Qolo recently joined Fintech Takes Banking for a live conversation with John Withrow, Head of Commercial Card at KeyBank, moderated by Kiah Haslett. The discussion covered why commercial banking infrastructure is having a moment, what banks risk by standing still, and what it actually looks like to modernize without ripping out a working core. Five points from that conversation are worth pulling out on their own.
1. The infrastructure layer is plumbing, not the facade, and most banks have been investing in the wrong one
The conversation kept returning to a simple distinction: banks have spent years investing in the parts of the experience clients can see, while the layer underneath, the ledger, the reconciliation, the connective tissue between systems, went largely untouched. As one infrastructure veteran on the call put it, infrastructure is the plumbing and the electrical. It is not the cool facade or the sliding windows. It is the required necessity that lets someone else finish the house.
That distinction matters because banks do not need to choose between modernizing the customer experience and modernizing what sits underneath it. They need the second one to make the first one possible. A bank cannot build real-time cash visibility, flexible account structures, or embedded card programs on top of a core that was designed to clear batch transactions decades ago. The infrastructure layer is where the actual constraint lives.
2. Modernization does not require a rip-and-replace
One of the more practical points in the conversation was about sequencing. The old assumption in banking, and in infrastructure, was that you had to make a hard switch on day one. Stand up the new system, migrate everything, deprecate the old one in one motion. That assumption is no longer true, and it was never the only path.
The model that actually works is standing the new infrastructure up side by side with what already exists. Move volume over deliberately. Harden the new system in production before deprecating the old one. This is not a theoretical point. It describes how KeyBank approached its own card and ledger modernization, and it is the reason the transition did not require choosing between stability and progress.
3. Banks already have a leaky bucket, whether or not they can see it
A recurring theme was that banks are not necessarily losing clients in a single dramatic event. They are losing share gradually, in ways that are easy to miss because the primary relationship, the deposits, the loan, often stays put. What moves is the value-added activity around it: payment flows, treasury tools, card programs.
Stripe’s Treasury business alone generates roughly $800 million a year. That number exists because banks left an opening. The institutions winning that share are not necessarily winning on relationships or trust. They are winning because they built a better product experience and made it easy to adopt. A bank that does nothing is making a decision, and that decision has a cost that compounds.
4. The real opportunity is in the money-adjacent problems clients cannot fully articulate
The most concrete example from the conversation was not about payment speed at all. It was about an executive assistant managing purchasing across 25 property management divisions, fielding every purchase request by text and call, then spending 20 to 30 hours a month manually reconciling a single corporate card statement line by line.
That is not a treasury management pitch. It is an operational pain point that has nothing to do with whether money moves in real time. It has everything to do with whether a bank can offer ledger infrastructure that segments funds by entity, reconciles automatically, and removes the manual burden entirely. Clients in this position are not asking for virtual account management by name. They are describing a problem that virtual account management solves. Banks that can hear that distinction find the deal. Banks that wait for clients to ask for the product by name will keep missing it.
5. Push and pull both work, but only if the business case accounts for the cost of doing nothing
Getting internal buy-in for infrastructure investment is its own challenge, separate from the technology itself. The conversation surfaced a useful diagnostic question for any bank building a business case: are you responding to clients who are already asking for this, or are you building ahead of demand you believe is coming.
Both paths can work. Apple did not wait for anyone to ask for an iPod. But a push strategy only succeeds when it is grounded in real client behavior, not designed in a vacuum. The detail that made KeyBank’s business case credible internally was not just total addressable market modeling. It was attrition data: a clear accounting of what client relationships actually cost the bank when this gap goes unaddressed. That is the number that moves a business case from theoretical to defensible.
The throughline
Every one of these points traces back to the same root cause: the gap between what commercial clients now expect and what most bank infrastructure can deliver is a ledger and rails problem, not a relationship problem. Banks are losing because the infrastructure underneath the relationship was not built for what commercial banking has become.
Listen to the full conversation here or get in touch with the Qolo team to start your path to modernization today.
