The two quantities are not comparable, and that is the finding
The question is usually posed as arithmetic: 40,000 readers in one industry against several million readers in general. Put like that it looks like a problem with an answer, and it is not, because the two numbers are not measurements of the same thing and no published dataset converts between them.
Take the national side first, where the figures are at least independent. Pew Research's newspaper series — the last fully published edition covers 2022 — puts estimated total US daily newspaper circulation, print and digital combined, at 20.9 million for weekday and 20.9 million for Sunday, down 8% and 10% on 2021. The top 50 papers averaged 8.8 million monthly unique visitors in the fourth quarter of 2022, a 20% fall year on year. Add the digital subscriptions of The New York Times, The Wall Street Journal and The Washington Post to the audited series and weekday circulation rises to roughly 43 million, which turns an 8% decline into a 12% increase — a reminder that the direction of the national-reach story depends on which three companies you include.
The trade side has no equivalent. Industry Dive, whose Retail Dive, Utility Dive, Waste Dive, Grocery Dive, Restaurant Dive, Hotel Dive and K-12 Dive titles are the archetype of the modern trade publication, states that it runs 30-plus publications across 30 industries with 100-plus journalists and reaches “18M+ decision makers”. Those are the publisher's own numbers on the publisher's own page. The Alliance for Audited Media, which describes itself as “the world leader in quantitative media audits, helping news, magazine and business media stand out”, does audit B2B titles, and BPA Worldwide exists for exactly this purpose — but the reports are commercial products. We could not open a free audited circulation figure for any named trade title, and BPA's own site would not complete a TLS handshake for us on 26 September 2026.
So the honest statement of the comparison is this. On the national side you have an independent, declining, general-population number. On the trade side you have a publisher-asserted, qualified, unaudited number. Anyone who puts those two on one axis has invented a unit. What can be compared is what each placement does afterwards, and that is where the evidence is thinner but more interesting.
What each does for search and AI citation
Here the result cuts against the usual hierarchy, and it is the strongest single finding in this guide.
No assistant operator publishes citation shares by domain, so this rests on vendor observation. The largest published dataset is Profound's, released 5 June 2025 and updated August 2025, covering 680 million citations across ChatGPT, Google AI Overviews and Perplexity from August 2024 to June 2025. Profound sells answer-engine optimisation software and has an obvious commercial interest in the finding that AI citations concentrate on a short, influenceable list of domains; the figures are reported here with that stake named.
What the top tens contain is not what a tier-one hierarchy predicts. ChatGPT's ten most-cited domains include Forbes at 1.1%, TechRadar at 0.9%, Business Insider at 0.8%, the New York Post at 0.7% and Reuters at 0.6%. Perplexity's include Forbes at 0.7% and PCMag at 0.5%. Google AI Overviews' include LinkedIn at 1.3%, Gartner at 0.7%, Forbes at 0.6%, Business Insider at 0.5% and Medium at 0.4%. A specialist consumer-technology title out-cites a global wire service in ChatGPT. And no general-interest newspaper of record — not the Times, the Journal, the Post, the FT or the Guardian — appears in any of the three published top tens at all.
% of all citations on that platform, 680 million citations, August 2024 – June 2025
Source: Nick Lafferty, “AI Platform Citation Patterns”, Profound, 5 June 2025, updated August 2025. Profound sells answer-engine optimisation software.
Bars are scaled to the largest value shown, 1.1%; for scale, ChatGPT's most-cited domain, Wikipedia, sits at 7.8%. Each study reports only that platform's top ten, so every other publication — including all general-interest newspapers of record — is below the reporting threshold and therefore unknown rather than zero. Profound does not publish its prompt set, so the query mix behind these shares is not known and cannot be matched to a B2B buyer's questions.
Three cautions, because this is the figure most likely to be misquoted. First, absence from a top ten is not absence from the index; it is a share below a cut-off. Second, part of the reason the papers of record are missing is that many of them have withdrawn from the open corpus altogether, either by blocking crawlers or by licensing their archive, both of which we cover in do news sites block AI crawlers and which publishers have AI licensing deals. A paper that is cited rarely because it is contractually unavailable is a different phenomenon from one that is cited rarely because it is unhelpful. Third, the whole news category is small inside AI answers regardless of tier: Pew Research, tracking 68,879 unique Google searches made by 900 US adults in March 2025, found news sites accounted for 5% of the sources cited in AI summaries, identical to their 5% share of ordinary result sources. The broader picture is in which sources AI assistants cite.
There is one more piece of evidence, and it comes from a filing rather than a study. Informa TechTarget's 2025 annual report on Form 10-K, filed 11 March 2026, describes an audience-development strategy that leans on “daily newsletters from our Dive brands, direct web engagement for premium editorial brands like Dark Reading, BrightTALK webinar channels, social communities, and increasingly AI search engine referrals”, and states in its opening business description that the company is “growing audience referrals from AI search channels”. That is the operator of more than thirty trade titles telling shareholders, under securities liability, that AI search is a growing source of its readers. It is not a measurement, and the company does not quantify it. It is still the closest thing to an insider confirmation that trade titles are being surfaced by assistants.
