How the structural disruption of digital publishing is reshaping brand visibility in AI — and what it means for the companies that depend on earned coverage to get recommended.
Dan Muirhead · Co-Founder, Head of Strategy
In our first Perspectives piece, The Invisible Publisher, we mapped the structural rupture in how information reaches people. The data was clear: search traffic is collapsing, AI platforms are consuming publisher content without returning meaningful value, and the publishers who built businesses on SEO-driven traffic and programmatic ads are fighting for survival.
But there was a question we didn’t fully answer — the one that matters most to the companies we work with every day: What does the disruption of digital publishing mean for the brands that depend on publisher coverage to show up in AI?
The answer turns out to be more urgent than we expected.
New research from Stacker and Scrunch — the largest study of its kind — confirms what we’ve been seeing in our own client data: earned media is the single most important input for AI visibility. Nearly two-thirds of AI citations come from third-party publisher sources. Earned placements are five times more likely to be the sole source of a brand’s AI visibility than the brand’s own website.
That finding creates a strategic paradox. The channel that matters most for AI visibility — earned media through publishers — is the channel being most acutely disrupted. And the disruption isn’t hitting all publishers equally. Different types of publishers operate on fundamentally different business models, face different levels of exposure, and are adapting in different ways. The brands that understand these differences will make smarter investments. The ones that treat “media” as a monolith will waste budget pitching outlets that may not exist in eighteen months.
We asked the AIVO Agent Team to go deep on the segmentation. What follows is a ground-level assessment of how different types of publishers are actually navigating this moment — how they make money, where they’re vulnerable, how their distribution strategies are shifting, and what all of it means for the brands trying to show up when AI answers the question.
The conversation about AI and publishers treats the industry as a single entity facing a single problem. It isn’t. A premium subscription newspaper, a programmatic-dependent lifestyle site, a B2B trade publication, and an independent creator-publisher have almost nothing in common commercially. They make money differently, reach audiences differently, and face AI disruption at entirely different scales.
Understanding the segmentation is the starting point for understanding what happens next.
NYT • WSJ • Financial Times • The Economist • Bloomberg • Condé Nast
How they make money: Diversified. Subscriptions are the center of gravity, increasingly supplemented by events, branded content, data products, and licensing. The NYT grew digital-only ARPU to $9.72 in early 2026 and raised its bundle price from $25 to $30. The New Yorker reached record revenue, profits, and subscribers in 2025. Bloomberg’s global forums delivered 30% sponsorship revenue growth. Across premium UK publishers, subscriptions overtook display advertising for the first time in Q1 2025, accounting for 34% of revenue versus display’s 31%.
AI exposure: Moderate, and manageable. Search traffic has declined — the NYT saw search’s share of traffic fall from 44% in 2022 to 37% in 2025 — but direct audience relationships provide a buffer. These publishers also command the largest AI licensing deals: the NYT–Amazon agreement is reportedly $20–25 million annually. Condé Nast, Bloomberg, and the AP have deals with multiple platforms.
AI engines trust them, cite them frequently, and will continue to do so. But placement is increasingly competitive and expensive, and these publishers are becoming more selective about branded content and partnership formats. The scarcity of stable, high-authority publishers is driving pricing power upward.
Forbes • Business Insider • BuzzFeed • HuffPost • Vox Media • Dotdash Meredith • USA Today
How they make money: Historically, scale-driven programmatic advertising, supplemented by branded content and affiliate commerce. That model is under severe pressure. Business Insider saw organic search traffic fall 55% between April 2022 and April 2025, leading to a 21% staff reduction. HuffPost lost half its search referrals over the same period. BuzzFeed warned of “substantial doubt about ability to continue as a going concern.”
But the picture isn’t uniformly bleak. Forbes is projecting roughly 9% overall revenue growth in 2026, with 12% growth in integrated media, led by video, social, live events, and branded content. Forbes produces more than 100 events per year. People Inc. built D/Cipher, a contextual targeting platform it’s now licensing to third-party publishers.
AI exposure: High, and bifurcated. The publishers in this tier that invested in original reporting, proprietary data, or distinctive editorial voices retain meaningful citation authority. The ones that built scale through SEO-optimized commodity content — aggregation, service journalism, listicles — are seeing that content absorbed and replaced by AI summaries.
Some of these publishers will consolidate, restructure, or close within the next two to three years. Others will emerge stronger with leaner operations and more premium positioning. The risk is investing in earned media relationships with publishers whose traffic, editorial quality, and AI citation authority may be declining. The opportunity is in identifying which Tier 2 publishers are successfully transitioning and building deeper partnerships with them before pricing reflects their stability.
