Since 2018, we have built our data rather than bought it
Everything you pull off this platform comes off a pipeline we own end to end — filings we parse ourselves, operating metrics our analysts read out by hand, and a sector and industry taxonomy that is our own work. It is an expensive way to run a data company, and it is the only way we know of to always be able to tell you where a number came from.
Six USPs
Each one exists because we built it in-house. That is also the reason we can extend it, correct it, or license it to you.
BQ-MICS, our own classification
A three-level sector and industry system for US-listed equities that tags a company in every industry it genuinely operates in, with one marked Primary so a clean single-industry view survives. Nobody else publishes it, and because it is ours it is also licensable.
See the system →The full SEC stack, end to end
A real-time filings feed, our own parsers, document and item-level extraction, and normalisation into our templates — every stage of it ours. Nothing waits on a vendor to process a filing first, which is how a figure can reach the platform minutes after it is filed, ahead of the feeds most of the market reads.
Where the data comes from →20,000+ KPIs read out by analysts
Store counts, deliveries by model, churn, take rates, backlog, ARPU. None of it carries a standard tag, so our equity research analysts pull it out of earnings decks, transcripts and supplements by hand, and hold every metric to one definition.
Browse the metrics →The whole API, on every paid plan
Not a separate five-figure product. A documented REST API and a realtime WebSocket feed, covering statements, filings, segments, estimates, screening, ownership, corporate actions, dividends and economic data.
Read the docs →Our own statement classification
We wrote the templates every statement is mapped to, standardised into four industry-specific ones and applied across five SEC filing types. That is what makes one company’s line items structurally comparable with the next, with no field mapping at your end.
See the statements API →Customisable licences
Almost all of it is built in-house, so no upstream vendor’s terms sit between you and what you want to do. Display, redistribution and enterprise licences are ours to write and ours to tweak. Paid plans also add new companies and KPIs on request.
Tell us the use case →A data company that happens to have a good front end
Reading a 10-K is not the hard part. Having twenty years of them already standardised, segment splits already broken out, and somebody who has typed a store count out of forty consecutive quarterly decks — that is the hard part, and it is what a professional research desk quietly pays for. Everything cheap enough for an individual investor stopped at a summary income statement.
We started in 2018 to close that particular gap, and decided early that we would not close it by reselling somebody else’s feed under a nicer interface. Buying data means inheriting another company’s gaps and another company’s definitions, and having nothing useful to say when two numbers disagree.
8 years later that decision still shapes nearly everything: what we can fix when you report a problem, how fast we can add a company you actually follow, and whether a licence for something you are building is a conversation or a refusal. It is also why most of what we spend our days on is the data, and the rest is making the data quick to reach.
Where our numbers turn up
Our data and research have been cited by financial publishers and market outlets since the platform’s early years.
Small enough that a real person reads your email
Engineers and equity research analysts, spread across time zones rather than gathered into one office, on a product where the difficult part is measured in datasets rather than headcount. Small is a choice here, not a stage we are trying to grow out of: it is what lets us ship in weeks, and it is why the person answering your email about an odd-looking figure is usually close enough to the data to go and check it.
Every message reaches someone on the team rather than a queue. Straightforward questions are often answered within minutes; anything that needs a number verified against its source takes as long as verifying it honestly takes. If something on the platform looks wrong we would far rather hear it than not — write to [email protected], or use the contact page.