What editors at each actually accept
The received wisdom is that trade titles are easier to get into and national titles are not. The published evidence for that is much weaker than the confidence with which it is asserted, and in one respect it points the other way.
The national end at least documents itself. Harvard Business Review publishes author guidelines setting five criteria — expertise, evidence, originality, usefulness and good writing — and is explicit that a piece must not be promotional of “your products or services”. Its originality test has been rewritten for the current moment: the ideas “should not be easily replicable by simply asking a large language model”. It says it receives “many more submissions than we can publish” and publishes no acceptance rate, no word ceiling and no response time. Multiple revision rounds are described as normal, and headline, art and placement decisions stay with the editor.
The trade end frequently documents nothing. Supply Chain Dive's editorial page — a representative modern trade title, in a category where a vendor byline would be obviously valuable — offers two routes in: a tip line for story ideas, and a paid channel to “post press releases to reach their audience”. It advertises no contributed-article programme, no op-ed guidelines and no word count. On this evidence the modern digital trade title is closer to a newsroom with an advertising department than to a byline marketplace, and the founder who expects to place a 900-word thought-leadership piece because the outlet is small has misread the model.
What neither side publishes is the number that would settle the question: submissions received against pieces run. We looked for an acceptance rate from a trade publisher and from a national one and found neither, which means the claim that trade is easier rests on practitioner experience rather than on any counted evidence. The adjacent numbers we do have are about inbound volume rather than acceptance, and we set them out in how many pitches a journalist gets; the editorial filter itself is the subject of how editors decide what to cover. The one thing that is clear from both sets of guidelines is that a promotional submission fails at the first gate in either category, which is a different proposition from paid contributor programmes — see what a contributor network is.
Who pays for each kind of journalism
This section is not a moral point. It is the mechanism that determines what a trade placement is and why its incentives differ from a national one's.
Informa TechTarget is the clearest available case because it files with the Securities and Exchange Commission and therefore has to be specific. Its portfolio includes the Industry Dive vertical titles, Computer Weekly, MicroScope, LeMagIT, Dark Reading, Light Reading, Telecoms.com, PharmaVoice and the Xtelligent healthcare sites — and, in the same company, the analyst firm Omdia. For the year to 31 December 2025 it reported total revenue of $486.8 million, made up of $355.8 million of marketing, advertising services and sponsorship, $77.0 million of intelligence subscription services, $52.4 million of advisory services and $1.6 million of exhibitor and attendee revenue. It also recorded a net loss of $1.0 billion for the year, driven by impairment.
Informa TechTarget revenue by category, US$ millions, year to 31 December 2025
Source: TechTarget, Inc. (Informa TechTarget) annual report on Form 10-K for the year ended 31 December 2025, filed with the US Securities and Exchange Commission on 11 March 2026.
Bars are scaled to the largest value. Total revenue $486.8 million; the largest category is 73% of the total. The company reports as a single operating segment, so these categories are a revenue disaggregation rather than divisional results, and no reader-paid subscription line is broken out. Net loss for the year was $1.0 billion, principally impairment.
Read the largest line honestly and the trade placement comes into focus. Roughly three dollars in four at a major B2B publisher come from vendors buying access to the audience the journalism assembles, and the 10-K says so in as many words: the Industry Dive brands provide “brands with a route to reach these audiences through outbound email sponsorship and content marketing in particular”. The company sizes the market it competes in at about $20 billion a year with roughly 45,000 potential customers — that is, it has 45,000 possible vendor clients and tens of millions of readers, and only the first group pays.
Two consequences follow for anyone weighing a trade feature. The first is that the commercial and editorial audiences are the same people, which is precisely what makes trade reach valuable and precisely what makes the line between earned and paid harder to see from outside than at a national title; the labelling obligations that govern it are in how paid content must be labelled. The second is that the same corporate parent can sell you editorial adjacency, purchase-intent data about the readers, and analyst coverage through Omdia — the filing says Omdia is positioned to serve clients “on both the buy-side and sell-side of the technology market”. That combination is the subject of our companion guide, do analyst firms affect what AI recommends.
Referral quality, and how buyers say they discover vendors
These are the two questions a B2B founder most wants answered and the two where we have the least to offer. Saying so is more useful than filling the gap.