Industry trade publications • B2B media • Specialized lifestyle • Regional and metro outlets
How they make money: This is the most varied tier. Revenue models range from events and sponsorships (dominant in B2B trade media) to programmatic display (dominant in ad-supported lifestyle), to membership and data products (emerging in specialized verticals). Local media companies have seen traffic declines of 25% to 50%.
The critical distinction within this tier is monetization model. A B2B trade publication that monetizes through events, sponsorships, and industry data can absorb search traffic loss because its revenue isn’t primarily tied to pageviews. A local news site or lifestyle publisher running on programmatic display is in the same structural trap as the most vulnerable Tier 2 publishers, but with fewer resources.
AI exposure: Extremely variable. Niche publishers with deep domain expertise have disproportionate citation value in AI. AI engines need specificity, and a trade publication that is the definitive source on, say, hotel technology or seafood supply chains is harder to replace than a generalist lifestyle site. But many mid-tier publishers never built that kind of authority.
Niche publishers with genuine domain authority are undervalued as citation sources. AI engines cite them because they’re often the only authoritative source on specific topics. But these publishers are also the most fragile commercially. A brand whose AI visibility depends on coverage from a trade publication that’s cutting staff and reducing output faces a silent risk that won’t show up in any dashboard until the citations stop.
Independent bloggers • Substack writers • Creator-led media • Single-person editorial operations
How they make money: Almost entirely through programmatic display (for blogs), sponsorships, paid subscriptions, affiliate commissions, and community access. This is where the destruction has been most acute. Travel blog The Planet D lost half its traffic after AI Overviews launched, then suffered another 90% plunge, forcing it to cease publication. A home improvement blog lost 70% of traffic resulting in a 65% decrease in ad revenue. Music blog Stereogum lost 70% of its ad revenue.
The SEO-to-programmatic pipeline that sustained hundreds of thousands of small publishers is effectively dead. The survivors are pivoting to creator-economy models: direct subscriptions, paid communities, merchandise, and live events at small scale. Substack grew 40% year-over-year and now has 100+ million monthly visitors. But the revenue distribution is extremely uneven.
AI exposure: Maximum vulnerability on the ad-supported side. But a counterintuitive dynamic is emerging: AI engines, particularly Claude and ChatGPT, have been increasing their citation of user-generated content, forum discussions, and independent voices. Reddit citations in ChatGPT increased 87% over the past year.
These publishers — travel blogs, product review sites, “best of” roundup sites — were often the sources AI engines cited for specific product and service recommendations. As they shut down, the content AI used to cite disappears or goes stale. In some verticals, the pool of third-party sources AI can reference is actively shrinking. This creates both risk (less coverage = less citation opportunity) and opportunity (the remaining sources become more valuable, and brands that invest in being covered by them gain disproportionate AI visibility).
The conversation about publisher disruption tends to focus on revenue. But there’s a parallel disruption happening in how publishers reach audiences — and it has direct implications for AI citation behavior.
For fifteen years, Google was the universal distribution mechanism for digital publishers. You published content, optimized it, and Google sent readers. That relationship is breaking. Google search referrals to publishers declined 34% globally and 38% in the U.S. between December 2024 and December 2025.
But publishers aren’t just losing search traffic — they’re losing confidence in search as a strategic channel. The Reuters Institute’s 2026 survey of 280 media executives found that newsrooms plan to deprioritize search optimization and invest more heavily in YouTube, newsletters, and what researchers call “liquid content” — modular content designed to flow across platforms rather than sit on a website waiting to be found.
This shift matters for AI visibility. AI engines, particularly Google’s AI Overviews and AI Mode, primarily pull from web-indexed content. If publishers deprioritize web publishing in favor of video, newsletters, and social, there’s less web-native content for AI retrieval systems to cite. The citation pool contracts.
Publishers are distributing content across an increasingly fragmented set of platforms: YouTube, TikTok, Instagram, Substack, LinkedIn, podcasts, and proprietary apps. Each platform has different content formats, different audience behaviors, and — critically — different relationships with AI systems.
