How RawCents Actually Works
RawCents runs out of a small research desk, not a newsroom with a masthead and a floor of staff writers. Nobody here answers to a bank, and there’s no advertiser we’re trying to keep smiling. What we’re chasing is simpler than it sounds: figure out where people are quietly losing money to the system, and say something before it costs them, not after.
Since there’s no bigger organization backing what we publish, the accountability lands directly on us. If a piece is wrong, that’s ours to own, in public. Below is roughly what has to happen before something goes up on the site.
Where the research actually comes from
We try not to write the same article that already exists on four other finance sites, just phrased differently. When a piece covers something like reward tier structures on a card, or a fee that’s technically disclosed but practically buried, the goal is to trace it back to source material rather than lean on someone else’s summary of it.
A lot of that starts with federal complaint databases. These aren’t polished; they’re raw case records of what actually goes wrong for customers, filed directly, without a PR filter smoothing anything over. Reading through enough of them, patterns start showing up — the same billing error, the same runaround, the same service breakdown repeated by people who’ve never spoken to each other. That’s usually where an idea for a piece begins.
Beyond the official records, there’s value in watching what people say in the moment, on forums, in scattered comment threads. Banks don’t always announce changes clearly. Credit limits get trimmed quietly. Fee structures shift. A lounge that used to have room suddenly doesn’t during peak hours. Customers usually notice these things weeks before any official statement catches up, so it helps to be watching the ground level rather than waiting on a press release to confirm what’s already happened.
When a product that’s been marketed heavily turns out to be unremarkable once you actually look at the terms, that gets said outright. There isn’t much interest in dressing up a forgettable financial product as a smart choice just because its ad spend is large.
Making sure the work holds up
There’s no separate editor sitting above this team double-checking the output, so the checking has to happen inside the process itself, and it tends to be more demanding because of that, not less.
A draft usually sits for a bit before it’s finalized. Coming back to it later, with some distance, makes it easier to spot a claim that isn’t quite earned or a line of reasoning that only works if nobody looks too closely.
Numbers get recalculated by hand rather than trusted on the strength of having been correct once already. Reward math, interest projections, payoff timelines — these get run again, separately, before anything ships.
Links pointing to legal disclosures, enforcement actions, or a bank’s own published fee schedule get opened and checked one by one. A citation pointing to the wrong page, or one that’s technically related but not quite what was claimed, doesn’t count as verification.
If something in that process doesn’t hold, the piece sits until it does.
What happens when something’s wrong
Rates change. Terms change. Banks update policy without much warning, and pretending this desk will always be current would be its own kind of dishonesty. What matters more is what happens once an error surfaces.
If a claim can’t be traced back to an actual document, disclosure, or verifiable source, it doesn’t get included, regardless of how likely it seems. A reasonable guess still isn’t a fact, and treating it like one just to fill out a paragraph isn’t something this team does.
If something looks outdated, a link stops working, or a bank changes terms overnight and the article hasn’t caught up yet, readers flagging it is genuinely useful. That kind of correction is welcome, not resented.
And corrections aren’t handled quietly in the background. When something material gets fixed, the change is noted directly in the piece, so anyone reading it later can see what shifted and roughly when.
None of this is a claim to being flawless. It’s closer to being upfront about the parts that weren’t, which matters more in the long run than pretending otherwise.