The Langford Institute https://www.thelangfordinstitute.org/ The Langford Institute for Organizational Transformation and Change Management: independent journalism, original research and annual honors on how organizations change. Every figure links to its original source. en Sun, 04 Oct 2026 08:00:00 GMT Where the 70 percent change failure rate came from, and what to measure instead https://www.thelangfordinstitute.org/articles/where-the-70-percent-change-failure-rate-came-from/ https://www.thelangfordinstitute.org/articles/where-the-70-percent-change-failure-rate-came-from/ Sun, 04 Oct 2026 08:00:00 GMT Research and Data The claim that 70 percent of change programs fail began in 1993 as a self-described unscientific estimate about reengineering. A peer-reviewed review found no reliable evidence for it, and the survey record points to better measures. For three decades, management books, Harvard Business Review, and consulting research have repeated that 70 percent of change programs fail, as a premise rather than a finding. The figure can justify a change budget or excuse a disappointing result before it arrives.

The Research Desk traced the figure back through its sources and set it against the survey evidence. It does not hold up. The research around it is more useful.

An estimate becomes a rule

The earliest widely cited source is Reengineering the Corporation (1993), in which Michael Hammer and James Champy offered what they called an unscientific estimate: as many as 50 to 70 percent of organizations that undertake reengineering fall short of the dramatic results they set out to achieve.

Each qualification matters: the estimate concerned reengineering, not change in general; it described falling short of dramatic gains, not failure; and it was an upper bound, unscientific by its authors' own account.

Two years later, in The Reengineering Revolution, Hammer and his co-author Steven Stanton wrote that the observation had been widely misread and turned from a description into a rule. Reengineering, they argued, has no inherent rate of success or failure; results depend on the quality, intensity, and intelligence of the effort.

The correction did not stop the drift. In a 1996 book, Harvey Robbins and Michael Finley reported Hammer and Champy as conceding that 70 percent of reengineering efforts fail to achieve any results. Falling short of dramatic results had become a failure to produce any. A 1998 book by Ronald Purser and Steven Cabana attributed a flat 70 percent failure rate for change efforts generally to a survey by Fortune magazine. Within five years, an upper bound for one method had become a failure rate for all change.

Assertions without a dataset

Two Harvard Business Review articles are often cited as sources of the figure. Only one contains it.

John Kotter's "Leading Change: Why Transformation Efforts Fail" (March-April 1995) gives no overall rate. Of the more than 100 companies he had watched, Kotter reported, a few change efforts were very successful, a few failed outright, and most fell in between, tilting toward the weaker end.

The other, Michael Beer and Nitin Nohria's "Cracking the Code of Change" (May-June 2000), is among the most widely cited sources of the flat figure. The authors wrote: "The brutal fact is that about 70% of all change initiatives fail." They gave no source, sample, or definition of failure.

Kotter put his own number on failure in 2008, 13 years after his article. In A Sense of Urgency, he estimated from years of study that more than 70 percent of needed change is never launched, never completed, or finishes late and over budget with its original aims unmet. That is an expert's estimate with a broad definition of failure, not a measured rate.

Consulting research then cited Kotter. A 2009 McKinsey article credited his research with finding that only 30 percent of change programs succeed, and a footnote in McKinsey's 2021 transformation survey quotes his 70 percent estimate. Each citation added authority without adding evidence.

In 2011, in the peer-reviewed Journal of Change Management, Mark Hughes of the University of Brighton examined five published instances of the figure. He accepted that a popular narrative of 70 percent failure exists but found no valid and reliable empirical evidence for it. He did not claim that most change succeeds. A 2015 review by Carlos Cândido and Sérgio Santos found that most estimates behind claims that 50 to 90 percent of strategic initiatives fail rest on evidence that is outdated, fragmentary, fragile, or absent.

Traced to its sources, the 70 percent figure is a chain of estimates, restatements, and assertions, not a measurement.

What the surveys measured

None of the large surveys reviewed here found that 70 percent of programs failed outright. Most set a demanding bar for success and group everything below it together.

Traced to its sources, the 70 percent figure is a chain of estimates, restatements, and assertions, not a measurement.

In McKinsey's July 2008 survey of 3,199 executives, about a third reported success, defined as a true step change in performance. Reading the other two-thirds as failures misstates the result: in the survey's unweighted response counts, about one respondent in 20 called the transformation not successful at all, and most called it somewhat successful. In McKinsey's 2015 survey, 26 percent of respondents said their transformation had been very or completely successful at both improving performance and sustaining the gains, rising to 79 percent among completed transformations that took all 24 actions the survey tracked. Its 2021 survey found fewer than a third successful on the same test.

IBM's 2008 Making Change Work study of 1,532 practitioners used a project manager's test: 41 percent of projects fully met their time, budget, and quality objectives, 44 percent missed at least one, and 15 percent missed all of them or were stopped. Boston Consulting Group's 2020 study of digital transformations found 30 percent met or exceeded their target value with sustainable change, 44 percent created some value but missed their targets, and 26 percent delivered less than half of it. BCG's own summary put the share falling short of objectives at 70 percent, a figure easily misread as a failure rate.

Bain & Company reported in April 2024 that only about 12 percent of business transformations achieve their original ambition. Missing an ambitious target is not failing, and a measure built on original ambition can penalize the clear, high aspirations that McKinsey's 2008 survey associated most strongly with success.

Three cautions apply. The research is self-reported, usually by people judging programs they took part in; definitions of success differ, so rates cannot be compared or averaged; and most of it comes from firms that sell transformation or change services. Even so, in the three studies here that publish a full breakdown, the largest group is the same: programs that partly succeeded.

What leaders should measure instead

The useful question is not how often change fails in general but what a given program is delivering, against what definition, and for how long. The research suggests four measures.

A definition of success agreed before launch. The studies above use at least five different tests, from time, budget, and quality to original ambition. In McKinsey's 2008 survey, nearly a quarter of respondents said their target had not been well defined, and nearly half said they would set clearer targets if they started again, the most common answer. McKinsey's 2021 respondents placed nearly a quarter of lost value at target setting.

The share of potential value captured. In McKinsey's 2021 survey, respondents at successful transformations estimated that they had captured 67 percent of the maximum financial benefit available, against 37 percent elsewhere. By their own estimates, even the successes left a third of the value behind.

Results after the program closes. The same respondents placed a fifth of lost value after implementation. A program measured only at go-live will miss it.

The conditions associated with results, read as signals rather than proof. In IBM's study, practitioners who always followed formal change management procedures reported a 52 percent project success rate, against 36 percent for those who improvised. In Prosci's benchmarking, 88 percent of practitioners who rated their change management excellent met or exceeded objectives, against 13 percent who rated it poor, and active, visible sponsorship ranked as the top contributor to success. McKinsey's 2015 survey tied success most closely to communication, active leadership, employee empowerment, and continuous improvement; Bain's research, to how well a company secures and develops the talent a transformation needs. These are correlations, often from respondents rating both inputs and outcomes. They are leading indicators, not guarantees.

The 70 percent figure is simple, alarming, and easy to repeat. Its replacement is less quotable: a definition of success set in advance, a measure of value captured, a check on whether gains last, and a few conditions to watch. That is harder to fit on a slide. It is also something a leader can manage.

