Quality Driven Customer Selection in BaaS
Roy Ng, co-founder and CEO of Bond, on BaaS's business model
Bond was signaling that early BaaS winners are pulled more by market demand and ecosystem fit than pushed by classic software sales. In practice, the best customers were companies that already had a real product, users, and transaction flow, then wanted to layer on cards, accounts, or payments. That makes customer acquisition look less like buying leads and more like careful filtering, because one breakout client can drive a disproportionate share of volume and revenue.
-
For all in one BaaS platforms, the product already bundles hard things, bank partner access, compliance workflows, processor integrations, and ledgering. That naturally creates inbound interest from fintechs and software companies that want to launch financial features without stitching together banks, KYC vendors, and processors themselves.
-
The real selection work happens after inbound arrives. Bond described deep KYB and fit screening, while other operators in the market framed BaaS portfolios like venture portfolios, where the job is to identify teams with funding, clear distribution, and a strong core business before they become the next Square like customer.
-
This approach also reflects BaaS economics. The platform often shares thin interchange revenue with banks, networks, and processors, so bad customers are expensive and risky, while good customers compound through higher spend, more products, and better retention. That is why larger brands and vertical SaaS companies became more attractive over time than startup serving startup models.
Going forward, customer acquisition in embedded finance will look even more concentrated around quality signals, existing distribution, and compliance readiness. The winning platforms will not be the ones that buy the most demand, but the ones that turn inbound interest into a repeatable process for picking durable customers, launching them quickly, and expanding them into higher value financial products over time.