Most-cited domain in AI Overviews (+34% in 6 months)
Leads citations across multiple AI platforms (+87% in ChatGPT YoY)
Cited by AI Overviews and Perplexity, especially for B2B topics
Content behind email delivery; not indexed by AI crawlers the same way
AI engines have limited ability to cite audio content today
The net effect: as publishers distribute across more platforms, the content that AI can actually retrieve and cite may represent a smaller share of their total output. A favorable review in a publisher’s newsletter reaches subscribers directly but may never enter the AI citation pool. A video review on YouTube might get cited in Google AI Overviews but won’t be attributed to the publisher in ChatGPT.
Getting covered by the right publisher isn’t enough. The coverage needs to exist in a format and on a surface that AI systems can retrieve. A feature article on a publisher’s website is citable. The same story told as a podcast episode or newsletter-exclusive may not be. Brands should negotiate for web publication when the publisher’s default is newsletter-only or social-only.
Every conversation about AI and publishers eventually arrives at the same question: Should publishers block AI crawlers, or should they cooperate and try to get paid? The data says neither extreme works, and the distinction between blocking and getting cited is weaker than most people assume.
BuzzStream’s March 2026 study of 4 million AI citations across 3,600 prompts found that blocking AI crawlers via robots.txt does not reliably prevent AI citation across ChatGPT, Gemini, AI Overviews, and AI Mode. AI systems cite content from blocked sites because they access it through indirect means — cached versions, third-party references, training data, and intermediary sources.
Meanwhile, roughly 30% of AI bot scrapes violate explicit robots.txt instructions. More sophisticated blocking through CDN-level enforcement is more effective but still imperfect.
“Robots.txt is as useful as a chocolate teapot.”
The structural problem that depresses the entire licensing market is Google’s refusal to separate its search crawler from its AI training infrastructure. As long as blocking AI training means blocking search indexing — total visibility or total exclusion — publishers can’t create the scarcity that would support meaningful licensing prices.
The UK’s Competition and Markets Authority has proposed conduct requirements that would mandate Google give publishers meaningful control over AI use of their content without sacrificing search visibility, but enforcement isn’t expected until late 2026 at the earliest.
per year in licensing
NYT, AP, Reuters, Condé Nast. Direct platform deals with multiple AI companies.
through platform deals or marketplaces
Larger digital-first publishers. Revenue varies widely based on content volume and citation frequency.
from direct licensing
Nothing from direct licensing. Modest-at-best revenue from collective intermediaries like News/Media Alliance.
Microsoft’s Publisher Content Marketplace, launched February 2026, is the most structurally promising model — pay-per-use with usage reporting. Early partners include AP, Vox Media, USA Today, Condé Nast, and People Inc. But payout details remain undisclosed.
For most publishers, licensing revenue is a rounding error — not a replacement for lost ad revenue. The companies adapting fastest are the ones that stopped waiting for licensing checks and invested in what they control: direct audience relationships, events, and content differentiation.
The blocking and licensing dynamics create an uneven and shifting citation landscape. Some publishers’ content is freely accessible to AI crawlers. Others are behind CDN-level blocks. Others have licensing deals that give certain AI platforms preferential access. Yext’s analysis found only 5% citation overlap across ChatGPT, Perplexity, and Gemini.
For brands pursuing AI visibility through earned media, this means a placement in a publisher with open crawler access may generate citations across all AI platforms, while a placement in a publisher with aggressive blocking may only generate citations in platforms with licensing deals. The earned media strategy needs to account for publisher-level crawler policies, not just editorial authority.
The disruption of digital publishing is not a publishing problem. It’s a brand visibility problem. Stacker’s March 2026 study — analyzing 87 stories across 30 clients, querying 2,600+ prompts across 8 AI platforms — found that distributing content through earned media channels produces a median 239% lift in AI search visibility. 64% of AI citations came from third-party publisher sources, not brand-owned content. Distributed versions of brand stories were 5.3x more likely to be the sole source of AI visibility than the brand’s own website.
This finding reframes the entire AI visibility conversation. Most brands approach AI visibility as an owned-content optimization problem — update your website, add schema markup, structure content for citation. That work matters. But the data says the larger driver is what other people say about you, particularly in editorially credible contexts.
1. The pool of citable publishers is shrinking. As small and mid-tier publishers shut down or reduce output, there are fewer third-party sources for AI engines to cite. This concentrates citation power in fewer publications, making placement in surviving outlets more valuable and more competitive.
2. Publisher stability is a leading indicator of citation risk. If a publisher in your citation ecosystem is losing traffic, cutting staff, or reducing editorial frequency, that’s a signal your future AI visibility may be at risk — even if current citations look healthy.