Sources

]]> AI's hardest problems are people problems. Change budgets say otherwise. https://www.thelangfordinstitute.org/articles/ai-hardest-problems-are-people-problems/ https://www.thelangfordinstitute.org/articles/ai-hardest-problems-are-people-problems/ Sun, 04 Oct 2026 08:00:00 GMT AI Adoption BCG's 2024 survey attributes about 70 percent of AI implementation challenges to people and process. Monitor Deloitte's 2025 study finds change management and communications get an average of 9 percent of transformation budgets. Boston Consulting Group surveyed 1,000 senior executives in 2024 and found that about 70 percent of the challenges companies face in implementing AI stem from people and process issues. Technology problems accounted for 20 percent of the challenges in BCG's survey, and the algorithms themselves for 10 percent.

Transformation budgets are weighted differently. Monitor Deloitte's 2025 Chief Transformation Officer Study found that change management and communications received an average 9 percent of transformation budgets, the smallest of six spending areas. Technology received 26 percent in one of the Deloitte report's charts and 28 percent in another.

The two findings do not measure the same thing. Deloitte's figure covers transformation budgets in general, not AI programs specifically, so the comparison is suggestive, not like for like. BCG's figure describes where problems arise, not where money is spent. Still, the pairing raises a fair question for any AI program: whether its budget is aimed at the problems it is most likely to meet.

What the evidence supports

BCG's 2024 survey covered CxOs and senior executives in 59 countries and more than 20 sectors. It found that 74 percent of companies had yet to show tangible value from AI, and it classed 26 percent as leaders. BCG's release names change management among the people and process capabilities that set those leaders apart.

The 70 percent figure needs careful handling. BCG also presents its 10-20-70 split as a rule for allocating resources, one it says leaders follow, and describes it as a long-held belief. It is best treated as a BCG heuristic backed by survey data rather than an independently validated measurement. It is also often misquoted as saying that 70 percent of AI's value comes from people. BCG's claim is about where the challenges lie.

Newer surveys point the same way. In McKinsey's 2026 State of AI survey, about 6 percent of respondents qualified as AI high performers, meaning they attribute 5 percent or more of EBIT to AI and describe its value as significant. Nearly three-quarters of those high performers reported fundamentally redesigning workflows because of their AI use, up from 55 percent in 2025. Among other respondents, the share was one-quarter. McKinsey also found high performers twice as likely to say senior leaders demonstrate commitment to AI and to have defined processes for measuring its impact.

Workers tell a similar story. BCG's 2026 AI at Work survey of 11,749 workers in 14 markets associates a clear strategy with a 25 percentage point lift in measurable business impact from AI, against about 5 points for better tools alone. Respondents in companies pursuing workflow redesign were 24 points more likely to see measurable improvement.

If the issue log and the budget disagree, the budget is the one to change.

The strongest case against

The best argument against this reading comes from Deloitte's own sample of 200 transformation executives. More than 90 percent had overseen three or more transformations, and more than 80 percent led programs that met all their targets. On average, these leaders reported giving 9 percent of their 2024 transformation budgets to change management and communications, and only 33 percent would increase that share in hindsight. If 9 percent were enough to sink a program, it would be odd to find it as the average in a sample dominated by leaders who met all their targets.

The budget line is also narrow. Deloitte counts talent as a separate spending area, and some people and process work, such as training or workflow redesign, may be booked under other lines. The 9 percent may understate what organizations spend on the human side of change.

The evidence that people and process work pays off has limits too. The McKinsey and BCG findings are self-reported associations, not controlled experiments. Companies already getting value from AI may find it easier to redesign workflows, rather than the reverse. McKinsey's high-performer group is small, so its percentages carry wide margins. BCG's measure of business impact is what employees perceive, not audited results. BCG, McKinsey and Deloitte are all consulting firms with an interest in this conclusion.

These objections weaken any claim of proof. They do not reverse the direction of the evidence. The AI-specific research points to people and process as the harder part of adoption. Deloitte's experienced leaders lean the same way in hindsight for transformations in general: 42 percent would raise investment in talent, and 33 percent would raise spending on change management and communications, the smallest line in the budget.

What to check in your own program

Industry averages matter less than an organization's own numbers. Four checks can show whether an AI program is funded for the problems it is likely to meet.

  • Split the AI budget the way Deloitte does, separating technology from change management and communications, and from talent. If the split cannot be produced, that is a finding in itself.
  • Tag each open problem in the program's issue log as people and process, technology or algorithm, the categories BCG used. Then compare the share of problems in each category with the share of money.
  • Count the workflows that have changed because of AI, and name an owner for each redesign. Report that count alongside the number of tools deployed.
  • Agree on how impact will be measured before the program scales. McKinsey's high performers were twice as likely to have defined processes for measuring AI's impact, and a measure set early lets later budget debates rest on results.

Run the first two checks side by side. If the issue log and the budget disagree, the budget is the one to change.

Sources

]]> Change fatigue has become a retention problem https://www.thelangfordinstitute.org/articles/change-fatigue-retention-problem/ https://www.thelangfordinstitute.org/articles/change-fatigue-retention-problem/ Sun, 04 Oct 2026 08:00:00 GMT Operating Models and Workforce Gartner found in 2022 that 43 percent of employees with above-average change fatigue intend to stay with their employer, against 74 percent of those with low fatigue. The gap measures intentions rather than exits, but it is wide enough to put change load on the retention agenda. Gartner found in 2022 that only 43 percent of employees with above-average change fatigue intend to stay with their organization, against 74 percent of employees with low fatigue. Change fatigue is easy to file under morale. On this evidence it belongs with retention, where risks have owners and numbers attached.

The Gartner result needs careful handling. It is a correlation, and the release does not say how many employees were surveyed, in which countries or when. Intent to stay is what employees say they plan to do, not measured turnover. The case for treating fatigue as a retention risk therefore depends on whether other evidence agrees.

Other evidence points the same way

A 2017 survey for the American Psychological Association, an online Harris Poll of 1,512 employed US adults, found that workers going through organizational change were more than three times as likely as those with no change to plan to leave within a year, at 46 percent against 15 percent. The same 2017 APA survey found them more than twice as likely to report chronic work stress, at 55 percent against 22 percent. These figures are dated and cover only the United States. The comparison group reported no recent, current or anticipated change, so the gaps are associations, not proof that change drives people out.

HR leaders see the strain from above. In a July 2024 Gartner survey of 473 HR leaders, 73 percent said their employees were "fatigued from change." That is the judgment of the people who manage the workforce, not a report from employees, and Gartner's 2024 figure should not be restated as 73 percent of employees being fatigued.

Change practitioners describe organizations close to their limit. In Prosci's 2023 benchmarking study, 80 percent of change professionals said their organization was near, at or past the point of change saturation. They were describing their organizations rather than themselves, and saturation was self-assessed, with no common threshold. Prosci sells change management methodology and training, and its respondents chose to take part.

Each source has a weakness. Between them they draw on employees and on the HR leaders and practitioners who observe them, in different years and by different methods, and none of the four points the other way.

The strongest case against

The weakest link is the outcome. None of these sources counts resignations. A stated plan to leave becomes an exit only if people have somewhere to go, and none of the surveys measures that.

The direction of cause is also open. Employees who have already decided to leave may describe change more harshly than those who plan to stay. Fatigue and intent to leave may also share a cause, such as a poor manager or weak trust in leadership.

Nor do these sources show the problem growing. Each is a snapshot from a different year, taken with a different method, so they cannot be joined into a trend. The argument here rests on the strength of the link, not on evidence that fatigue is rising.

Gallup's work on trust is the most serious challenge to a story about volume alone. Gallup reported in 2024 that among US employees who had experienced significant disruption, those who strongly agreed they trusted their organization's leadership were 4.5 times as likely to be engaged and 62 percent less likely to feel burned out. Those 2024 Gallup figures are correlations, drawn from Gallup Panel web surveys of US employees conducted in 2023, and both are relative measures rather than percentage points. They suggest that how change is led matters alongside how much of it there is. If trust explains much of the gap, cutting the number of initiatives will do less than changing how they are run.

