How Nathan Andersen Built a Different Kind of Financial News Operation
Nathan Andersen ran a newsroom that treated financial reporting like an engineering problem rather than a literary one. He hired people who could read a 10-K and a balance sheet without getting distracted by narrative flourishes. The entire approach centered on speed combined with source verification, and it changed how a lot of small-to-mid-size financial publications operate now. The core system was simple but expensive to implement well. Every claim in a piece had to trace back to a primary document or a on-the-record source. No anonymous sourcing except in rare circumstances, and even then the editor had to note why anonymity was necessary in the story itself. This meant slower initial output but dramatically fewer retractions and corrections. In my experience, publications that adopted this model saw their correction rate drop from roughly 4-6% per month down to under 1%, which matters a lot when you're competing against outlets that publish fast and apologize later. Andersen's team used a shared editorial database that tracked every source, every document, and every claim across stories. When a reporter filed an article, the fact-checker pulled each assertion against the database before it went live. If a source wasn't already in the system, the reporter had to provide verifiable contact information and the original document before the piece advanced. This took more time upfront but prevented the embarrassing situation where a story falls apart days after publication because a key claim couldn't be verified.
One thing most people miss about this approach is that it doesn't just affect accuracy. It fundamentally changes what stories your team can and cannot pursue. A lot of financial journalism relies on pattern recognition and contextual inference. Andersen's model deliberately avoided that. If you couldn't prove it from a document or a verified source, you didn't publish it. This sounds like it would make for dry reporting. It didn't, because the stories that made it through were almost always more compelling than the ones that got cut. I worked with a team that tried to replicate this system for a sector-specific newsletter covering biotech earnings. We hit a wall pretty quickly with companies that had unclear ownership structures and shell entities across multiple jurisdictions. The verification process that worked fine for standard US-listed corporations broke down entirely when we were dealing with Cayman Islands holding companies and variable interest entity arrangements. The workaround was straightforward but tedious. We stopped trying to verify every claim independently and instead required that any story about these kinds of structures include a documented methodology section explaining exactly what sources were consulted, what documents were reviewed, and what assumptions were made. Readers could then judge the reliability themselves rather than trusting editorial judgment alone. This added about twenty minutes to every article but eliminated a whole category of error that we were making before we realized we were making them. Source hierarchy was the other critical component. Andersen's team ranked sources in this order: regulatory filings and official company documents came first, followed by on-the-record company statements, then analyst reports from known research firms, then industry interviews, and finally general news reports. Each level carried different weight. A claim sourced to a 10-Q filing needed no additional corroboration. A claim sourced to an analyst report required at least one independent confirmation. This hierarchy wasn't arbitrary. It reflected the actual reliability differences between source types, which most outlets ignore at their peril.
The downside of this model is that it slows things down considerably. When a major market event happens and competitors are publishing within minutes, Andersen's team might take two to three hours to run the same story through verification. In fast-moving markets, that delay can mean missing the initial audience entirely. The strategy accepts this tradeoff deliberately. The assumption is that readers who care about accuracy will return for follow-up coverage rather than consuming the first draft available online. This works if you have an established audience and brand credibility. It does not work if you're trying to build traffic from scratch or compete with wire services on breaking news. Another practical challenge is personnel cost. This system requires editors who understand financial documents well enough to spot inconsistencies and reporters who can parse primary sources without relying on secondary summaries. These are specialized skills. Salaries for people with this capability run noticeably higher than the average business reporter rate, and the talent pool is smaller than most publishers realize. I've seen several outlets attempt this model and abandon it within eighteen months because they couldn't hire or retain staff with the right skill set. If you're considering implementing anything like this, start with the verification workflow rather than trying to copy the entire operation. Build a source database. Establish a clear hierarchy for what counts as evidence. Require documentation for every major claim. These three steps alone will improve your output quality significantly without the full overhead of Andersen's system. You can scale up from there.
Get the Full Details
The broader implication is worth noting. Financial journalism has drifted toward opinion-driven content because opinion drives engagement and ad revenue. Andersen's model proves that there is a viable audience for rigorously verified financial reporting, but the audience is smaller and less lucrative on a per-reader basis than the opinion crowd. Any organization adopting this approach needs to understand that they are playing a different game, not a better version of the same game. The metrics that matter are different. Retention rates matter more than page views. Trust compounds more slowly than virality but lasts longer. These are the tradeoffs that define the model.
