The statistic everyone cites regarding change management is a 70% failure rate. I spent a day reviewing the studies that are supposed to support this figure, as well as related ones. Some are very well-founded: a sample, a method, a data collection date. For others, including the most frequently cited one, there’s nothing to back it up.
The Numbers, and What Lies Behind Them
| Number | What It Measures | Scope | Source | Year |
|---|---|---|---|---|
| 70% | Change programs that fail to meet their objectives | No sample, no methodology, no study cited in the article | McKinsey, Changing Change Management | 2015 |
| 70% | Transformations that fail | “Research shows,” unnamed study; McKinsey Transformation Practice | McKinsey, interview with Jon Garcia | 2022 |
| No valid evidence | Critical review of five publications citing the 70% rate | Academic literature review, Journal of Change Management, pp. 451–464 | Mark Hughes, University of Brighton | 2011 |
| 30% | Digital transformations that met or exceeded their value targets with sustainable change | 825 executives surveyed internationally + internal data on 70 client companies | BCG, Flipping the Odds of Digital Transformation Success | 2020 |
| 44% and 26% | “Worry zone” (partial value) and “woe zone” (less than 50% of the target) | Same sample of 825 executives | BCG | 2020 |
| 93% vs. 15% | Projects meeting or exceeding their objectives, depending on whether change management is rated as excellent or poor | Participants in Prosci’s Best Practices studies, international, self-reported | Prosci, The Correlation Between Change Management and Project Success | 2021 |
| 79% vs. 27% | Objectives achieved based on the effectiveness of the executive sponsor | Prosci Best Practices in Change Management, 12th edition | Prosci | 2023 |
| Less than 70% | Participants reporting that they follow a structured methodology, the lowest level since 2011 | Prosci Best Practices in Change Management, 12th Edition | Prosci, 6 Notable Shifts | 2023 |
| +24 points | Increased probability of change success with an “open source” approach rather than a top-down approach | Gartner HR practice, figure reported without a published sample | Gartner, press release dated October 14, 2020 | 2020 |
| Capacity halved | Volume of change an employee can absorb without becoming fatigued, 2020 compared to 2019 | Gartner HR practice, sample not published in the press release | Gartner | 2020 |
| 74% | French employees who believe that change has not improved their well-being at work | 1,044 employees from public and private companies with more than 500 employees, France, quotas + regional stratification | Empreinte Humaine Barometer / IFOP | 2016 |
| 68% | Employees who believe that the change did not make their work more efficient | Same sample of 1,044 employees | Empreinte Humaine / IFOP | 2016 |
| 63% | Employees who did not have the opportunity to express their views on the change | Same sample of 1,044 employees | Empreinte Humaine / IFOP | 2016 |
| 80% to 100% | Percentage of a project’s benefits that depend on a change in people’s practices (“adoption contribution” from the Prosci CMROI model) | Prosci proprietary model, based on experience, no published sample | Prosci, Why Change Management | 2021 |
| 74% vs. 42% | Executives reporting that they involved employees, vs. employees who feel involved | Figure reported by Prosci, primary source not found | Prosci, Why Change Management Fails | 2024 |
The 70% figure has no source
McKinsey wrote in July 2015 that “70% of change programs fail to meet their objectives, largely due to employee resistance and a lack of management support.” Third sentence of the article. No footnote, no sample size, no year of data collection. In March 2022, the firm republished the figure in an interview with Jon Garcia, a senior partner in its Transformation practice: “Research shows that their efforts fail about 70% of the time.” What research? We’ll never know.
Mark Hughes, a researcher at the University of Brighton, did the legwork as early as 2011. He reviewed five publications that cited this rate and tracked down the data behind each one. His verdict, in the Journal of Change Management: there is no valid and reliable empirical evidence to support this narrative. His article is fifteen years old. The figure is still on slides.
That doesn’t mean transformations succeed. It means that 70% is a common belief, not a measurable fact.
BCG, at least, publishes its methodology
The 2020 BCG study takes the opposite approach, and I find it honest. It surveyed 825 executives, plus internal data from seventy companies they advised. They were asked to rate their transformation on a scale of 1 to 10 based on targets achieved and value created. The result: 30% in the “win zone,” 44% in the “worry zone,” and 26% in the “woe zone.”
Look at who’s doing the rating: the executives who led the transformation. I’m not saying they’re sugarcoating things. I’m saying that a program manager who stayed within budget and delivered their deliverables on schedule checks “goals achieved”—and from their perspective, they’re right.
The scenario I see that doesn’t fit into any of these categories
A medium-sized manufacturing company with four sites. ERP go-live on a Monday in March. By Friday: zero blocking incidents, orders are processing, payroll is out. The vendor wraps up, and the project team moves on to something else. In any analysis of this table, this project counts as a success.
In October, I’m on-site for another matter. The paint shop scheduler opens the ERP system to check the quantities, then switches to an Excel spreadsheet she maintains herself, which includes the actual dates and machine downtime. The ERP says one thing, her spreadsheet says another—and it’s her spreadsheet that drives the shop floor. She’s been doing this since June. No one asked her to do it or told her not to: no one knew about it.