That narrows the argument without overturning it. Whether the driver is volume, trust or both, the cost lands in the same place: employees who say they will not stay. Governing change load and repairing trust are two responses to one retention risk.

Governing change load and repairing trust are two responses to one retention risk.

What to measure

The practical step is to test the link inside your own organization rather than rely on other people's averages.

  1. Add a change fatigue question and an intent-to-stay question to the next pulse survey, so results can be compared team by team.
  2. List the planned changes that reached each team in the past 12 months, from restructures to system replacements.
  3. Compare fatigue and intent to stay between teams carrying light and heavy change loads.
  4. Check actual attrition in the same teams after six and 12 months, to see whether stated intentions became exits.

The results answer the question for your organization in a way that external surveys cannot. If teams under heavy change report lower intent to stay but lose no more people, fatigue is a morale cost worth managing. If attrition rises as well, change load is a retention risk and should be reported as one.

Two practices have some support. Prosci's 2023 study found that respondents who actively managed their portfolio of change projects met or exceeded objectives more often, which argues for a single view of all the change landing on each team. Gartner's 2022 release says involving employees throughout a change can cut the risk of change fatigue by 29 percentage points, an approach Gartner calls "open-source" change. The release gives no sample size, so treat that figure as a direction to test rather than a forecast.

Exit reviews already ask why people left. Recording how much planned change each leaver absorbed in the preceding year would supply the outcome data these surveys lack.

Sources

]]> CEOs say AI depends on adoption. They estimate a quarter of staff use it. https://www.thelangfordinstitute.org/articles/ceos-ai-adoption-quarter-of-staff/ https://www.thelangfordinstitute.org/articles/ceos-ai-adoption-quarter-of-staff/ Sun, 04 Oct 2026 08:00:00 GMT AI Adoption In IBM's 2026 CEO Study, 83 percent of chief executives said AI success depends more on people's adoption than on technology. The same leaders estimated that only 25 percent of their workforce uses AI regularly. Of the chief executives in IBM's 2026 CEO Study, 83 percent said AI success depends more on people's adoption than on technology. By the same leaders' estimate, only 25 percent of their workforce uses AI regularly. Yet 86 percent believe their employees have the skills.

The IBM Institute for Business Value and Oxford Economics surveyed 2,000 CEOs and equivalent senior leaders in 33 geographies and 21 industries between February and April 2026. The 25 percent is the CEOs' estimate for their total workforce, not measured usage. IBM, which published the study, sells AI software and consulting.

Read together, the answers say something about where CEOs think the problem lies. If employees have the skills but only a quarter use AI regularly, then on the CEOs' own account something other than skills is holding usage back.

A chief executive who believes adoption decides AI's success should not have to estimate it.

Other surveys ask other people

Other 2026 surveys report higher use and wider benefits. They asked different people different questions, and their figures should not be set against IBM's as if they measured the same thing.

BCG's 2026 AI at Work survey of 11,749 workers in 14 markets found that 74 percent of frontline employees use AI regularly. BCG defines frontline employees as individual white-collar employees without managerial responsibilities, not deskless or shop-floor workers. The figure comes from workers describing their own use, in a narrower group than the total workforce IBM's CEOs were estimating for. The two numbers are not comparable, and the distance between them does not show that CEOs underestimate use.

BCG also found that 42 percent of frontline employees who use AI regularly save at least a full workday a week, by their own estimate. Yet 66 percent say they get limited or no guidance on what to do with that time. BCG's release does not state the base for the 66 percent; it appears to be regular frontline users.

McKinsey's 2026 State of AI survey of 1,719 participants in 97 nations is a third population again. It found that 80 percent of respondents say AI has improved their individual productivity, while only 37 percent attribute any EBIT impact to AI, essentially unchanged from 2025. Both are self-reported perceptions, not audited financials.

SourceWho answeredFigureWhat it measures
IBM CEO Study, 20262,000 CEOs and equivalent leaders83%Say AI success depends more on people's adoption than on technology
IBM CEO Study, 2026Same CEOs25%Their estimate of the workforce using AI regularly
BCG AI at Work, 202611,749 workers in 14 markets74%Frontline white-collar non-managers who say they use AI regularly
BCG AI at Work, 2026Same survey (base not stated)66%Get limited or no guidance on using the time AI saves
McKinsey State of AI, 20261,719 respondents in 97 nations80%Say AI has improved their individual productivity
McKinsey State of AI, 2026Same respondents37%Attribute any EBIT impact to AI

Measure adoption instead of estimating it

Every figure in the table comes from what people say, not from what systems record. A chief executive who believes adoption decides AI's success should not have to estimate it.

  • Define regular use, then report it monthly from system data by function and level. Pair it with an anonymous survey to catch use of personal tools, which system data will not show.
  • Ask regular users what happens to the time AI saves them, and decide team by team what those hours are for. BCG's 66 percent figure suggests many users are left to work that out alone.
  • Test the skills assumption. Most of IBM's CEOs believe their employees have the skills; a short practical assessment in one function would show whether that holds.
  • Report one business measure next to usage. McKinsey's respondents report personal productivity gains far more often than any EBIT impact, so usage and value need separate measures.

Sources

]]> Employees report AI productivity gains. Most companies can't find them in operating profit. https://www.thelangfordinstitute.org/articles/ai-productivity-gains-operating-profit/ https://www.thelangfordinstitute.org/articles/ai-productivity-gains-operating-profit/ Sun, 04 Oct 2026 08:00:00 GMT AI Adoption In McKinsey's 2026 survey, 80 percent of respondents say AI improved their productivity. Only 37 percent attribute any EBIT impact to it. In McKinsey's 2026 State of AI survey, 80 percent of respondents say AI has improved their individual productivity. Only 37 percent attribute any EBIT impact to AI, a share essentially unchanged from 2025. The gains are easier to report than to find in operating profit.

The survey's details matter. McKinsey fielded it online from May 4 to June 8, 2026, among 1,719 participants in 97 nations, and weighted the results by each nation's share of global GDP. Of the respondents, 36 percent work at organizations with revenue above $1 billion. EBIT, earnings before interest and taxes, is a standard measure of operating profit. Both figures are self-reported perceptions, not audited financials, and the EBIT measure counts any positive contribution, however small. The 37 percent is a share of respondents, not of companies or AI projects, so it should not be read as a failure rate.

The distance between the two numbers is where management comes in. Saved time reaches operating profit only when it turns into more or better output that customers pay for, or into lower costs. Neither happens by default.

Where the saved time goes

Employees report the time savings directly. BCG's 2026 AI at Work survey of 11,749 workers in 14 markets found that 42 percent of frontline employees who use AI regularly save at least a full workday a week. Yet 66 percent say they get limited or no guidance on what to do with that time, and over half do not redirect it to strategic work.

Two caveats apply. BCG defines frontline employees as individual white-collar employees without managerial responsibilities, not deskless or shop-floor workers. The time savings are self-estimated, and BCG's release does not state the base for the 66 percent, which appears to be regular frontline users.

If BCG's figures are close to right, much of the time AI saves arrives without instructions. Some of it may go into better work that no survey captures. Some may simply be absorbed. An organization that has not decided what the saved time is for has little reason to expect it to show up in EBIT.

What separates the high performers

McKinsey's survey does identify a group that reports the connection. About 6 percent of respondents qualified as AI high performers, meaning they attribute 5 percent or more of EBIT to AI and describe its value as significant. That share is unchanged from 2025. It is often misquoted as the share of companies getting any value from AI, when 37 percent report some EBIT impact.

