Geyto Research · Mariano D'Angelo · Evidence review · · 2 min read

The 70% myth. What the evidence says about change failure

The claim that 70% of organizational changes fail is repeated widely and has rarely been shown. What we know, and what to measure instead.

In brief: the claim that "70% of organizational changes fail" is widely cited in change management. An academic review traced its sources and found no valid and reliable empirical evidence to support it. That does not mean changes go well. It means the figure is no help in understanding why they go wrong.

A figure that travels on its own

The figure shows up in presentations, consulting reports and leadership books. Its appeal is obvious: it is round, alarming, and it justifies investment in any change program.

In 2011, Mark Hughes, of the University of Brighton, did something unusual: he looked for where it came from. He reviewed five publications that cited it as fact. His conclusion was that there is no valid and reliable empirical evidence for a 70% failure rate. What exists is a narrative that gets repeated.

The problem is not only the number

Even if the figure were correct, it would have little practical value. "Failure" can mean very different things:

  • that the system was never fully installed;
  • that it was installed but is not used;
  • that it is used but did not change the expected result;
  • that it worked in one site and not in another.

Each of those situations has different causes and calls for different responses. A single rate mixes them together and hides what matters: where, in which part of the work, and why the change did not take hold.

What believing it leads to

An alarming figure pushes people to respond with more of the same: more communication, more training, more pressure on the organization. Those responses are sometimes right. But the number does not say whether the problem lies in the organization's capacity or in what the initiative asks for, in what order and with what support.

It also creates a fatalistic expectation. If almost everything fails, a stall looks normal and stops being investigated.

What to look at instead

Instead of asking whether a change "failed", it is more useful to ask questions that have answers:

  • Which part of the work changed, and which part is the same?
  • Where is the old way still alive, and what does it solve?
  • Does what the initiative asks for match the capacity available in each place where the work happens?
  • Which conditions are missing: time, judgement, coordination, support?

These questions are less striking than a percentage, but they make it possible to act.

A test for reading other figures

Hughes's review leaves a useful rule for any market statistic on change and technology adoption: before using it, ask who measured it, how, and on what sample. If no one can answer, the figure is an opinion formatted as data.

Geyto Research publishes essays, evidence reviews and market notes on what happens after go-live. More in our approach.

Frequently asked questions

Is it true that 70% of organizational changes fail?

There is no valid and reliable empirical evidence to support that figure. A review published in the Journal of Change Management traced its sources and concluded that it is a repeated narrative, not a finding.

Where does the 70% figure come from?

It has appeared in management books and articles since the 1990s. The sources that cite it do not present a study that measures it systematically.

What should be measured instead of a failure rate?

Whether the change has become part of the work and holds: which routines changed, where the old way is still alive, and which conditions are missing in each place.

Sources

  1. Hughes, M. (2011). Do 70 per cent of all organizational change initiatives really fail? Journal of Change Management, 11(4), 451–464.

Share