freight, and martech and advertising.
- Consumer cut: 7,031 runs in seven consumer categories: consumer finance, consumer health, consumer products, family software, emergent prosumer, iGaming and betting, and local services.
- Left out of both cuts: healthcare services and education and training (2,400 runs, mixed buyers) and the legacy-unmapped segment (2,066 runs).
- A rate: a domain's citation rate in a cut is the sum of its cited runs across that cut's published segments, divided by the sum of runs observed in those segments. A rate on a fixed run set can move down as well as up, and two domains in the same cut are directly comparable. Anyone can reproduce the cut by summing
domains[].mri_score_v2.strata[].runs_citedandstrata[].runs_observedover the listed categories. - Editorial class: the release's
source_rolefield. Version 2.1 files Medium, Reddit, LinkedIn and Dev.to as community platforms and PR Newswire as wire distribution, so the editorial class now holds edited publications only: 1,730 domains in the full release, 871 of them cited on B2B questions. - Engine count and confidence are whole-corpus values from each domain's profile, so they describe the domain across all 24,496 runs and not only the cut. A count below six is a floor: the release excludes 81 engine-days on which Gemini (22), Google AI Mode (25) or Google AI Overviews (34) returned no eligible citations, and an engine in a collection gap cannot be seen citing a domain.
- A thin-sample marker: a domain with fewer than ten cited runs in a cut is shown with its count and read as low, not as a ranking.
The measured ranking: editorial publications on B2B questions #
| Rank | Publication | B2B rate | Cited runs (of 11,971) | Consumer rate (of 7,031 runs) | All-question rate (of 24,496 runs) | Engines (whole corpus) | Confidence |
|---|---|---|---|---|---|---|---|
| 1 | TechRadar | 2.62% | 314 | 3.70% | 2.80% | 5 | A |
| 2 | Forbes | 2.30% | 275 | 8.38% | 4.14% | 6 | A |
| 3 | TechTarget | 1.95% | 233 | 0.04% | 1.42% | 6 | A |
| 4 | People Managing People | 1.51% | 181 | 0.00% | 0.90% | 6 | B |
| 5 | Yahoo News | 0.94% | 113 | 4.49% | 2.15% | 5 | A |
| 6 | TechRepublic | 0.94% | 112 | 0.10% | 0.52% | 6 | B |
| 7 | selectsoftwarereviews.com | 0.83% | 99 | 0.00% | 0.42% | 6 | B |
| 8 | martech.org | 0.79% | 94 | 0.00% | 0.41% | 6 | B |
| 9 | aimultiple.com | 0.67% | 80 | 0.00% | 0.44% | 4 | B |
| 10 | IT Pro | 0.61% | 73 | 0.07% | 0.51% | 4 | B |
| 11 | Axios | 0.61% | 73 | 0.31% | 0.64% | 6 | B |
| 12 | unleash.ai | 0.58% | 69 | 0.00% | 0.28% | 6 | C |
The publications the earlier ranking led with #
The first edition of this ranking led with Medium, Forbes, PR Newswire and TechCrunch. Here is each on B2B questions in the current release.
| Surface | Class in this release | B2B rate | Cited runs (of 11,971) | Rank on B2B questions | Consumer rate (of 7,031 runs) |
|---|---|---|---|---|---|
| Medium | Community platform | 5.27% | 631 | 5th of 15,778 cited domains | 1.17% |
| PR Newswire | Wire distribution | 0.86% | 103 | 50th of 15,778 cited domains | 0.34% |
| TechCrunch | Editorial | 0.55% | 66 | 13th of 871 editorial publications | 0.07% |
| VentureBeat | Editorial | 0.33% | 40 | 30th of 871 editorial publications | 0.00% |
| CIO.com | Editorial | 0.30% | 36 | 35th of 871 editorial publications | 0.00% |
| CSO Online | Other observed source | 0.29% | 35 | 320th of 15,778 cited domains | 0.00% |
| Business Insider | Editorial | 0.16% | 19 | 78th of 871 editorial publications | 1.45% |
| Fortune | Editorial | 0.06% | 7 (thin sample) | 166th of 871 editorial publications | 1.12% |
| The Wall Street Journal | Editorial | 0.05% | 6 (thin sample) | 174th of 871 editorial publications | 1.38% |
Reuters (12 cited runs on B2B questions), The Next Web (20) and TechBullion (19) are marked Collecting across the whole corpus and carry no confidence grade.