The high performers stand out for redesigning work. Nearly three-quarters of them report fundamentally redesigning workflows because of their AI use, up from 55 percent in 2025, compared with one-quarter of other respondents. McKinsey also found them twice as likely to say senior leaders demonstrate commitment to AI and to have defined processes for measuring its impact. In their commentary, McKinsey's authors argue that adoption lags the technology because organizations struggle to absorb change.

BCG's results run in the same direction. Its 2026 survey associates a clear strategy with a 25 percentage point lift in measurable business impact from AI, against about 5 points for better tools alone. Respondents in companies pursuing workflow redesign were 24 points more likely to see measurable improvement and 22 points more likely to save a full day a week. The pattern fits the arithmetic of saved time. Redesign is the point at which an organization decides what the freed hours will do.

Neither finding is proof. Redesign correlates with higher reported impact, but companies already getting value may simply find redesign easier. The high-performer subgroup is small, so its percentages carry wide margins. BCG's lifts compare groups of respondents, not the results of controlled experiments, and its measure of impact is what employees perceive.

Saved time reaches operating profit only when it turns into more or better output that customers pay for, or into lower costs.

A measurement problem as well as a management one

The strongest objection to this argument is that the gap is partly an artifact of measurement. EBIT moves for many reasons at once, and isolating AI's share is hard. A company could gain real value from AI while survey respondents cannot point to it in EBIT. McKinsey's respondents also cite benefits that EBIT does not capture, such as innovation and employee satisfaction. BCG found that respondents in companies pursuing workflow redesign were 20 points more likely to report higher job satisfaction, a gain no income statement records.

A second objection is timing. In the same McKinsey survey, 44 percent report AI scaling across the enterprise, up from 38 percent. Benefits from that scaling may not have reached the accounts yet. The flat EBIT figure offers a partial answer: scaling rose over the year while the share reporting any EBIT impact stayed about the same. One year is too short to settle the question.

Both objections lead to the same practical step. If AI's value is hard to see in the accounts, measure it where it occurs, one workflow at a time, before scaling further.

  1. Choose one workflow where AI use is already regular, and record its baseline cost and output.
  2. Decide in writing what the saved hours are for, and tell the people doing the work. In BCG's survey, 66 percent said they get limited or no guidance on what to do with saved time.
  3. Redesign the workflow around that decision instead of adding AI to the old process.
  4. Track cost and output each month, and report the change in terms the finance team would accept as a contribution to EBIT.

A program that cannot show gains in a single workflow is unlikely to find them in operating profit.

Sources

]]> Sponsorship remains the strongest signal in the change data https://www.thelangfordinstitute.org/articles/sponsorship-strongest-signal-change-data/ https://www.thelangfordinstitute.org/articles/sponsorship-strongest-signal-change-data/ Sun, 04 Oct 2026 08:00:00 GMT Leadership and Sponsorship In Prosci's 2023 benchmarking, projects with extremely effective sponsors met objectives 79 percent of the time, against 27 percent with extremely ineffective ones. Sponsorship has topped Prosci's list of contributors to success in every study since 1998, but the data come from practitioners rating their own projects. In Prosci's 2023 benchmarking report, projects with extremely effective sponsors met their objectives 79 percent of the time, against 27 percent for projects with extremely ineffective sponsors. The finding has a long record. Active and visible sponsorship has been the top-cited contributor to success in every Prosci study since 1998, and the executive summary of Prosci's 2023 edition calls it "the single greatest contributor."

Prosci reports a similar pattern for change management as a whole. In its benchmarking research, as reported in 2023, 88 percent of practitioners who rated their change management excellent met or exceeded project objectives, against 13 percent of those who rated it poor, roughly a sevenfold difference. Prosci's 2023 figures put practitioners who rated it good or fair in between, at 73 and 39 percent.

Other research points the same way, less sharply. McKinsey's 2015 survey found four practices most closely tied to transformation success: communicating effectively, leading actively, empowering employees and continuous improvement. Active leadership, the closest match to sponsorship, made the list, but communication contributed most. Sponsorship is the most consistent signal in Prosci's data and one of the strongest in McKinsey's.

How the research was done

Prosci's benchmarking surveys change practitioners who choose to take part. Its 2023 study, the 12th edition, drew 2,668 respondents from 101 countries. Each respondent judges whether a project met its objectives and rates the sponsor and the change management behind it. Prosci sells change management methodology and training, and its data come from self-selected practitioners rating their own projects.

The evidence is consistent enough to act on and uncertain enough to test.

The sevenfold figure has two stated bases. Prosci's 2023 executive summary presents it as drawn from more than 10,800 respondents across its benchmarking research, while a 2023 Prosci blog post reporting it cites more than 2,600 practitioners in the latest study.

McKinsey's 2015 results come from an online survey fielded in November 2014 among 1,946 executives, 1,713 of whom had been part of a transformation in the previous five years. McKinsey weighted the self-reported answers by national GDP. Like Prosci's data, the 2015 survey shows correlation, not cause.

What it does not show

Neither source shows that sponsorship causes success. In Prosci's data the same person rates the sponsor and judges the outcome, so a project that went well can make its sponsor look better in hindsight. Practitioners who opt into a benchmarking study may also differ from those who do not.

The headline comparison takes the two ends of Prosci's sponsor-effectiveness scale. It shows the widest gap on that scale, not the typical one. The common shorthand that sponsorship triples success overstates what was measured.

A second figure from Prosci's 2023 study matters as much to leaders: only 48 percent of participants said they had an effective or very effective sponsor.

What to do with it

The evidence is consistent enough to act on and uncertain enough to test. Leaders can do both with the programs they already run.

  • Rate the sponsor before the result is known. Ask the change lead and the program lead to score the sponsor at launch and at each major milestone, and keep the scores. Compare them with outcomes at close. That removes the hindsight problem in the benchmark data and builds evidence of your own.
  • Define active and visible in observable terms. For each sponsor, list the messages they will deliver in person and the decisions they own, with dates. Then check whether those things happen.
  • Write the sponsor's part in communication into the plan. McKinsey's 2015 survey found communication contributed most to success. Specify which messages the sponsor delivers and to whom.
  • Treat a weak sponsor as a program risk. In Prosci's 2023 data, fewer than half of participants reported an effective sponsor, so a sponsor who falls short is common. Put the gap on the risk register with an owner and a remedy.

Sources

]]> Ten planned changes a year, up from two https://www.thelangfordinstitute.org/articles/ten-planned-changes-a-year/ https://www.thelangfordinstitute.org/articles/ten-planned-changes-a-year/ Sun, 04 Oct 2026 08:00:00 GMT Operating Models and Workforce Gartner counted 10 planned enterprise changes per employee in 2022, against two in 2016. Over the same years, willingness to support change fell from 74 to 38 percent, on Gartner's own figures. Gartner reported in 2023 that the average employee experienced 10 planned enterprise changes in 2022, up from two in 2016. Over the same period, the share of employees willing to change their work behaviors to support organizational change fell from 74 percent in 2016 to 38 percent in 2022, according to Gartner's Workforce Change Survey.

Measure20162022Source
Planned enterprise changes experienced by the average employee in the year210Gartner
Employees willing to change work behaviors to support organizational change74%38%Gartner Workforce Change Survey

On Gartner's figures, the change load rose fivefold between 2016 and 2022 while willingness to support change roughly halved. The count is an average per employee per year, not a tally of programs per organization. In a 2023 article in Harvard Business Review, Gartner's researchers gave a restructure and the replacement of a legacy technology system as examples of a planned enterprise change.

