Spade

Spade sits at a specific layer of the payments stack, and most of the confusion around it comes from comparing it to the wrong product. Plaid and Finicity solve connectivity: they link an account and return balances and transactions. Spade assumes the transaction record already reached you — from your card processor, your issuer feed, or a third-party source — and its job is to make that record legible. Merchant name, category, and exact geolocation get appended before the authorisation decision closes, which is the whole point of the 40ms P99 figure. The four workflows on the site are the clearest guide to fit. Risk & Authorization is about having merchant and location context inside the approval window, which is what Coast describes: real-time controls at the pump that approve legitimate fuel purchases without creating a new class of false declines. Rewards & Attribution is about matching each purchase to a programme at merchant and location granularity — Bilt's neighbourhood-level rewards on its platform spend is the reference case. Analytics & AI is about consistent merchant context across cards, transfers, and third-party sources, so your models learn from one vocabulary rather than five. User Experience is the quiet one with real ROI: if the customer recognises the merchant, the dispute often never gets filed. The published numbers are worth stating plainly and worth verifying: 99.9% merchant coverage in the U.S. and Canada, 99% enrichment accuracy, exact geolocation on 95% of physical transactions, and a 99.9% uptime SLA. That last figure is the one to hold Spade to — if your authorisation path depends on enrichment, an enrichment outage becomes an authorisation outage. The honest limits are geographic and operational. Coverage is stated for the U.S. and Canada only, so if a meaningful share of your volume is cross-border you should ask how those transactions come back before you design around it. Exact geolocation resolves on 95% of physical transactions, meaning roughly one in twenty will not carry an exact location — build your rules with that margin rather than assuming a perfect field. And the low-configuration API model, with no per-merchant custom setup, is an advantage for shipping speed but means your edge cases get handled by the standard enrichment rather than bespoke mapping rules. Where it fits: fintechs building rewards with location precision, issuers tightening real-time authorisation, fleet and fuel-card providers applying merchant and location controls, expense platforms replacing hand-maintained merchant rules, and AI teams that need a clean ground-truth transaction record. Where it doesn't: batch or month-end enrichment jobs that don't care about milliseconds, buyers shopping for account connectivity, and teams whose volume sits predominantly outside North America.