Owning Banking vs Owning Workflows
Post-Brex Ramp vs Mercury
The split at the infrastructure layer reveals that Mercury is optimizing for owning banking economics, while Ramp is optimizing for shipping software faster. Mercury already makes most of its money from net interest on customer deposits, so taking the charter path lets it keep more of that spread and depend less on partner banks. Ramp, by contrast, is pushing finance teams into higher value workflows like procurement, bill pay, travel, treasury, and AI driven automation, where the bottleneck is product velocity, not bank ownership.
-
Mercury looks more like a future digital bank than a pure fintech front end. It reached about $650M annualized revenue by September 2025 on roughly $20B in deposits, applied for an OCC national bank charter in December 2025, and had already migrated banking partners after the Evolve disruption, making infrastructure control more valuable.
-
Ramp looks more like a finance operating system. It reached about $1B annualized revenue by August 2025 by selling a bundle that starts with cards and expense management, then adds bill pay, procurement, travel, and treasury. That mix favors software attachment and AI features that help finance teams review receipts, code spend, close books, and manage approvals faster.
-
The clean comparison is SoFi or Nubank on Mercury's side, and Coupa or AI native back office software on Ramp's side. Mercury's charter can widen margins by removing partner bank revenue share. Ramp's upside comes from getting more workflows per customer and being valued more like enterprise software than a spread business.
Going forward, the category is likely to separate into two winning models. One owns the balance sheet and turns deposits into durable bank economics. The other owns the workflow and turns finance data into software revenue and AI leverage. Mercury is moving deeper into the first model. Ramp is moving deeper into the second.