Why we use 38 percent

Two versions of the willingness figure circulate. Gartner's own publications give 38 percent for 2022: its October 2022 article "What Will HR Focus on in 2023?" and the first-quarter 2023 issue of Gartner Business Quarterly. An article in Harvard Business Review by the same Gartner researchers, published in May 2023, gives 43 percent for the same comparison, and 43 percent is the version most often repeated online. We use 38 percent for 2022 because Gartner published it under its own name, in two places. We do not mix the two figures.

A second pairing adds to the confusion. A separate 2022 Gartner finding also uses 43 and 74 percent: only 43 percent of employees with above-average change fatigue intend to stay with their organization, against 74 percent of those with low fatigue. That measures intent to stay, not willingness to change.

What the numbers do not show

Gartner has not published the sample size behind either series or how it counted planned changes, and it has not disclosed which countries the willingness survey covered. Both Gartner figures describe 2022, and the planned-changes count in particular is often repeated in 2025 and 2026 as if it were current. Gartner has since taken the original pages down, so our links to them go to archived copies.

The two lines moving in opposite directions do not show that more change caused less willingness. Much else changed at work between 2016 and 2022, and these data cannot separate the effects.

Later evidence comes from a different vantage point. In Prosci's 2023 benchmarking study, 80 percent of change professionals said their organization was near, at or past the point of change saturation. Those respondents were practitioners describing their organizations, not employees describing their own experience, and saturation was self-assessed with no common threshold.

We use 38 percent for 2022 because Gartner published it under its own name, in two places.

What to count

The best benchmark is an organization's own count, repeated over time.

  • List every planned enterprise change that reached employees in the past 12 months, and estimate how many each role experienced. Compare the result with Gartner's 2022 figure of 10 only loosely, because Gartner has not published its counting method.
  • Ask the willingness question in a pulse survey, in Gartner's terms: whether employees are willing to change their work behaviors to support organizational change. Track it each quarter against the change count for the same teams.

Prosci's 2023 study gives one reason to keep the list current: respondents who actively managed their portfolio of change projects met or exceeded objectives more often. A quarterly count of your own would show what has happened since 2022.

Sources

]]> Nine percent: what transformation budgets say about change management https://www.thelangfordinstitute.org/articles/nine-percent-transformation-budgets/ https://www.thelangfordinstitute.org/articles/nine-percent-transformation-budgets/ Sun, 04 Oct 2026 08:00:00 GMT Practice and Methods Monitor Deloitte's 2025 study finds change management and communications get 9 percent of transformation budgets, and a third of leaders would raise it in hindsight. The wider evidence says more about how to spend that money than how much. Change management and communications received an average of 9 percent of transformation budgets in Monitor Deloitte's 2025 Chief Transformation Officer Study, the smallest of six spending areas, against 26 percent for technology. In hindsight, 33 percent of the transformation leaders Deloitte surveyed for the 2025 study said they would increase the change share. More of them, 42 percent in the same study, would raise investment in talent, the other area Deloitte places at the bottom of the budget.

Deloitte's 2025 report is not fully consistent on technology, which appears as 26 percent in one chart and 28 percent in another, but change management and communications is 9 percent in both.

Who answered matters. Monitor Deloitte's 2025 study surveyed 200 transformation executives across industries and geographies. Of those 2025 respondents, more than 90 percent had overseen three or more transformations and more than 80 percent led programs that met all their targets, so the sample skews toward experienced and successful leaders. The budget shares are their own reports of 2024 averages. Deloitte sells transformation consulting.

Set against where executives say transformations run into trouble, 9 percent looks small. The evidence on what change money buys is thinner than that comparison suggests, and it says more about how to spend the money than about how much.

Where the trouble sits

BCG's 2024 survey of 1,000 senior executives found that about 70 percent of the challenges companies face in implementing AI stem from people and process issues, 20 percent from technology and 10 percent from algorithms. The comparison with Deloitte's 2025 figure is loose. BCG's figure covers AI programs only, and people and process is a wider category than change management and communications. BCG's 2024 release does name change management among the people and process capabilities that distinguish AI leaders. BCG also presents its split as a rule for allocating resources that AI leaders follow, and calls it a long-held belief. It is best read as a BCG heuristic backed by survey data, not an independent measurement. The direction is still plain. BCG's executives place most of the difficulty in AI programs with people and process, and Deloitte's 2025 respondents gave the two budget lines most directly about people the lowest allocations.

McKinsey's data on lost value point the same way. In McKinsey's 2021 survey of 1,034 people who had taken part in a transformation, fewer than a third rated it successful at improving and sustaining performance. Respondents to McKinsey's 2021 survey placed nearly a quarter of lost value at target setting and 55 percent during and after implementation, including 20 percent after it. In the same 2021 survey, successful transformations captured on average only 67 percent of their potential financial benefit, against 37 percent for the rest, by respondents' own estimates. Implementation and the months after it are where change management does much of its work. On these estimates, they are also where more than half the lost value leaks.

A leader who met every target with a small change line has some reason to doubt that a bigger one was needed.

The most direct evidence on method is older. IBM's 2008 Making Change Work study of 1,532 practitioners found 52 percent project success where formal change management procedures were always followed, against 36 percent where people improvised. Success in IBM's 2008 study meant fully meeting time, budget and quality objectives, a strict test.

The strongest case against

The strongest counter-evidence comes from Deloitte's own sample. The leaders behind the 9 percent average were, by their own account, mostly successful: more than 80 percent of Deloitte's 2025 respondents led programs that met all targets. Most did not say they would raise the change budget. A leader who met every target with a small change line has some reason to doubt that a bigger one was needed.

Budgets are also growing. Deloitte's 2025 study reports increases of up to 2.5 times in transformation budgets over two years, so a steady share can still mean more money.

The evidence for more spending has limits of its own. Every study cited here is self-reported, and the links they draw between practice and results are correlational. All four come from firms that sell consulting. IBM's figures are from 2008, and they complicate the idea that more is better: in that 2008 study, the two groups between always following formal procedures and improvising reported 39 and 38 percent success, barely above the 36 percent for improvisers. If that pattern holds, partial use of a method buys little, and a larger budget spent on inconsistent practice would buy little too.

Finally, Deloitte's 9 percent is a budget category as its 2025 respondents reported it. It cannot show how much change work is paid for under other headings, such as training funded from a technology line.

What to do on Monday

None of these studies identifies the right share. The practical response is to establish your own.

  • Find your number. Calculate the share of each transformation budget explicitly allocated to change management and communications. Compare it with Deloitte's 2025 average of 9 percent with care: that figure describes a sample of mostly successful leaders and is not a target.
  • Ask the hindsight question early. Deloitte's leaders answered after the fact. Ask your transformation leads now which line they would raise, while the money can still move.
  • Fund the later stages. McKinsey's 2021 respondents placed 55 percent of lost value during and after implementation. Hold part of the change budget for the months after go-live rather than spending it all on launch.
  • Buy consistency before volume. If your organization has a change method, check whether every program applies it. In IBM's 2008 data, the clear gain appeared where formal procedures were always followed.
  • Track adoption against spend. Record what each program's change budget paid for and what adoption followed, so the next allocation rests on your own evidence.

