The 70 Percent Problem

Digital banking transformation has one of the worst success rates in financial services. McKinsey's research consistently puts transformation failure rates at 70 percent or higher — and in banking, where legacy infrastructure, regulatory constraints, and risk aversion compound the difficulty, the actual failure rate among programmes that set out ambitious digital-core modernisation goals is higher still.

I have spent the last eight years advising banks through transformation programmes. The failures are not random. They cluster around three structural patterns that repeat so consistently across institutions that I can now identify a programme at risk within the first planning workshop. Understanding those patterns is the difference between a transformation that delivers and one that spends three years and £50 million producing a shiny mobile app sitting on top of an unchanged COBOL core.

Failure Pattern 1: Technology-Led Without Business Architecture

The most common transformation failure is what I call the technology-first trap. A bank's leadership team approves a digital transformation programme, names a Chief Digital Officer or Chief Transformation Officer, and immediately begins a technology vendor selection process. RFPs go to Temenos, Thought Machine, Mambu, and Salesforce. A systems integrator is appointed. The work begins.

Six to twelve months later, the programme is in trouble. Not because the technology was wrong — often it was fine — but because the bank never articulated a clear business architecture before choosing the technology. The questions that should have been answered first were never asked: Which customer segments are we actually trying to serve differently? Which products are the growth priorities over the next five years? What are the unit economics we need to hit? How will our distribution model change?

Without answers to these questions, technology selection becomes arbitrary. Monzo's early growth trajectory is instructive here. Before Monzo's technology stack was famous, its business model was clear: serve a previously ignored customer segment — the cost-conscious, mobile-native young professional — with radical transparency on fees and a current account that actually worked for frequent travellers. The technology was an expression of that business model, not a bet made in advance of it.

The fix is to run business architecture work before technology selection, not after. Define the customer segment, the product strategy, and the target operating model in enough detail to actually constrain the technology choice. Banks that do this well shorten their vendor selection cycles by months and avoid the misalignment that produces failed implementations.

Failure Pattern 2: Ignoring the Operating Model

Legacy banks have operating models built around their legacy technology. Branch networks exist because account opening required physical presence. Manual credit underwriting processes exist because automated decisioning required integration between core banking and credit bureaus that was never built. Operations teams exist at scale because the technology never automated the exception-handling flows.

Digital transformation changes the technology. It rarely changes the operating model at the same pace — and that mismatch is the second major failure pattern.

A bank can deploy a cloud-native core banking platform, automate onboarding flows, and launch a modern mobile app — and still operate with a headcount structure, a governance model, and a risk management process designed for a 1990s operation. Both McKinsey and BCG framework for transformation emphasise operating model redesign as a parallel workstream to technology delivery. In practice, it is almost always underfunded relative to the technology investment.

Starling Bank's operating model is structurally different from a legacy retail bank — not just technologically different. Starling runs a far higher ratio of automated decisioning to manual review. Its operational exception handling is designed into product flows, not bolted on after launch. Its regulatory reporting is event-driven, generated automatically from the core system rather than compiled monthly from aggregated data.

Legacy banks cannot simply adopt Starling's technology and expect equivalent operational efficiency. The operating model — the processes, the organisational structure, the governance cadences, the risk appetite — has to change alongside the technology. Transformations that treat the operating model as a future consideration reliably fail.

Failure Pattern 3: Governance That Cannot Keep Pace

Digital product development operates on a two-week sprint cadence. Legacy bank governance processes operate on monthly risk committee cycles, quarterly investment approval boards, and annual strategic planning processes.

The governance mismatch is not just an irritant — it is a programme killer. When a product team needs a commercial decision to unblock a sprint, and the decision requires a paper to be prepared for the next risk committee meeting four weeks away, product velocity collapses. Engineers get demoralised. The transformation programme starts to look indistinguishable from the legacy change management processes it was meant to replace.

BCG's transformation research identifies governance model redesign as one of the three highest-impact levers available to transformation programmes. The interventions that work are: delegated authority structures that allow product teams to make commercial decisions below defined thresholds without committee approval; standing architecture review boards that meet weekly rather than monthly; and clear ownership accountability at the product-domain level, with named executives accountable for delivery and outcome metrics.

Monzo's scaling challenges between 2021 and 2023 included governance growing pains as the company moved from start-up speed to the operational rigour required by a fully licensed retail bank. The lesson from Monzo's evolution is that governance needs to scale — it needs to become more rigorous as complexity increases, but it cannot revert to legacy cadences without killing product velocity. Finding that balance is one of the hardest leadership challenges in digital banking.

What the Winners Do Differently

The transformations I have seen succeed share three characteristics that cut across the failure patterns above.

They start with a clear business case at the customer segment level. Not "we need to modernise our core banking system" but "we are losing market share in SME current accounts because our onboarding journey takes five days and Mettle does it in fifteen minutes. Our programme goal is to reduce commercial onboarding to forty-eight hours and reduce the cost of onboarding by thirty percent." Specific, measurable, connected to a customer outcome.

They treat operating model redesign as a first-class workstream. The transformation programme structure includes an operating model design track with dedicated resource, a budget, and named accountability. The technology workstream and the operating model workstream are designed to land together, not sequentially. Banks that sequence technology first and operating model later spend twelve months trying to retrofit the organisation to a system it was never designed to run.

They redesign governance for digital cadences before the programme begins. This typically means fighting a political battle internally — existing governance committees do not give up decision rights without resistance. The banks that execute well make this fight early, win it, and establish the delegated authority structures that allow product teams to move. Programmes that try to work around governance rather than change it end up with eighteen-month timelines for decisions that should take eighteen days.

The Leadership Imperative

Digital banking transformation is not a technology programme. It is an organisational transformation that requires technology to enable it. The difference is not semantic. It changes what the leadership team needs to sponsor, what gets measured, and what constitutes success.

Core banking modernisation is necessary but not sufficient. A bank with a modern core banking system and an unreformed operating model will deliver marginal improvement on customer experience and no improvement on unit economics. The banks that win — Starling, Monzo at maturity, Allica in the SME segment — are the ones whose leadership teams understood the full scope of what transformation requires and funded, governed, and sponsored accordingly.

The question for any bank leadership team contemplating a transformation programme is not which vendor to select. It is whether the organisation has the governance structures, the operating model clarity, and the business architecture discipline to make the technology investment deliver. Most do not — yet. The ones that invest in answering that question first are the ones that end up on the right side of the 70 percent statistic.

Need a transformation roadmap? Request a task with Nadia.

Nadia leads Digital Banking Transformation advisory at Aicura Consulting, specialising in core banking modernisation programmes, operating model redesign, and transformation governance for retail and commercial banks.