NORBERT BIEDRZYCKI

The Transformation ROI: Why Most Change Programs Fail to Deliver Their Promised Value

The Transformation ROI: Why Most Change Programs Fail to Deliver Their Promised Value

In July, Gartner predicted that at least 30 percent of generative AI projects will be abandoned after the proof of concept stage by the end of 2025, citing poor data quality, inadequate risk controls, escalating costs and unclear business value.

The last reason is the most interesting. Not technology failure. Unclear value. And it is not new. It is the same reason a large share of digital transformations, ERP implementations and operating model redesigns have disappointed for twenty years.

The gap between implementation and adoption

Most transformation failures are not failures of implementation. The platform goes live. The new organization chart is published. The training is completed. Every milestone on the plan is green.

The value simply does not appear.

The reason sits in the gap between implementation and adoption.

A platform can be deployed without changing the process. The old workflow continues, now with an additional system on top of it.

A structure can change without changing decisions. New boxes, same approvals, same meetings.

Training can be completed without changing behavior. Certificates are issued, habits remain.

Robert Solow observed in 1987 that you could see the computer age everywhere except in the productivity statistics. The explanation then was the same as now. Technology creates potential. Organizations decide whether that potential becomes value.

Some expensive lessons

The history of large programs is full of warnings. Lidl abandoned a major SAP based project in 2018 after spending a reported half a billion euros, in part because the company was not willing to adapt its own processes to the system. The technology was not the main problem. The unresolved question of who should adapt to whom was.

At the other end of the spectrum, Klarna announced in February that its AI assistant was handling two thirds of customer service chats in its first month, doing work it said was equivalent to 700 full time agents, with an estimated 40 million dollar profit improvement this year. Whether those numbers hold over time remains to be seen. But note how they are expressed: in volume, in cost and in profit, not in number of licenses deployed.

Measuring the right thing

Most organizations measure transformation by inputs and outputs: money spent, systems launched, people trained, use cases piloted. Those are measures of activity.

The measures that matter are outcomes: productivity per employee, revenue growth, margin, customer retention, cycle time, working capital. They are harder to attribute and slower to appear, which is why they are avoided.

From the perspective of a board or a private equity owner, the discipline is straightforward.

Every major initiative has a value hypothesis expressed in P&L terms before it starts.

A named business owner, not a project manager, is accountable for that value.

Value tracking continues after go live, for at least a year. This is where most programs stop measuring, and exactly where the value is won or lost.

Funding for later phases depends on value from earlier ones.

None of this is sophisticated. It is simply rare.

Leadership question: Why do companies measure transformation by what they implement rather than by the business value they actually create?

Read next

  1. Transformation Is Not a Project: Why Organizations Need Continuous Reinvention

    Transformation and Leadership series, written in 2026

  2. The Transformation Imperative: Why Incremental Change Is No Longer Enough

    Transformation and Leadership series, written in 2026

  3. The Great Reset: Why Returning to Normal Is the Wrong Strategy

    Transformation and Leadership series, written in 2026