When Neobanks Should Build Their Stack
Founder of neobank company on the importance of picking the right sponsor bank
This is the moment where a neobank stops being a reseller of someone else’s stack and starts becoming a real financial software company. Early on, an all-in-one BaaS package buys speed, but as volume grows the expensive pieces become obvious, especially card manufacturing, ledgering, compliance tooling, and processor fees. The biggest gains come from unbundling those line items, owning the sponsor bank relationship, and keeping more of the interchange economics.
-
The practical trigger is not just size, it is complexity. Teams usually start moving off the bundled stack when they want a new product, custom KYC and AML rules, more bespoke physical card programs, or direct dialogue with the sponsor bank. That is when off the rack infrastructure stops fitting.
-
The money case is straightforward. BaaS providers make much of their revenue from interchange splits, monthly account fees, and subscriptions, and those economics compress as fintech customers scale and negotiate. What looks like convenience at launch can turn into millions per month of avoidable cost later.
-
The migration path is usually modular, not a full rewrite. A company may keep the sponsor bank and some compliance workflows in place, but swap in a specialist like Lithic or Marqeta for card processing, use its own ledger or a separate ledger provider, and move card printing or dispute operations to cheaper vendors.
The likely end state is a more unbundled market, where BaaS wins the first launch and specialist infrastructure wins the scaled customer. The strongest neobanks will keep using partners for regulated balance sheet access, but they will increasingly own the product logic, data flows, and unit economics that matter most.