On referral quality, we could find no published dataset comparing bounce rate, session depth, pages per visit, form completion or pipeline contribution between trade-publication referrals and national-outlet referrals. Analytics platforms hold this data by the terabyte and none of them publishes it cut that way. The argument that trade referrals convert better is a good argument — the reader is by construction in the industry, often reached through a subscribed newsletter rather than a search result — but it is an argument, not a measurement, and it should not be quoted as a statistic. What is measured is how little referral traffic a placement of any kind produces and how fast it decays, which we set out in what happens to placement traffic and, for the search side, in do AI Overviews reduce clicks. Pew's figures are the sharpest available: with an AI summary present, 8% of searches produced a click on a traditional result link and 1% a click inside the summary, against 15% when no summary appeared.
On how B2B buyers say they discover vendors, the widely circulated figures — the share of the buying cycle spent with any supplier, the size of the buying group — originate with Gartner, and we could not open Gartner's pages to verify them. Gartner's website returned an automated-access block to every attempt we made on 26 September 2026, including its robots file. We are not going to quote a statistic from a source we could not read, so those numbers do not appear here, and any guide that reproduces them without saying where it got them should be treated with suspicion. The surveys that are publicly readable in this area are published by companies selling demand-generation software to the marketers being surveyed, and we could not open a methodology page for any of them either. Our existing treatment of what the buying-side evidence does and does not support is in does media coverage affect B2B buying and, for the assistant-mediated version of the same question, AI search and B2B discovery.
What survives is a structural observation rather than a survey. A national outlet sells attention to a general population; a trade outlet sells attention to a population that a vendor could otherwise only reach by buying a list. That is why 73% of a major trade publisher's revenue is vendor money and why national publishers increasingly depend on reader subscriptions instead. The audience composition, not the audience size, is the asset. Whether that converts is unmeasured in public.
Methodology and limitations
Citation shares are from “AI Platform Citation Patterns: How ChatGPT, Google AI Overviews, and Perplexity Source Information” by Nick Lafferty, Profound, published 5 June 2025 and updated August 2025, covering 680 million citations observed between August 2024 and June 2025. Profound sells answer-engine optimisation software. The study reports only each platform's top ten domains, so every publication outside those lists — including all general-interest newspapers of record — is below the reporting threshold and its share is unknown rather than zero; the prompt set is not published, so the query mix cannot be matched to a B2B buyer's questions; and the products change continuously, so the shares are a period observation. Search-behaviour figures are from Pew Research Center, 22 July 2025, based on browsing data from 900 US adults who installed tracking software, covering 68,879 unique Google searches made in March 2025 with data collected 7 to 17 April 2025; it is a US-only opt-in panel and measures consumer search rather than B2B research. Newspaper circulation figures are from the Pew Research Center newspapers fact sheet, whose last fully published series covers 2022 and is therefore four years old at the time of writing; the augmented 43 million weekday figure depends on adding three named publishers' digital subscriptions to the audited series and is not comparable with the unaugmented number. Publisher economics are from TechTarget, Inc. (Informa TechTarget) annual report on Form 10-K for the year ended 31 December 2025, filed 11 March 2026, which is a primary regulatory filing but describes one company and should not be generalised to all trade publishers; the company reports as a single operating segment, so the revenue categories quoted are a disaggregation rather than divisional results. Industry Dive's audience figures — 30-plus publications, 30 industries, 18 million-plus decision makers, 100-plus journalists — are the publisher's own claims on its own website, unaudited and uncorroborated, and are labelled as such throughout. Editorial guidelines are quoted from Harvard Business Review's published author guidelines and Supply Chain Dive's editorial page, both read 26 September 2026; a single title is not a category, and the absence of a contributed-content programme on one trade title's page does not establish that trade titles generally refuse bylines. Four gaps are material and we have left them open rather than bridged. We could not open a free audited circulation figure for any named trade title: the Alliance for Audited Media's circulation reports are commercial products and BPA Worldwide's site would not complete a TLS handshake for us. We found no acceptance-rate figure for any publication in either category. We found no dataset comparing referral quality between trade and national placements. And we deliberately excluded the most-quoted B2B buying statistics in the field because they originate with Gartner, whose site returned an automated-access block to every request we made, so we could not read them at source.
Conclusion
For a B2B company the trade feature is the better instrument and the national mention is the better ornament, and the evidence supports that split more weakly than anyone selling either would like. What is genuinely established is that specialist titles are cited by AI assistants at rates that embarrass the traditional hierarchy — TechRadar above Reuters in ChatGPT, no paper of record in any published top ten — and that the trade publisher's business model is the sale of its readers' attention to the vendors it covers, which is both why its audience is worth reaching and why its editorial boundary is harder to see. What is not established is anything about conversion: no audited trade circulation figure that a buyer can check, no acceptance rate on either side, no referral-quality comparison, and no readable primary survey of how B2B buyers actually find vendors. The defensible position is therefore modest. Pursue the trade feature because the readers are the buyers and the page is durable and category-specific, not because a study says it converts. Pursue the national mention for the recognition it carries with investors, recruits and procurement committees, which is what what makes an outlet tier one and are tier-one media features still worth it examine in detail. And treat any number that ranks the two against each other as an estimate someone has dressed up, because as of today nobody has published the measurement.
