The Ten-Minute Window
A claim starts circulating online at 9:14 a.m. By 9:24, it has been repeated by a handful of accounts with real followings, stripped of whatever hedge or caveat the original source attached to it. By 9:40, the caveat is gone entirely, and the claim reads as settled fact. A newsroom deciding whether to write about that claim is working inside a ten-minute window that keeps shrinking, and the pressure to publish something — anything — before the window closes is real and constant.
Epoch Signal’s standards desk exists because of that pressure, not despite it. Our rule is simple to state and harder to follow: we would rather be second and correct than first and wrong, even when being second costs us the traffic that comes with being first.
What Verification Actually Involves
Fact-checking sounds like a single step — someone checks the facts — but in practice it is a series of specific, boring questions asked about every claim in a story before it runs. Where did this number originate, and is that source the number’s origin or just the latest place it was repeated? Does the document actually say what the story claims it says, or does it say something adjacent that got rounded up? If a source is quoted making a claim about someone else, has that person been given a chance to respond?
None of these questions are exotic. They are the same questions a careful reporter should be asking while writing the story in the first place. The standards desk exists as a second set of eyes precisely because a writer who has spent three weeks convinced of a story’s thesis is the worst-positioned person to spot the hole in it.
The Cost of Getting It Right
Slowing down has a price, and we do not pretend otherwise. Stories that clear our verification process sometimes run a day or two after a competitor’s version, by which point the news cycle has often moved on to the next thing. We lose some readers who wanted the fastest version of the story rather than the most accurate one. That is a trade we have decided to keep making, but it is worth being honest that it is a trade, not a costless virtue.
What we get in exchange is fewer corrections and, we would argue, a publication readers can cite without hedging. Both of those are hard to measure in a traffic report, and both matter more to us than the numbers a traffic report can show.
A Case in Point
Early last year, a claim spread that a regional hospital system had quietly closed a maternity ward without notice. It fit a familiar narrative about rural health care, and by the time it reached us, it had already been shared by outlets with far more readers than we have. Two calls — one to the hospital system’s communications office, one to a county health official — established that the ward had been reduced from around-the-clock staffing to scheduled hours, not closed. That is a real story, arguably still a concerning one, but it is not the story that had already gone out under different bylines. We ran the accurate, less dramatic version. It performed worse than the original claim had. We ran it anyway.
When We Get It Wrong Anyway
Verification reduces errors; it does not eliminate them. When a fact in a published story turns out to be wrong, we correct it at the top of the piece, not in a note appended at the bottom that most readers will never scroll to. The correction states what was wrong, what the accurate information is, and when the fix was made. We do not quietly edit a sentence and hope no one notices the difference between the version they read and the version that is live now.
This is not generosity on our part. A publication that hides its mistakes is asking readers to trust a track record it is not willing to show them honestly.
Speed Has a Place
None of this is an argument against speed itself. Some stories genuinely are time-sensitive, and holding a verified, important story for an extra round of checks that do not change its substance is its own kind of failure. The distinction we try to draw is not between fast and slow; it is between verified and unverified. A story can be both fast and checked if the reporting underneath it was done carefully from the start, which is really the only way to have both at once.
The most effective version of this happens early, not at the end. Standards-desk involvement that starts when a reporter first pitches a story — agreeing up front on what threshold of evidence a given claim will need before publication — tends to save time relative to a verification pass bolted on the night before deadline, when a hole in the reporting is far more expensive to fix and far more tempting to paper over.