Sources

]]> Go-live is not adoption: what to measure in the six months after cutover https://www.thelangfordinstitute.org/articles/go-live-is-not-adoption/ https://www.thelangfordinstitute.org/articles/go-live-is-not-adoption/ Sun, 04 Oct 2026 08:00:00 GMT Enterprise Technology A vendor-neutral scorecard for ERP and core-system programs that tracks usage, workarounds, process conformance and benefits for six months after cutover, long after hypercare ends. Even the transformations that their own participants rated a success fell well short of what they could have delivered. In McKinsey's 2021 survey of 1,034 people who had taken part in a transformation, respondents at successful programs estimated that their organizations captured on average 67 percent of the potential financial benefit, against 37 percent at the rest. Respondents placed nearly a quarter of the lost value at target setting and about 20 percent after implementation. For an ERP or core-system program, that is the period after cutover, when the program team is being stood down.

The figures are self-reported estimates from an online survey fielded in mid-2021 and weighted by national GDP, and they cover transformations of every kind rather than system programs alone. They do not tell you how much value a given ERP program loses after go-live. They do show that the people closest to these programs saw value slipping away at both ends: when targets were set, and after the work was declared done.

What go-live metrics miss

The dashboards that govern cutover and hypercare, the weeks of intensified support after go-live, track defects, ticket volumes, system availability and transactions posted. Those numbers tell you whether the system is stable. They cannot tell you whether buyers raise purchase orders before invoices arrive, whether planners use the system forecast or keep their own spreadsheet, or whether anyone still owns the business case.

Leaders' own assessments suggest the gap is wide. Gartner reported in 2025 that only 32 percent of mid-to-senior business leaders said the last change they led achieved healthy adoption by employees. Gartner defines healthy adoption as employees acting on a change, on time, without undue stress or harm to performance and engagement. The figure is the leaders' own judgment, it covers changes of all kinds, and Gartner's release does not give the size of the leader sample.

For a core system, adoption means people doing their work in the system as designed, and the business seeing the results the case promised. Both need measuring for months after cutover. The scorecard below sets out what to collect. It contains no target percentages. A sensible threshold for a high-volume accounts payable process differs from one for a monthly planning cycle, so thresholds have to come from your own baseline and your own business case.

A site at full use and a site half in spreadsheets can average out to a reassuring number.

The scorecard and its thresholds

Four measures sit at the core: usage, workarounds, process conformance and benefits. Five supporting measures explain why the core numbers move.

MeasureWhat it tells youHow to collect itWhen
Usage: core transactions completed in the new system, by role and siteWhether work has actually moved into the system; logins alone do not show thisTransaction logs matched to a list of who should perform each taskWeekly to month 3, then monthly
WorkaroundsWhere work is leaving the system: spreadsheets, manual journals, email approvals, legacy lookupsManual-entry and override reports, legacy access logs, floor walks, manager check-insEvery two weeks to month 3, then monthly
Process conformanceWhether cases follow the designed path, such as purchase order before invoiceException reports, or event logs compared with the process designMonthly from month 2
BenefitsWhether the business case is being realizedBenefit owners report against the pre-cutover baseline; finance validatesMonths 3 and 6, then quarterly
Breadth of useWhether users run the full designed process or only the minimumCount of designed transactions used per roleMonthly
Support demand by typeWhether problems are defects, access or know-howTicket categories, with how-to tickets tracked by roleWeekly in hypercare, then monthly
ProficiencyWhether users are getting faster and more accurateCycle times and rework rates per task, from system timestampsMonthly
Data qualityWhether the data the process depends on stays cleanAutomated rules on master and transaction dataMonthly
User confidenceWhether people feel able to do their job in the systemShort pulse survey by roleEnd of months 1, 3 and 6

Process conformance deserves particular care because many benefits depend on a sequence of steps rather than a single transaction. Automated invoice matching pays off only if the purchase order exists before the invoice arrives. System event logs record when each step of a case happened, and comparing them with the designed process shows where work departs from it and how often.

Workarounds seldom appear in ticket queues, because a person who has found a way around the system has no reason to log a ticket. Look for indirect signals instead: rising manual journal entries, requests to keep legacy access, bulk uploads replacing on-screen entry, and approvals given by email and keyed in afterward.

A scorecard is only as useful as the baselines and thresholds behind it. Seven rules keep them honest.

  • Capture the baseline before cutover. For each benefit and operational measure, record the legacy value over a full business cycle, such as a quarter or several month-end closes. Where the legacy system cannot produce a measure, record the gap and take the baseline from the first stable month after cutover, labeled as such.
  • Turn the business case into behaviors. Each benefit assumes people will work differently: fewer manual journals, invoices matched automatically, planners using the system forecast. Make each assumption a measure with a threshold. If the case assumes manual matching disappears, manual matching volume is the number to watch.
  • Agree on the dip in advance. If the plan expects performance to fall after cutover, the process owner and finance should agree before go-live how deep and how long a fall is acceptable, and what happens if it goes further.
  • Set thresholds by role and site. A site at full use and a site half in spreadsheets can average out to a reassuring number. McKinsey's 2021 survey found that goals adapted for employees at all levels were among the actions most predictive of value capture, though that is a correlation in self-reported data.
  • Freeze the definitions. Write down the numerator, denominator, data source and owner for each measure at go-live. A definition changed after disappointing results costs the scorecard its credibility.
  • Compare like periods. Month-end, quarter-end and seasonal peaks move every measure, so compare a close with a close.
  • Attach an action and an owner to every threshold. A breach should trigger something specific: retraining for a role, a design fix, a data cleanup or an escalation to the sponsor.

The six-month sequence

  1. Before cutover. Agree on the scorecard, definitions and owners. Capture baselines and have finance sign off the benefits baseline. Decide what legacy access will remain, make it read-only where possible, and log its use.
  2. Weeks 1 to 4. Track usage by role and site every week. Classify every ticket as defect, access or how-to. Walk the floor and ask managers where work is leaving the system. Run the first confidence pulse in week 4. Hold back benefits reporting, because backlog clearing and data corrections distort early figures.
  3. Month 2. Take the first conformance reading on the main end-to-end processes. Separate deviations caused by design gaps, which need a fix, from those caused by habit or skill, which need coaching.
  4. Month 3. Exit hypercare when the criteria are met rather than when the calendar says so: defect and how-to tickets have stabilized, and usage meets the agreed threshold in each role. Hand each measure to a named process owner. Take the first benefits reading against baseline and run the second pulse.
  5. Months 4 and 5. Retire the workarounds. Switch off remaining legacy access, close shadow spreadsheets with the teams that built them, and retrain roles where conformance lags. Recheck data quality, which can drift once the project team stops watching it.
  6. Month 6. Hold a formal adoption review with the sponsor, process owners and finance, using the third pulse alongside the system data. Review every measure against its baseline and threshold, then decide whether to close, extend support for named roles or sites, or reopen the design. Move the measures that drive the benefits case into business-as-usual reporting, reviewed quarterly until the case is closed.

The first step costs nothing. Open the business case, list the behavior each benefit depends on, and check whether anyone has been asked to measure it once hypercare ends.

Sources

]]> Managers are absorbing the most disruption, and their engagement is falling https://www.thelangfordinstitute.org/articles/managers-absorbing-disruption/ https://www.thelangfordinstitute.org/articles/managers-absorbing-disruption/ Sun, 04 Oct 2026 08:00:00 GMT Operating Models and Workforce In a 2023 survey of US workers, Gallup found leaders and managers 56 percent more likely than individual contributors to report extensive disruptive change. Global manager engagement, measured separately, fell to 22 percent in 2025. Leaders and managers are 56 percent more likely than individual contributors to experience extensive disruptive change in their organization, Gallup reported in 2024. That is a relative difference. It does not mean that 56 percent of managers face extensive change, or that the gap is 56 percentage points.

