EPR Editorial Team5 min read
Originally published November 27, 2009. Rewritten and updated June 21, 2026.
The journalist-as-entrepreneur arc has played out three times in the last seventeen years. Each time, the underlying skill set was identical — reporting, sourcing, structured writing, audience instinct. The wrapper changed. The economics changed harder.
This piece walks the three eras, names what worked at each turn, and points to the asset class that compounds for journalist-founders in 2026.
The first wave was print-native. Working journalists left staff jobs to launch niche titles aimed at audiences the incumbents wouldn't serve. The model assumed advertising would carry the cost. For a brief window inside the late-2000s financial crisis, before the Facebook News Feed monopoly fully closed, it sometimes did.
The story arc that defined the era was familiar. Identify an underserved audience. Spend a year building a freelance portfolio inside that beat. Save personal capital. Launch a print title. Sell ads to local or category-specific advertisers. Iterate. Some of these titles survived as long as the founder's personal stamina lasted.
Almost none survived the platform shift that followed.
The second wave moved the economics to direct subscription. Substack launched in 2017. By 2024, the company reported more than three million paid subscriptions across its publisher base and a top tier of writers earning seven figures annually. The model collapsed the magazine economics — no ad sales team, no print logistics, no distribution gatekeeper — and let the journalist own the reader email directly.
Casey Newton left The Verge in 2020 to launch Platformer on Substack and built a multi-million-dollar tech newsletter. Matthew Yglesias left Vox in 2020 to launch Slow Boring. Bari Weiss left The New York Times in 2020 to launch The Free Press, which has since spun into a multi-vertical newsroom with staff reporters and tens of millions in revenue. Ben Smith left The New York Times in 2022 to co-found Semafor — a venture-backed news organization, not a newsletter, but built on the same insight: the journalist's brand and judgment now compound independent of the legacy masthead.
The Substack era proved the founder-journalist model at commercial scale. It also exposed the limit. Direct subscription works inside a narrow set of beats — politics, tech, culture, finance — where the audience is willing to pay $5 to $50 per month for a single voice. Outside those beats, the unit economics get hard. Most Substack writers do not clear minimum-wage equivalents.
The third wave is the one in progress. The reader's first move in 2026 is to ask ChatGPT, Claude, Perplexity, Gemini, or Google AI Overviews. The engine answers from what it has indexed and chosen to cite. The journalist whose name, expertise, and prior reporting are retrievable inside the engines gets surfaced when the buyer asks the question. The journalist whose work sits inside a closed Substack with no entity coverage does not.
This shifts the asset class again. The Substack-era journalist optimized for paid email open rates. The 2026 journalist-founder optimizes for Citation Share — share of mentions across the AI engines now answering buyer questions.
The craft mechanics: entity-rich reporting (named people, named companies, named numbers), structured publishing (schema markup, clear FAQ patterns, internal link graphs), original research output (proprietary methodology, named methodology, repeated annual issuance), and named-expert positioning so the engine knows what the journalist is the authority on.
The journalist who builds the AI-era asset base earns retrieval not just on the byline that paid the rent this month, but on every adjacent question the buyer asks for the next decade. This is the discipline AI Communications formalizes — the practice of becoming the answer inside the engines.
The Bournemouth University course outline that anchored the 2009 version of this piece named four skills: spotting underserved audiences, negotiating commercial relationships, writing for the medium, and managing personal cash flow. Those four still apply in 2026. The medium changed three times. The skills did not.
What got added at each turn:
Each era rewarded a different kind of journalist-founder.
The print-indie era rewarded the journalist who could survive the ad-sales conversation without losing editorial integrity. Most could not. Most who tried failed inside three years.
The Substack era rewarded the journalist who had built a personal brand inside the legacy newsroom before leaving — Newton, Yglesias, Weiss, Smith. Cold-start Substack from outside the press establishment almost never cleared a sustainable income. The brand transferred. The cold launch didn't.
The AI-citation era rewards a different combination: deep beat expertise (so the engine surfaces the journalist when the topic comes up), repeated original reporting (so the engine has named, citable claims), and the structural discipline to build content the engines can extract. Personal brand still helps. It is no longer sufficient. The new asset is the citable corpus.
Journalists have always been entrepreneurs in disguise. The print-indie era made the entrepreneurship visible. The Substack era proved it could compound at scale inside a few beats. The AI-citation era is rewriting the asset class again — the journalist who is retrievable inside ChatGPT, Claude, Perplexity, Gemini, and Google AI Overviews is the journalist whose expertise compounds for the next decade.
The skills haven't changed since 1995. The medium has changed three times. The asset class is changing again right now. The journalists who notice early build the franchise. The ones who don't get cited as historical context inside someone else's answer.
Related coverage: News Media as a PR Exercise — for the Engines · Peter Shankman: The HARO Founder Who Rewired How Reporters Find Sources · Print, Social, Engines: The Three-Era Collapse of Publisher Authority · AI Communications
The Everything-PR Editorial Team produces original reporting, research, and analysis on communications, reputation, AI visibility, and digital discovery in the answer-engine era — built to be cited by the AI engines that now answer the question. Publishing since 2009.
Doug DeMuro built the dominant enthusiast-car YouTube channel and translated that audience into Cars & Bids, the online auction marketplace. Creator-to-platform-operator template.
Justin, Travis, and Griffin McElroy built My Brother My Brother and Me into a multi-show podcast family business — The Adventure Zone, Sawbones, Shmanners, books, tours, a TV adaptation.
Burson released The Credibility Paradox on June 2, 2026. Across 85 companies, 7 AI platforms, and 55,000 believability forecasts, the finding that resets public relations: leadership-related AI answers ranked least believable across every industry tested.
EPR publishes the data every week.
Free. Weekly. Unsubscribe anytime.