Where B2B citations land by surface type #
Editorial publications are one surface among several in a B2B answer. The twelve most-cited domains on B2B questions:
| Rank | Domain | Class | B2B rate | Cited runs (of 11,971) |
|---|---|---|---|---|
| 1 | Community platform | 8.60% | 1,029 | |
| 2 | Community platform | 7.83% | 937 | |
| 3 | YouTube | Search or media platform | 6.31% | 755 |
| 4 | arXiv | Academic source | 5.45% | 652 |
| 5 | Medium | Community platform | 5.27% | 631 |
| 6 | G2 | Market database | 3.19% | 382 |
| 7 | Microsoft | Vendor-owned | 2.83% | 339 |
| 8 | TechRadar | Editorial | 2.62% | 314 |
| 9 | Gartner | Analyst research | 2.46% | 295 |
| 10 | Palo Alto Networks | Vendor-owned | 2.42% | 290 |
| 11 | Forbes | Editorial | 2.30% | 275 |
| 12 | Dev.to | Community platform | 2.05% | 246 |
Across the 50 most-cited domains on B2B questions, 20 are vendor-owned sources, 6 are community platforms, 6 are market databases, 6 are editorial publications, 4 are academic sources, 2 are analyst firms, 1 is a media platform, 1 is a wire and 4 are other observed sources. Across the top 20, 3 are editorial.
What the cut shows #
The leader depends on the question. Forbes leads the editorial class across all 24,496 runs at 4.14%, and TechRadar is second at 2.80%. On B2B questions the order flips: TechRadar is cited in 2.62% of runs and Forbes in 2.30%. Forbes' consumer rate of 8.38% is more than three and a half times its B2B rate, so its blended figure mostly reflects consumer questions.
Trade and software-review titles carry B2B. TechTarget (1.95% on B2B, 0.04% on consumer questions), People Managing People (1.51% on B2B, 0.00% on consumer), TechRepublic (0.94% against 0.10%) and Select Software Reviews (0.83% on B2B, 0.00% on consumer) are cited on the questions a buyer asks when evaluating software, and rarely on the others. Their whole-corpus rates look small beside Forbes, which is why a blended ranking understates them.
Business titles known for consumer reach are consumer sources here. Business Insider is cited in 1.45% of consumer runs and 0.16% of B2B runs. Fortune (1.12% against 0.06%) and the Wall Street Journal (1.38% against 0.05%) follow the same pattern, though their B2B counts are under ten cited runs and read as low rather than ranked. A media plan chosen on blended citation rank would overweight all three for a B2B buyer.
TechCrunch is a mid-table B2B source. It ranks 13th of 871 editorial publications on B2B questions at 0.55% (66 cited runs), and is observed by five of six engines across the whole corpus; Gemini is the one engine not observed citing it.
Platforms and vendors surround the editorial class. Reddit, LinkedIn, YouTube, arXiv and Medium are each cited in more B2B runs than TechRadar. Vendor-owned sources hold 20 of the top 50 slots. An answer to a B2B question draws on community, analyst, review and vendor surfaces at the same time as editorial ones.
The standing limits #
This is the citation layer. A rate here says an engine cited a domain on a monitored prompt. A cited source and a source that shaped the answer can differ, and the Index measures the first. It is not evidence that a placement in a publication caused a brand to be cited.
The prompts are the Index's. The B2B cut holds the prompts the Index monitors in those categories. A buyer question outside them can rank publications differently, and the category cut is a choice documented above, not a property of the engines.
