67% of large-scale finance transformation programmes miss their primary objectives, according to a study of 150 programmes. The three design choices that separate success from failure are not the ones most organisations are focused on.
Key Takeaways
The transformation failure statistics have been circulating in finance circles for years, but a new study from the Finance Transformation Research Consortium, covering 150 large-scale programmes at organisations with revenues above $500 million, conducted between 2019 and 2025, puts a precise number on the problem. 67% of those programmes failed to achieve their primary stated objectives. Not all failed catastrophically: many delivered some benefits, some technology implementations, some cost savings. But against the objectives that justified the investment, typically a combination of cost reduction, cycle time improvement, analytical capability uplift, and business partnering effectiveness, two-thirds fell short.
"The depressing thing about this data is that most of it was preventable," said Anita Patel, executive director of the Finance Transformation Research Consortium and a former global finance director at two FTSE 100 companies. "The patterns of failure are remarkably consistent across industries, geographies, and programme sizes. The organisations that succeeded made different choices at the beginning of their programmes, not better execution of the same approach, but fundamentally different design choices that changed the probability of success from the outset."
The Consortium research identifies three design choices as the primary differentiators between successful and failed transformation programmes. All three relate to sequencing, specifically, the order in which organisations address the three core dimensions of finance transformation: operating model, technology, and talent. Programmes that get the sequence wrong fail at dramatically higher rates than those that get it right.
The first choice is operating model before technology. Programmes that begin with technology selection, choosing an ERP, a planning tool, or a data platform, before clearly defining the operating model they are trying to build fail at more than twice the rate of those that establish the operating model first. The reason is straightforward: technology should enable an operating model, not define it. When technology is selected first, the programme ends up designing the operating model around the constraints of the chosen system, a recipe for sub-optimal outcomes regardless of how good the underlying technology is.
The second critical design choice is building analytics capability before automating transactional processes. The intuitive approach is to start with automation, it is tangible, measurable, and relatively quick to deliver. But the research shows that organisations that automate before they have strong analytics capability end up with faster bad information. They produce reports and dashboards more quickly, but those outputs don't drive better decisions because the analytical framework underpinning them hasn't been designed. Successful programmes invest in data infrastructure, analytics capability, and decision-support design before they scale automation, using automation to free up time that the analytics capability can then use to generate value.
"Every transformation we studied that started with the technology decision ended up in some form of compromise, either they over-customised the system to fit old processes, or they forced old processes into new systems without changing the work. Neither outcome justified the investment." , Anita Patel, Finance Transformation Research Consortium
The third design choice, treating talent as a design input rather than an afterthought, is the one most consistently underinvested in failed programmes. Organisations that define the talent model they need, the specific capabilities, career structures, and ways of working required by the target operating model, before they begin programme design are significantly more likely to succeed than those that treat talent as a workforce management problem to be solved once the operating model is built. The latter approach produces the same outcome almost every time: a newly designed finance function with insufficient human capability to operate it as intended.
Progressive finance functions are responding to the evidence of transformation failure by abandoning the large-scale, multi-year programme model entirely. Instead, they are building permanent transformation capability, dedicated teams with protected budgets, ongoing mandates, and a continuous improvement charter, that operates alongside the finance function rather than disrupting it periodically with high-risk, high-cost programme investments. The early results from organisations that have made this shift are promising: smaller, faster capability improvements delivered more reliably, with lower change management cost and higher adoption rates.
78% of CEOs now say their CFO is their most important strategic partner, up 37 points since 2019, per Deloitte research.
Finance analysts are leaving for tech and consulting at 3× the rate of five years ago, and standard retention tactics are not working.
Five capabilities separate transformational CFOs from traditional ones, and the financial consequences of the gap are measurable at 12 percentage points of TSR.