The same Gallup study adds detail. Seven in 10 US workers reported disruptive change in their organization in the past year, Gallup found in 2024, but that count includes any degree of disruption, and only 20 percent reported change to a large or very large extent. In Gallup's 2024 data, over half of managers said their organization had restructured teams. Gallup's 2024 study also found that employees facing extensive disruptive change were 67 percent more likely to cite an issue with a leader or manager, another relative measure.

Managers' engagement has also fallen. Gallup's State of the Global Workplace 2026 report found that global manager engagement fell from 27 percent in 2024 to 22 percent in 2025. Gallup calls it the largest year-over-year drop in manager engagement, and it left the figure nine points below its 2022 level. According to the same 2026 Gallup report, the fall in manager engagement accounted for most of the decline in overall employee engagement, to 20 percent. Gallup says managers have largely lost the engagement advantage they once held over individual contributors.

HR leaders doubt that managers are ready. In a July 2024 Gartner survey of 473 HR leaders, 74 percent said their managers were not equipped to lead change. Change practitioners single out the middle layer: in Prosci's 2023 benchmarking study, 43 percent of participants named mid-level managers as the group most resistant to change.

How the research was done

Gallup's disruption findings come from web surveys of the Gallup Panel, US adults working full or part time, conducted in the first three quarters of 2023. Gallup ties its headline seven-in-10 figure to the third-quarter survey of 18,665 employees, fielded August 9 to 24, 2023, with a margin of error of plus or minus 1.1 points. The engagement figures in Gallup's 2026 report are global estimates from Gallup World Poll data covering more than 140 countries, and "managers" there includes every level, not only middle managers. Gartner's 74 percent comes from 473 HR leaders, a subset of the 1,403 it surveyed in July 2024. Prosci's 43 percent comes from 2,668 change practitioners who chose to take part in its 2023 study.

What it does not show

The headline pairs two separate Gallup data sets. The disruption finding covers US employees in 2023. The engagement decline is global and covers 2024 and 2025. Gallup's engagement measure is general, not a reaction to change, so these data do not show that disruption caused managers' engagement to fall. Gallup suggests that organizational flattening and cuts to management roles, especially in South Asia, may be one factor. That is Gallup's interpretation, not a measured cause.

The other two figures record views. Gartner's 2024 figure of 74 percent is HR leaders' judgment of their managers. Prosci's 2023 figure of 43 percent records how practitioners see managers, not how managers behave. Managers go through change while leading it for others. What practitioners call resistance may reflect overload or an unclear role.

Read together, the findings are consistent with managers under strain. They do not establish why.

Managers go through change while leading it for others.

What to do with it

  • Give managers a change plan of their own, separate from the plan for their teams. Prosci says much manager resistance can be reduced by addressing managers explicitly in change plans.
  • Ask managers whether they feel equipped to lead the next change. Gartner's 2024 figure is HR's view, and the managers' own view is missing from it.
  • Count the changes each manager is leading and the changes each is going through, and review the totals before approving new initiatives that land on the same layer.
  • Report manager engagement separately from team engagement, by level, so that a fall like the one Gallup found globally would show up in your own data.

Sources

]]> Companies abandoning most AI projects rose from 17 to 42 percent in a year https://www.thelangfordinstitute.org/articles/companies-abandoning-ai-initiatives/ https://www.thelangfordinstitute.org/articles/companies-abandoning-ai-initiatives/ Sun, 04 Oct 2026 08:00:00 GMT AI Adoption S&P Global found that 42 percent of companies had abandoned most of their AI initiatives before production, up from 17 percent a year earlier, in a survey fielded in late 2024. The share of companies abandoning most of their AI initiatives before production rose from 17 percent to 42 percent in a year, S&P Global Market Intelligence's 451 Research reported in 2025. The average organization in the survey scrapped 46 percent of its proofs of concept before production.

The data are older than the 2025 label suggests. The survey was fielded online from October 21 to November 25, 2024, among 1,006 mid-level and senior IT and line-of-business professionals in North America and Europe, and compared with 1,001 respondents surveyed in late 2023. The survey cannot say whether the rise continued after 2024.

The increase is far larger than the survey's margin of error of plus or minus 3 points. But it counts companies that abandoned most of their initiatives, not failed projects, and is often misquoted as 42 percent of AI projects failing. Staff resistance to AI was cited as a challenge by 28 percent of respondents, among the top five obstacles.

A proof of concept stopped because it failed a test set in advance is portfolio discipline.

What other studies add

SourceEvidence basePeriodKey figures
S&P Global (451 Research)Survey of 1,006 IT and business professionals, North America and EuropeOct to Nov 202442% of companies abandoned most AI initiatives before production (17% a year earlier); average organization scrapped 46% of proofs of concept
MIT NANDA (preliminary)300+ public initiatives, 52 organizations interviewed, 153 conference survey responsesJan to Jun 20255% of custom enterprise AI tools reached production
McKinsey State of AISurvey of 1,719 respondents in 97 nationsMay to Jun 2026Nearly three-quarters of high performers redesigned workflows; one-quarter of others did

MIT's Project NANDA reported in July 2025 that 95 percent of the organizations it studied were getting zero return on generative AI, despite an estimated $30 billion to $40 billion in enterprise investment. It found that 5 percent of custom enterprise AI tools reached production.

The report is labeled preliminary findings. Its evidence came from a review of more than 300 public AI initiatives, interviews with representatives of 52 organizations and 153 survey responses from senior leaders at four industry conferences, collected between January and June 2025. That is not a representative sample. The authors caution that the figures are indicative, drawn from interviews rather than official company reporting, and that definitions of success vary. Success meant marked, sustained productivity or P&L impact.

The study is often misquoted as showing that 95 percent of AI pilots fail. Its 5 percent figure refers to custom or vendor-sold enterprise tools reaching production. General-purpose tools such as ChatGPT and Copilot were widely piloted and deployed.

In McKinsey's 2026 State of AI survey, about 6 percent of respondents qualified as high performers, attributing 5 percent or more of EBIT to AI and describing its value as significant. Nearly three-quarters of them reported fundamentally redesigning workflows because of their AI use, against one-quarter of other respondents. That is an association within a small subgroup.

The three studies measure different things in different populations, and their figures should not be combined into a single failure rate.

What to count in your own portfolio

A rising abandonment rate is not necessarily bad news. A company that stops weak proofs of concept early may be showing discipline. The published S&P figures do not separate those cases from projects that simply stalled. A company's own records can.

  • Count the AI proofs of concept stopped before production in the past year, as a share of those started. Compare it with S&P's 46 percent only loosely, since definitions differ.
  • Classify each stop. A proof of concept stopped because it failed a test set in advance is portfolio discipline. One that stalled without a decision is a cost with nothing learned.
  • Record the obstacles behind each stop, including staff resistance, which S&P's respondents placed among the top five.
  • Before approving the next proof of concept, name the workflow it would change and the person accountable for redesigning it. In McKinsey's survey, high performers were far more likely to have redesigned workflows.

Sources

]]> People choices made before day one shape integration value https://www.thelangfordinstitute.org/articles/integration-value-people-choices/ https://www.thelangfordinstitute.org/articles/integration-value-people-choices/ Sun, 04 Oct 2026 08:00:00 GMT M&A and Integration Retention, role design and the first leadership appointments shape what a deal delivers. These are the decisions integration leaders should settle before day one rather than defer. Only 16 percent of respondents in PwC's 2026 M&A Integration Survey said they had assessed leadership and talent before signing or carried out people and organizational diligence. The survey, published on September 29, draws on 530 C-suite executives and senior dealmakers at companies with annual revenue of $1 billion or more. They named clear organization design and decision rights as their leading people-related challenge, cited by 56 percent. The figures are self-reported, and PwC sells deal and integration services. Even so, they imply that most acquirers do not assess leaders and talent before signing, a habit the research below gives reason to avoid.