I ask her why. “When I first started, I tried it—it took me twenty minutes to figure out where to enter a delay, and I had a team waiting.” Just once. It got stuck at the wrong moment, so she reopened Excel. She never went back.
Another site, another tool, same pattern. A CRM system deployed with a real sponsor and team liaisons—exactly what Prosci recommends. Six months later, the records are filled out on Fridays around 5 p.m., in batches, from memory. Monday’s forecast is built on that. Project successful. Data without value.
My Disagreement with the Prevailing View
The usual take: transformations fail because people resist, so we need more communication and more sponsorship.
I don’t buy into that—and not for ideological reasons. Almost all of these studies measure the same thing: the project’s failure at completion, as assessed by those who led it. None of them measure the actual use of the software six months later. That’s where it falls apart, and it’s the only moment that matters to the company that paid for the license.
The Empreinte Humaine/IFOP survey is the only one in this table that surveys employees rather than project leaders, and it paints a very different picture. Of 1,044 employees at companies with more than 500 people, 74% say the change did not improve their well-being, and 68% say it did not make their work more efficient. That was in 2016, but I haven’t seen anything since then that contradicts that on the ground.
As for resistance to change, the term bothers me more and more. The planner wasn’t resisting anything. She had a team waiting and an interface that took twenty minutes to navigate. She made a decision in three seconds, just like anyone else would. Calling that “resistance” shifts the blame onto the person when the problem actually lies with the screen.
If I had only one metric: the percentage of active users at the six-month mark who complete the target task within the tool, without using a separate file. It’s measurable, no one publishes it, and I suspect the figures are significantly less flattering than BCG’s 30%.
A word about what’s missing, because it’s noticeable. There’s no average cost in this table, no quantified ROI. I looked for them. What’s circulating is either unverifiable or published by those selling the service. The only figure of this kind that I was able to source is Prosci’s 80% to 100% range, derived from their own model and not from a sample-based measurement. It is included in the table with this caveat. The same logic applies to the Standish Group’s CHAOS database: since it’s sold as a paid report, there are no verifiable figures, so none are cited.
FAQ
Is the 70% failure rate for change management true?
It has not been proven. In 2011, Mark Hughes reviewed five publications citing this rate and found no valid empirical evidence to support it. McKinsey repeated the claim in 2015 and 2022 without citing any studies.
What is the most reliable figure regarding the failure of digital transformations?
The one from BCG in 2020: a 30% success rate, based on 825 executives surveyed, with a published methodology. It remains anecdotal and is assessed by the project leaders themselves.
What do French employees say about change in the workplace?
The 2016 Empreinte Humaine / IFOP survey from 2016, conducted among 1,044 employees at companies with more than 500 employees, indicates that 74% believe the change has not improved their well-being and that 63% did not have the opportunity to voice their opinions on it.
Sources
- McKinsey, Changing change management, July 1, 2015 — https://www.mckinsey.com/capabilities/people-and-organizational-performance/our-insights/changing-change-management
- McKinsey, Common Pitfalls in Transformations: A Conversation with Jon Garcia, March 29, 2022 — https://www.mckinsey.com/capabilities/transformation/our-insights/common-pitfalls-in-transformations-a-conversation-with-jon-garcia
- Mark Hughes, Do 70 Percent of All Organizational Change Initiatives Really Fail?, Journal of Change Management, 2011, pp. 451–464 — https://doi.org/10.1080/14697017.2011.630506
- BCG, Flipping the Odds of Digital Transformation Success, 2020 — https://www.bcg.com/publications/2020/increasing-odds-of-success-in-digital-transformation
- Prosci, The Correlation Between Change Management and Project Success, December 8, 2021 — https://www.prosci.com/blog/the-correlation-between-change-management-and-project-success
- Prosci, 7 Change Management Best Practices (12th edition), May 26, 2023 — https://www.prosci.com/blog/change-management-best-practices
- Prosci, 6 Notable Shifts in the Best Practices in Change Management, 2023 — https://www.prosci.com/blog/6-notable-shifts-in-the-best-practices-in-change-management
- Prosci, Why Change Management, May 20, 2021 — https://www.prosci.com/blog/why-change-management
- Prosci, Why Change Management Fails — https://www.prosci.com/blog/why-change-management-fails
- Gartner, Gartner Cautions HR Leaders That the Risk of Change Fatigue Among Employees Has Doubled in 2020, October 14, 2020 — https://www.gartner.com/en/newsroom/press-releases/2020-10-14-gartner-cautions-hr-leaders-that-the-risk-of-change-fatigue-among-employees-has-doubled-in-2020-this-year
- Empreinte Humaine / IFOP Barometer on Human-Centered Change Management, May 2016 — https://www.ifop.com/publication/barometre-sur-laccompagnement-humain-du-changement/ (study report: https://www.ifop.com/wp-content/uploads/2018/03/3400-1-study_file.pdf)
I am the director of AdoptionLayer, which focuses on software adoption in the enterprise. The real-world scenarios described in this article are composite; the figures are not—each is cited from its source above.