The Avoidable Reliability Tax of Retail Banks

by Tobi Olayiwola

 From the first pay cheque you ever received to the estate your children will eventually settle, a retail bank is present at almost every financial inflection point of a person's life. No other commercial relationship spans that kind of arc. And yet most retail banks still treat trust as a brand metric, something to be managed in the communications department, measured in quarterly NPS surveys, and periodically refreshed with a new campaign.

It is not a brand metric. It is a margin lever. And the distinction matters more than most boards have yet acknowledged.

The Quiet Downgrade

Here is what most churn models miss: unhappy banking customers don't leave. They redistribute.

When a customer loses confidence in their bank through accumulated friction, they don't close accounts. They quietly restructure the relationship.

The salary still lands in the same place it always has, because changing a salary account is friction most people won't tolerate. But within 48 hours, a significant portion of that balance is elsewhere. The savings pot is with a fintech offering better rates and a cleaner interface. The investment account is at a platform the bank doesn't own.

The customer remains on the books. The economics of the relationship have quietly left the building.

This is what I call the reliability tax. It is not captured in churn reports because the customer never churned. It shows up in deposit stickiness, in cross-sell conversion rates, in cost-to-serve ratios that are higher than they should be for a customer who has supposedly been "retained," and in net interest margin that underperforms what the customer's balance sheet would suggest.

The Measurement Problem

Part of why reliability remains under-invested is that its financial consequences are diffuse and lagged. A customer who loses confidence in January may not move their savings until April, may not take the mortgage elsewhere until the following year, may not fully exit the relationship for another three years. By the time the margin impact is legible, the connection to a service failure eighteen months prior is difficult to establish.

Banks, in the meantime, optimise for what can be reported in the short term: satisfaction scores, complaint resolution rates, digital adoption figures. These are not without value, but they are proxies, and imperfect ones. A customer can register as “satisfied” after a resolution and still think less of the institution for having required one.

Reliability, as a financial concept, rarely arrives at the board in a form that commands serious capital allocation.

The EBITDA Case

Retail banking strategy conversations tend to orient around acquisition: new customers, new segments, new products. Acquisition is legible. It has a clear ROI model and a straightforward narrative for the board.

Reliability does not, which is partly why it is underfunded relative to what the economics would justify.

A fully engaged retail banking customer, one who holds their primary current account, savings, mortgage, and at least one insurance or investment product with the same institution, is worth several multiples of a customer who holds only the current account. The distance between those two states is not primarily a product problem or a pricing problem. It is a trust problem.

If even a fraction of customers who have quietly downgraded their relationships were re-engaged through consistently reliable service, the effect on cross-sell ratios, deposit stickiness, and net interest margin would be material within the current reporting cycle, not some future strategic horizon.

The question that belongs at board level is not how to improve the NPS score. It is what the reliability gap is costing in financial terms, and what closing it would require.

The Bottom Line

In retail banking, trust is not built through advertising. It is built through reliability, through being the institution that does what it said it would do, every time, without the customer having to verify.

That is not a customer experience objective. It is not an operational one. It is a commercial strategy. And reliability remains the most under-priced driver of EBITDA on the board agenda.

The reliability tax is real. The question is whether enough banks will name it before the redistribution becomes irreversible.