What Publishers Need to Know About the AI Citation Economy

Sixty-nine percent of news-related Google searches now end without a click (widely referenced across industry studies). Users get what they need from the search page itself.

This is more than a traffic problem. It’s a complete visibility parameter shift. Citations are the new currency, and publishers who understand this are building advantages that transcend page view metrics.

Ken Romano, Stacker's SVP of Product and Distribution, revealed to me how Stacker’s model works in the age of GEO (Generative Engine Optimization) and why brands pay millions for it.

Stacker built a business offering publishers free content while brands pay for distribution.

In the past year, the company’s pitch changed. "If we talk to 10 COMs (Chief Marketing Officers), seven of them, their first question is going to be about GEO,” Romano told me in a Zoom call in early February. Brands want to be cited when ChatGPT, Gemini, and Perplexity answer questions in their category.

Stacker partnered with Scrunch (a vendor that optimizes content for AI visibility) to track brand citations across AI platforms, claiming 325 percent increases. That's why brands pay - and why publishers need the same capability.

Stacker stories include clear brand sponsorship disclosure at the end. The problem however is: This transparency gets lost in AI search results. When I asked whether these disclosures make it into ChatGPT answers, Romano was direct: "Into the ChatGPT answer, no."

Brands distributing across 4,000 publisher sites create exactly the pattern AI looks for. As Pete Pachal writes in the The Media Copilot, when AI sees the same narrative across sites, it gains confidence.

That model works because AI engines prioritize pattern recognition. Publishers need the same advantage - but through journalism. Branded content should only be a supplement.

A Princeton University study (cited by Search Engine Land) analyzed 10,000 search queries and found specific tactics that increase visibility by 30 to 40 percent:

Here's the strategic opportunity: Studies cited by Pete Pachal in The Media Copilot show that AI engines prioritize journalistic content far above commercial content. And they reward topical authority as much as domain authority.

When AI sees that a site or journalist has covered the same topic consistently from multiple angles, it strengthens the authority signal. Beat journalism and focused publications gain advantage because specialized coverage builds the pattern AI recognizes as expertise.

But you need "surface area" - distribute stories across formats so AI systems see them. Newsletters, speaking, video, podcasts - the advantage belongs to those who tell the most distinct story in enough places that pattern recognition systems notice.

“Reuters Institute says service journalism is the lowest priority for publishers through lens of AI,” Romano wrote in a LinkedIn post. “But this is probably just the stigma of people thinking service journalism is entirely "what's a good credit score" stories.” Jessica Roy from the San Francisco Chronicle added in the ensuing discussion thread: "Don't answer what people are already asking, because a lot of them are just asking AI. Answer what they don't know they want to know yet." Her example: California's PTO rollover rules (Paid Time Off) - people didn't know to search for them until they saw the headline.

Dom DiFurio noted on the Texas floods: "I'm not convinced AI has the granularity to pull together a list of grassroots organizations the way a dedicated reporter that knows their community could."

Local knowledge, reporting infrastructure and editorial judgment is journalism AI can't replicate and will continue citing.

The dynamics described here extend beyond Stacker. Content syndication platforms globally are repositioning around GEO value. While Stacker operates only in North America (US and Canada), European publishers face parallel offerings - platforms like StoryChief.io now market syndication explicitly on AI citation gains. The strategic questions apply regardless of vendor: Where and when should you use content syndication in the age of GEO?

Publishers using Stacker strategically get value for specific gaps. Gray Media's lean national desk (originally one person supporting more than 120 local TV stations) selectively pulls one or two stories weekly that consistently meet or exceed benchmarks - with standout pieces hitting 100K+ views and zero editorial issues such as factual errors or standards violations (case study). 

Seattle Medium, a small Black-owned newsroom, generated more then 62,000 views from 277 pickups, with Stacker stories performing nearly double their average and local content hitting two-and-a-half times normal engagement (case study). Both news orgs used Stacker to maintain production velocity without sacrificing editorial standards or community trust.

But dependency on this model becomes riskier as AI systems evolve to distinguish brand-funded from independent journalism. OpenAI already separates ads from answers. TollBit's Q4 2025 data shows 30 percent of AI bot scrapes bypass robots.txt permissions entirely, with ChatGPT's RAG agent violating explicit blocks 42 percent of the time. If you're building workflow dependency on content whose value relies on AI not making distinctions, you're betting those systems won't get smarter.

The traffic landscape is shifting dramatically. A January 2026 Reuters Institute survey of 280 news executives from 51 countries found publishers expect search traffic to decline by more than 40 percent over the next three years. Chartbeat data tracking more than 2,500 publisher sites confirms the trend is already underway - Google search referrals dropped 33 percent globally and 38 percent in the US between November 2024 and November 2025.

The composition of web traffic itself is changing. TollBit data from Q4 2025, cited by Digiday, shows the bot-to-human ratio increased 60 percent in just two quarters - from 1 AI bot visit per 50 human visits in Q2 to 1 per 31 human visits by Q4. Human traffic fell 5 percent Q3 to Q4 while AI bot scraping grew 20 percent in the same period.

Even AI licensing deals aren't solving the referral problem. TollBit found sites with AI content licensing agreements saw click-through rates drop 6.5 times in 2025 - from 8.8 percent in Q1 to 1.33 percent in Q4. Overall click through rates (CTRs) from AI tools fell nearly 3 times (from 0.8 percent to 0.27 percent) in the same period. AI tools currently send only 0.12 percent of publishers' total referral traffic - while Google still delivers 678 human visitors for every single visitor from an AI application.

Social referrals have collapsed even faster: Facebook is down 43 percent and X (Twitter) is down 46 percent over the past three years.

Forward-thinking publishers are responding by building direct monetization that doesn't depend on platform traffic:

The metrics have shifted fundamentally: Success now includes your news brand’s mentions, citation frequency in AI responses, direct traffic, email subscribers, APIs for data and consulting inquiries - not just page views.