Frequently Asked Questions
Is a trade feature worth more than a national mention for a B2B company?
On the evidence that exists, usually yes for the buying decision and usually no for everything else. The trade title reaches a pre-qualified audience and produces a durable, indexable, category-specific page; the national mention reaches an audience of whom almost none are buyers but carries recognition value with investors, recruits and procurement committees. Nobody has published a controlled comparison of the two, so any ranking stated with confidence is a preference dressed as a finding.
Which kind of outlet actually gets cited by AI assistants?
Specialist titles do better than the framing suggests. On Profound's dataset of 680 million citations from August 2024 to June 2025, ChatGPT's ten most-cited domains include TechRadar at 0.9% — ahead of Reuters at 0.6% — and Perplexity's include PCMag at 0.5%. No general-interest newspaper of record appears in any of the three platforms' published top tens.
How large is a trade publication's audience really?
Almost always unverifiable from outside. Industry Dive states it operates 30-plus publications reaching 18 million-plus decision makers with 100-plus journalists, but those are the publisher's own figures. The Alliance for Audited Media and BPA Worldwide audit B2B media, and their circulation reports are paywalled; we could not open a free audited figure for any named trade title.
Do trade editors accept contributed articles more readily than national editors?
This is widely assumed and poorly evidenced. Harvard Business Review publishes explicit author guidelines requiring original, evidence-backed ideas that are not promotional of “your products or services”, and states it receives “many more submissions than we can publish” without giving a rate. Supply Chain Dive's editorial page offers a tip line and a press-release channel and advertises no contributed-article programme at all. Neither category publishes an acceptance rate.
Who pays for trade journalism?
The vendors being covered, for the most part. Informa TechTarget, which owns the Industry Dive titles, Computer Weekly, Dark Reading and the analyst firm Omdia, reported $486.8 million of 2025 revenue of which $355.8 million — 73% — came from marketing, advertising services and sponsorship. That does not make the journalism bad, but it means the commercial and editorial audiences are the same people.
Is trade referral traffic better quality than national referral traffic?
Probably, and nobody has published the numbers. We could find no dataset comparing bounce rate, session depth, form completion or pipeline contribution between trade and national referrals. The claim is plausible from first principles and unmeasured in public, and it should be described that way rather than quoted as a fact.
Sources
- regulatory filing TechTarget, Inc. (Informa TechTarget), 2026. Annual Report on Form 10-K for the fiscal year ended December 31, 2025. US Securities and Exchange Commission, filed 11 March 2026. Source for the $486.8 million total revenue and its categories, the brand portfolio including Omdia, the “AI search engine referrals” language and the $20 billion / 45,000-customer market sizing.
- vendor Lafferty, N., 2025. AI Platform Citation Patterns: How ChatGPT, Google AI Overviews, and Perplexity Source Information. Profound, 5 June 2025, updated August 2025. 680 million citations, August 2024 to June 2025; top ten domains per platform only. Profound sells answer-engine optimisation software.
- research Chapekis, A. and Lieb, A., 2025. Google users are less likely to click on links when an AI summary appears in the results. Pew Research Center, 22 July 2025. 900 US adults, 68,879 unique searches in March 2025, data collected 7–17 April 2025. Source for the 8%/1%/15% click figures and the 5% news-site share of AI summary sources.
- research Pew Research Center. Newspapers Fact Sheet. State of the News Media. Source for 20.9 million weekday circulation in 2022, the 8.8 million monthly unique visitors across the top 50 papers in Q4 2022 and the augmented 43 million figure. Latest fully published series covers 2022.
- operator Industry Dive. About Industry Dive. Read 26 September 2026. Publisher's own claims: 30+ publications across 30 industries, 18M+ decision makers, 100+ journalists. Unaudited.
- editorial policy Harvard Business Review. Guidelines for Authors. Read 26 September 2026. Five publishing criteria, the prohibition on promoting “your products or services”, the large-language-model originality test and the “many more submissions than we can publish” statement. No acceptance rate published.
- editorial policy Supply Chain Dive. Editorial team. Read 26 September 2026. Offers a tip line and a press-release channel; advertises no contributed-article programme, guidelines or word count.
- industry body Alliance for Audited Media. Alliance for Audited Media. Read 26 September 2026. Confirms that B2B media are audited; circulation reports are commercial products and no free audited figure for a named trade title was reachable.
- research Reuters Institute for the Study of Journalism, 2025. Digital News Report 2025 — Executive Summary. 48 markets. Used for context on the national-news environment: trust in news at 40%, news avoidance at 40%, and social media (54%) overtaking news websites (48%) as a US news source.