She Spent Years Correcting Strangers Online for Free. One Day, the New York Times Called.
For nearly a decade, Dominique Ferris spent her evenings doing something most people would consider a waste of time: methodically dismantling false claims in online comment sections, one citation at a time. She had no journalism degree, no press credentials, and no expectation that anyone important was paying attention.
Someone was.
Today, Ferris leads a verification unit at one of the country's largest news organizations. She has a staff, a budget, and a methodology that's been adopted by three other major outlets. She built all of it from a secondhand laptop in a one-bedroom apartment in Akron, Ohio, doing work she never expected to get paid for.
The Habit That Started as Irritation
Ferris will be the first to tell you that her fact-checking career didn't begin with noble intentions. It began with annoyance.
"I was on a neighborhood Facebook group in 2011," she said in a 2022 interview with the Columbia Journalism Review. "Someone posted a completely fabricated statistic about crime rates, and everyone was just sharing it. I knew it was wrong because I had literally read the actual report two days earlier for work. So I posted the link. And then someone argued with me about it. And then I stayed up until two in the morning finding every primary source I could."
She won the argument. More importantly, she discovered she was good at this — and that she liked it in a way that felt almost compulsive.
Ferris worked days as an administrative coordinator at a regional insurance company. The job was steady and unremarkable. In the evenings, she started gravitating toward larger forums: Reddit threads, comment sections on news sites, early social media debates where misinformation spread fast and corrections rarely followed. She developed a system. Not just for finding accurate information, but for documenting how a false claim had traveled — which source it originated from, how it mutated as it moved, which phrases made it more shareable.
She started keeping records. Spreadsheets, mostly. Then a private blog where she'd write up her findings in plain language.
Building a Reputation Nobody Assigned Her
The thing about doing rigorous, consistent work in public — even when that public is a niche corner of the internet — is that people notice. Slowly, then suddenly.
By 2014, Ferris had a modest but unusually loyal following on her blog, which she'd named Primary Source and updated several times a week. Her readers weren't casual. They were researchers, librarians, a handful of journalists who'd started quietly citing her work without attribution. She was meticulous in a way that felt almost old-fashioned: she linked every claim to its original document, she noted when sources had been updated or corrected, and she was scrupulous about distinguishing between what the evidence showed and what it suggested.
That last habit — the intellectual honesty about uncertainty — is what set her apart from the crowded field of online debunkers who were gaining traction in the mid-2010s.
"A lot of fact-checking at that time was essentially point-scoring," said Marcus Webb, a media critic who wrote about Ferris for the Nieman Lab in 2019. "Dominique was doing something different. She wasn't trying to win. She was trying to be accurate. Those aren't the same thing, and readers could feel the difference."
She was also, critically, willing to correct herself. On at least four documented occasions, she published detailed posts walking back or refining earlier conclusions when new evidence emerged. In the attention economy, that kind of self-correction is rare. In professional newsrooms, it's considered essential. Ferris had arrived at it on her own.
The Call She Didn't Expect
In 2017, a senior editor at a major national newspaper reached out through her blog's contact form. He'd been following her work for two years. His organization was building a dedicated verification desk — one of the first of its kind at a publication of that scale — and he wanted to know if she'd consider interviewing for the role.
Ferris thought it was spam. She almost deleted it.
"I read it three times," she said. "Then I Googled his name to make sure he was real. Then I called my mom."
She got the job. She was thirty-one years old, had never worked in a newsroom, and had no formal training in journalism. What she had was a seven-year paper trail of methodical, documented, publicly available work that spoke more clearly than any résumé.
What She Built From the Inside
Ferris's verification unit has since become a model studied by journalism schools and adopted — in modified forms — by several other major outlets. Her core methodology isn't complicated, but it's demanding: every factual claim in a story must be traced to its primary source, not a secondary report. Uncertainty must be stated explicitly. And when a story can't be fully verified, the piece doesn't run until it can be.
That last principle has killed some stories. It has also, on at least two occasions, prevented her organization from publishing claims that later turned out to be false — claims that other outlets ran and had to retract.
"She brought a kind of discipline that's honestly hard to teach in journalism school," said a senior colleague who asked not to be identified. "Because in school, you're taught to be fast and accurate. Dominique came up in a world where she was only accountable to herself, and she chose accuracy over speed every single time. That's a different kind of character."
The Longer View
What Ferris built in those Akron evenings wasn't a brand or a platform in the modern influencer sense. It was a body of work — unglamorous, painstaking, largely unrecognized for years. She wasn't trying to get hired. She was trying to get it right.
The internet has a way of surfacing that distinction, eventually. The people who are doing the work because the work matters tend to look different, over time, from the people who are doing it to be seen.
Dominique Ferris spent nearly a decade being invisible. The record she left behind was anything but.