3. Format and distribution channel affect citation reach. A feature article on a publisher’s website is indexable and citable across AI platforms. The same story told as a newsletter exclusive, a podcast episode, or a social media thread may reach the publisher’s audience but never enter the AI citation pool.
4. Earned media strategy needs to become engine-aware. Different AI engines cite different publishers. ChatGPT shows strong preference for established knowledge sources — Wikipedia alone accounts for nearly 8% of its citations. Gemini shows the strongest preference for first-party brand content. Google AI Overviews pull from a broader set of sources, with two-thirds of citations coming from pages that wouldn’t rank on page one of the traditional SERP.
5. The brands that invest now build a compounding advantage. AI citation behavior has momentum. Brands that build a broad base of earned media coverage today create an authority signal that compounds over time. As competing publishers and competing brands lose coverage, the ones that remain become the default citation targets. This is analogous to the early years of SEO. The window is open now. It won’t be open forever.
Map your citation ecosystem. Identify which publishers AI engines currently cite when answering questions in your category. Track this across ChatGPT, Perplexity, Google AI Overviews, Claude, and Gemini. This gives you a baseline of which earned media sources actually drive AI visibility.
Assess publisher health. For the publishers that appear in your citation mapping, evaluate their commercial stability. Are they growing or contracting? Have they had layoffs? Is their editorial frequency increasing or decreasing?
Align earned media investment to citation value. Direct PR, content partnerships, and earned media budgets toward the publishers that (a) AI engines cite in your category, (b) are commercially stable, and (c) publish in formats that AI systems can retrieve.
Ensure earned coverage exists in citable formats. When you secure a placement, pay attention to where and how it’s published. Web-published articles with structured markup, clear entity references, and open crawler access have the broadest AI citation potential.
Build owned content that earned media amplifies. Your website should contain well-structured, citation-ready content on the topics you want to be known for. Earned media then amplifies and validates that content in the eyes of AI engines. One without the other underperforms.
The relationship between earned media and AI visibility is now empirically documented. Publisher disruption is real, measurable, and disproportionately affecting the mid-tier and small publishers that historically generated the most granular coverage of products, services, and experiences. The distribution fracture is moving publisher content away from web-indexed formats and toward channels that AI systems currently can’t cite as effectively. These trends are structural, not cyclical.
We believe the contraction of the publisher ecosystem will lead to increased citation concentration — fewer sources cited more frequently — which will create pricing power for surviving publishers and strategic advantages for brands that build earned media relationships early. We believe format-aware earned media strategy (prioritizing citable formats) will become a differentiator within the next twelve months. We believe publisher-level crawler policies will increasingly affect which brands get cited and which don’t, creating a fragmented visibility landscape that requires engine-by-engine strategy.
Whether AI engines will develop better capabilities for citing non-web content (newsletters, podcasts, video transcripts) that would change the distribution calculus. Whether licensing marketplaces like Microsoft PCM will scale enough to stabilize mid-tier publishers. Whether regulatory intervention (UK CMA, EU AI Act) will force Google to separate its crawlers, creating the conditions for a functioning licensing market. Whether the “Content Collapse” scenario — where AI undermines the economic incentive to create original content, degrading its own training data — will become a meaningful constraint on AI quality.
If earned media is the primary input for AI visibility, and the ecosystem that produces earned media is contracting, then AI visibility is a structural dependency on an information ecosystem that’s being actively undermined by the platforms that depend on it.
The brands that recognize this dependency — and invest in the health and diversity of their earned media ecosystem — will have more durable AI visibility than the ones that treat publisher coverage as a commodity to be acquired at the lowest possible cost.
The publishers that survive will be the ones that are genuinely irreplaceable — the ones that produce original reporting, proprietary analysis, and distinctive editorial perspectives that AI cannot generate and brands cannot replicate. Everything else will be absorbed. The brands that understand this — and build earned media strategies around the publishers that matter — will have durable AI visibility. The rest will wonder where their citations went.
Sources and methodology notes: This analysis draws on primary research from Stacker/Scrunch (March 2026, 87 stories, 2,600+ prompts, 8 AI platforms), BuzzStream/Citation Labs (March 2026, 4 million citations, 3,600 prompts), Chartbeat/Reuters Institute (2,500 news sites), Digiday+ Research publisher revenue surveys (2025–2026), AOP/Deloitte Digital Publishers’ Revenue Index, and AIVO platform data. Publisher revenue figures are sourced from company earnings calls and executive statements as cited.
AIVO maps your visibility across ChatGPT, Google AI, Perplexity, and more — then builds the strategy to close the gaps.