Beyond the Failure Rate: A Standard for Measuring Transformation Outcomes
Five tests in place of one failure rate, with a scorecard boards can request at every stage gate.
Research and data
Original surveys, a quarterly index and evidence reviews, each published with its method, questionnaire and data. Hypotheses for the annual study are pre-registered, and every report is reviewed by two external experts.
The Langford Transformation Outcomes Study
How often do large transformations meet their objectives, and what do the ones that succeed do differently? Each respondent reports on one specific program in an organization of 1,000 or more employees, including programs that were canceled or paused, so the results are not limited to survivors.
Success is measured with five separate tests rather than a single failure rate, and respondents say whether each outcome is documented or estimated. Findings are reported as associations, not causes.
- Sample
- 1,000 executive sponsors, program leads, change leads and business-unit leaders; at least 150 per program type and 300 from regulated industries
- Program types
- AI adoption, enterprise technology, M&A integration, operating model redesign
- Success tests
- Scope delivered; cost within budget; 80% of planned benefits at 12 months; adoption target met; attrition at or below baseline
- Pre-registration
- Open Science Framework, by January 22, 2027
- Fieldwork
- February 1 to March 19, 2027
- Publication
- June 15, 2027: report, data tables, questionnaire and methods appendix, free without registration
- Peer review
- Two external academics before publication
The Langford Change Load Index
How much change employees in large organizations are asked to absorb, relative to their capacity to absorb it. Each quarter, 2,500 employed adults in organizations of 250 or more staff across the United States, the United Kingdom, Germany, India and Australia.
The index is the simple average of four shares on a scale of 0 to 100, where higher means heavier load. Movements within the margin of error, about two points, are not reported as changes. First release: Tuesday, April 13, 2027, at 10:00 a.m. US Eastern.
- ConcurrencyShare reporting three or more significant changes to their work in the past 90 days
- Capacity gapShare who lack the time and support to adapt
- Clarity gapShare who do not understand why the changes are being made
- StrainShare who say the changes have made it harder to do their job well
Library
Where the 70 Percent Change Failure Rate Came From, and What to Measure Instead
Traces the most repeated statistic in change management to its sources and sets it against three decades of survey evidence.
Read
Change Load: What the Evidence Says About Saturation, Fatigue and Intent to Stay
Separates what employees report from what leaders estimate, two measures that are often confused.
Sponsoring an AI Rollout: The Decisions an Executive Cannot Delegate
The calls a sponsor has to make in person, from workflow redesign to how adoption is measured.
Validated Change: Transformation Under GxP, Model Risk and Safety Rules
How regulated firms plan change when validation, audit trails and training records set the pace.
The First 100 Days of an Integration: People Decisions That Set the Value Case
Retention, role mapping, leadership appointments and the order of communications, with templates.
The Middle-Manager Squeeze: Disruption, Engagement and a Role No One Redesigned
What the research shows about managers in sustained change, and what it does not.