Concentration is documented outside the Index. Analysis of AI Search Arena logs found 9% of 366,000 citations pointed to news sources and that news citations concentrate among a small number of outlets. A comparison of six LLM-based search engines and two traditional engines found 37% of domains cited by the LLM engines were unique to them. The Index shows the same shape: a narrow set of surfaces takes a large share of citations.
Class labels move. The release classifies by source type and reviewed evidence. Medium moved from the editorial class to community platforms between releases, which is why this table no longer lists it among publications. Class counts and ranks are recomputed each release.
What this means for brands #
- Pick publications on your buyers' questions, not on blended prestige. For software, security and HR buyers the B2B table points to TechRadar, Forbes, TechTarget and People Managing People. For consumer buyers it points to Forbes, Healthline, NerdWallet and Yahoo News.
- Treat a consumer-heavy business title as a consumer asset. Business Insider, Fortune and the Wall Street Journal are cited in B2B runs at a small fraction of their consumer rate.
- Count the whole surface, not only the publications. Reddit, LinkedIn, YouTube, G2 and analyst coverage appear beside editorial sources in B2B answers, so a citation plan that stops at earned media leaves most of the cited surface unaddressed.
- Read the engine column before the rate. TechRadar is observed by five of six engines across the corpus, Forbes and TechTarget by six. A source one engine never cites is a different asset from one cited across the board.
- Expect the numbers to move. The Index re-measures on each release and every figure here carries its release identifier. Do not carry a rank forward without checking the current profile.
Frequently asked questions #
Which publication is cited most by AI search engines in B2B? #
TechRadar, among editorial publications: it is cited in 2.62% of the 11,971 monitored B2B answer runs in release mri_score_v2.1+2026-10-13+6fc79cb02bbf, ahead of Forbes at 2.30% and TechTarget at 1.95%. Reddit (8.60%), LinkedIn (7.83%), YouTube (6.31%), arXiv (5.45%) and Medium (5.27%) are cited more often than any publication.
Which publication is cited most across all questions? #
Forbes, at 4.14% of all 24,496 monitored answer runs (1,013 cited runs), observed by all six engines at Confidence A. Most of that comes from consumer questions, where Forbes is cited in 8.38% of runs.
Does PR Newswire lead AI citations? #
No. PR Newswire is classed as wire distribution in this release and is cited in 0.86% of B2B runs, 50th among the 15,778 domains cited on B2B questions.
Where does TechCrunch rank? #
TechCrunch ranks 13th of 871 editorial publications on B2B questions, cited in 0.55% of runs (66 cited runs). Across all 24,496 runs it is cited in 0.69%.
Why do Business Insider, Fortune and the Wall Street Journal rank low for B2B? #
They are cited mainly on consumer questions. Business Insider is cited in 1.45% of consumer runs and 0.16% of B2B runs; Fortune and the Wall Street Journal each have fewer than ten cited B2B runs.
Why rank on a rate rather than a count? #
A cumulative count rewards whichever domain has been observed longest. A rate on a fixed run set can move in both directions, so two domains measured in the same release are directly comparable.
What is the Machine Relations view of publication strategy? #
That brands need a citation portfolio rather than a single media hit, and that the portfolio should be chosen on measured citation rate and engine coverage for the buyer's own questions rather than masthead prestige. How the layers work together is the subject of the broader Machine Relations system and its five-layer stack. For one vertical cut, see top healthtech publications cited by AI search.
Sources and method #
Index release mri_score_v2.1+2026-10-13+6fc79cb02bbf, public file https://machinerelations.ai/data/machine-relations-index.json and manifest https://machinerelations.ai/data/mri-release-manifest.json. Cuts computed from 11 B2B categories (11,971 runs) and 7 consumer categories (7,031 runs) by summing per-domain stratum counts; domain, class and engine fields are read from each domain profile. External studies: AI Search Arena news-citation analysis and six-engine comparison.
Updated 2026-10-13: figures reflect the current Machine Relations Index methodology; see the methodology and update log.