Why Most Change Efforts Fail, and What the Research Actually Says

You've done this before. A strategy session, a new set of priorities, maybe OKRs. There was energy in the room. People nodded. And then, somewhere between the kick-off and the end of Q1, the initiative quietly died. Not dramatically, just slowly, as everyone drifted back to what they were doing before.

If that sounds familiar, you're in the majority. Research across more than 200 companies found that 75% of change efforts fail to make dramatic improvements. The cause is usually neither a wrong diagnosis nor bad intentions. It is something more specific, and more fixable.

The Track Record Is Worse Than Most Leaders Realize

The data on organizational change is not encouraging:

• Reengineering efforts in Fortune 1000 companies succeed in only 20–50% of cases
• 50–75% of mergers and acquisitions fail to meet the expectations set at the start
• Even five years after a completed merger, 70% of surviving organizations remain underperformers in their industry
• Only 15% of IT projects are commonly rated as successful
• 50% of companies that downsized saw productivity fall, not rise

None of these are niche interventions. These are the most common, most resourced change efforts organizations attempt. And they fail at rates that should be making leadership teams ask harder questions about the model, not just the execution.

One finding stands out: success is highly correlated with visible sponsorship from leadership. Failure is consistently associated with missing or conflicted leadership at the top. Managers in the same survey agreed far more clearly on why change succeeds than on why it fails, which suggests the path forward is clearer than organizations usually admit.

The Problem Is What Happens After the Framework

Most change initiatives fail not in the planning room but in the months that follow. A new strategy is announced, a new system is launched, a new set of OKRs is set, and then leadership returns to their regular workload. The initiative becomes one more thing on a long list.

There's a pattern here that organizational research has been describing for decades. Rensis Likert, studying organizations at the University of Michigan in the 1950s, noticed that companies would commission diagnostic surveys, receive clear findings, and then do nothing with the results. He called it the "ask-but-don't-act" pattern, and his research showed it produced greater frustration among employees than not asking at all.

The same dynamic plays out in strategy and OKR rollouts today. A leadership team that commissions a process and then doesn't visibly follow through fails to make progress and actively erodes the trust needed to try again.

You Can't Change One Thing and Leave Everything Else Intact

There's a second failure mode, and it's subtler. Organizations consistently treat change as if it can be localized: fix the strategy, leave the culture. Launch OKRs, leave the accountability dynamics. Restructure the team, leave the way decisions get made.

A system doesn't work that way. A change in any part of an organization changes other parts, with both predictable and unpredictable consequences. When the technical layer (plans, structures, metrics) changes but the human layer underneath (alignment, conflict capacity, ownership, trust) doesn't, the technical change doesn't hold. The new strategy lives in a document. The OKRs exist until week 6. The restructure creates new friction instead of resolving the old kind.

This is a systems problem, not a people problem. And it explains why leaders who have done everything "right" on process still find themselves in the same place twelve months later.

What Makes Change Stick

The organizations that successfully navigate change share a few consistent characteristics, none of which are about having a better framework.

Ownership sits with the team, not the consultant. Edgar Schein, one of the most respected thinkers in this field, built his entire approach around one principle: the client owns the problem and the solution, and the client is the people who have to live with the outcome, not the consultant or the change manager. When an external partner writes the strategy and hands it over, the team receives it but doesn't own it. A strategy you receive and a strategy you built are two completely different things, regardless of the quality of the content.

Diagnosis leads somewhere. Assessment without a committed action phase is not neutral. It raises expectations and then fails to meet them, which damages the trust needed for the next attempt. Any diagnostic (an engagement survey, a team effectiveness assessment, an OKR health check) should come bundled with a clear path to action. If that path isn't defined before the data is collected, the process has already started to fail.

Sponsorship is sustained, not performed. The kick-off all-hands is not sponsorship. Sponsorship is the CEO who says no to a new initiative because it conflicts with the three priorities they committed to in January. It's the leadership team that references the strategy in a budget decision in March. It's the visible, repeated, public demonstration that the change is real, and that it costs something.

What This Means for Your Next Initiative

Before launching the next strategy process, the next OKR cycle, or the next team restructure, ask a few honest questions:

Who will hold this publicly when the energy fades, and are they committed to doing that?

Is this a technical change (new system, new process) or a behavioral change (how decisions get made, how the team holds each other accountable)? The two require very different interventions at very different depths.

If we commissioned an assessment, do we have a clear plan for what happens after we see the results, including what we're prepared to do differently?

The data on change failure is sobering, but it describes a pattern most organizations repeat without examining; it is not a verdict. The organizations that break it are the ones willing to hold the change visible, sustained, and honestly owned by the people who matter, not the ones with better frameworks.