The most direct evidence concerns the first leadership appointments. In a study of 96 acquisitions that took place between 1980 and 1984, Albert Cannella and Donald Hambrick found that departures of acquired executives harmed post-acquisition performance, with the loss of the most senior, such as chief executives and chairmen, appearing most damaging. Giving one or more acquired executives a place on the combined firm's top management team was associated with better performance (Strategic Management Journal, 1993). The data are old and observational. They still point to a choice made early, in the first round of appointments, that is hard to reverse once the people concerned have left.

What the research shows

Departures after a deal are not random. James Walsh followed the top managers of acquired companies for five years and found turnover significantly higher than normal, with visible, very senior executives leaving sooner than less visible colleagues (Strategic Management Journal, 1988). A later study with John Ellwood found that target top-management turnover was higher than normal in the two years after a deal and was unrelated to how well the targets had performed before it (Strategic Management Journal, 1991).

Why people leave matters as much as how many. Studying 430 executives in 97 acquired firms, Hambrick and Cannella found evidence that relative standing, or local social status, helps explain which acquired executives depart (Academy of Management Journal, 1993). The mechanism reaches below the executive suite. Using patent data on 3,933 inventors at acquired pharmaceutical companies, Srikanth Paruchuri, Atul Nerkar and Hambrick found that integration generally impaired inventors' productivity, most severely for those who lost the most social status and centrality in the combined company (Organization Science, 2006). If a new structure lowers people's status or centrality, the study suggests their productivity may suffer, which makes role design a productivity decision as well as a retention one.

In 228 US bank acquisitions, Maurizio Zollo and Harbir Singh found that replacing the acquired firm's top managers hurt performance, all else being equal, while deeper integration and codified integration know-how improved it (Strategic Management Journal, 2004).

On communication, there is experimental evidence. In a field experiment by David Schweiger and Angelo DeNisi, employees at one plant of a company that had just announced a merger received a "realistic merger preview," a program of realistic communication about the merger, while employees at another plant received limited information. Across four rounds of data collection, the preview reduced the merger's dysfunctional outcomes, and the effects lasted for the duration of the study (Academy of Management Journal, 1991). With only two plants compared, the study shows what is possible rather than what is typical.

Inventors who lost the most status and centrality in the combined company suffered the steepest productivity drops.

The strongest counter-evidence

Most of these studies are archival, many use samples that are decades old, and they show associations. A departure may be a symptom of a deal going badly as much as a cause. Some exits are the point of the deal, where removing duplicate roles is the synergy.

Reviewers also question retention as a default. Jeffrey Krug, Peter Wright and Mark Kroll concluded that the link between turnover and post-acquisition performance is more complex than existing studies imply, argued that there are good theoretical grounds to think replacing executives may be an equally important source of value creation in some acquisitions, and wrote that the decision to retain or replace target executives "is largely a matter of context" (Academy of Management Perspectives, 2014). A meta-analysis of 112 studies by Tsvetomira Bilgili and colleagues found that turnover in the acquired top team and turnover of its chief executive have opposite effects on post-acquisition performance (Journal of Management, 2017). In the authors' accepted manuscript, top-team turnover is associated with worse performance and CEO turnover with better, although the CEO estimate rests on only two studies.

Cultural differences, which a substantial body of research treats as an obstacle to integration, cut both ways. A meta-analysis by Günter Stahl and Andreas Voigt of 46 studies covering 10,710 deals found that cultural differences affect sociocultural integration, synergy realization and shareholder value in different and sometimes opposing ways (Organization Science, 2008).

People decisions are also only one influence among many on how a deal performs. A meta-analysis by David King, Dan Dalton, Catherine Daily and Jeffrey Covin found that, on average, acquiring firms' performance does not improve with their acquisition activity and is modestly hurt, and that variables researchers had not identified may explain significant variance in results (Strategic Management Journal, 2004). Integration choices may be among them; the meta-analysis does not say.

Finally, "before day one" has a legal boundary. Staff of the US Federal Trade Commission wrote in 2018 that until a merger closes, the parties "must continue to operate independently," and that competitively sensitive information needed for integration planning should pass through clean teams and other safeguards. A buyer that effectively takes control of the target before closing risks what US antitrust enforcers call gun jumping. Before closing, integration leaders can plan and decide; the decisions take effect afterward.

What to settle before day one

  • The top team. Name the leaders of the combined organization and decide, deal by deal, whether executives of the acquired company sit on the top team. The evidence leans toward keeping the wider top team; on the acquired chief executive it is thin and divided.
  • The roles the deal depends on. Identify the roles, not just the people, that carry the investment case: the scientists, engineers, client leads and specialists it assumes will stay productive. Settle their scope, reporting lines and decision rights before designing retention packages. A retention payment keeps someone on the payroll; the role decides what they contribute.
  • Retention offers that follow from the role design. Offer them to the people in those roles, with terms tied to the integration milestones that matter, and decide now how you will measure whether they work. In a self-reported 2024 study of 159 respondents by WTW, which advises on retention pay, almost 40 percent of respondents said they did not track retention rates, and only 15 percent tracked retention beyond the end of the retention agreement.
  • A realistic communication plan. Tell employees what has been decided, what has not, and when the rest will be decided, and repeat it as decisions land.
  • A lawful sequence. Run planning through clean teams where sensitive data is involved, settle the decisions before closing, and make them take effect at closing.

Sources

]]> In regulated industries, the change plan can be part of the compliance record https://www.thelangfordinstitute.org/articles/regulated-industries-change-plan-compliance-record/ https://www.thelangfordinstitute.org/articles/regulated-industries-change-plan-compliance-record/ Sun, 04 Oct 2026 08:00:00 GMT Regulated Industries Where regulators require validation, audit trails and training records, much of what a change team produces becomes compliance evidence. A close reading of the rules shows what they demand and where they stop. Europe's guide to good manufacturing practice (GMP) does not stop at telling medicine makers to train their staff. Its chapter on personnel says continuing training should be given, that "its practical effectiveness should be periodically assessed," and that training records should be kept. In the United States, the FDA's electronic records rule, which dates from 1997, requires a determination that people who develop, maintain or use electronic record systems "have the education, training, and experience to perform their assigned tasks."

Set those provisions beside the requirements for validated systems and audit trails, and much of what a change team produces on a regulated system program starts to look like compliance evidence. In this publication's reading, who was trained, on which version of which procedure, before receiving system access, belongs to the quality record. Part 11 makes computer systems, their controls and their "attendant documentation" subject to FDA inspection. None of these rules uses the phrase change management in its organizational sense, so the link is an interpretation, and it has limits.

The vocabulary also holds a trap. In pharmaceutical quality systems, change management already has a defined meaning. ICH Q10, published by the FDA as guidance in 2009, defines it as "a systematic approach to proposing, evaluating, approving, implementing, and reviewing changes" and applies it across a product's lifecycle. It also asks for an evaluation after implementation to confirm that a change achieved its objectives. We see that evaluation as the point where a quality unit's change control and the people side of change meet. In our reading, a new system that staff do not use as designed has not achieved its objectives.

What the life-sciences rules require

In US drug manufacturing, 21 CFR 211.25 requires each person engaged in manufacturing, processing, packing or holding a drug product to have the education, training and experience, or any combination of them, to perform their assigned functions. Training must cover the operations they perform and GMP, including the written procedures that relate to their functions, and GMP training must be given "on a continuing basis and with sufficient frequency" to keep employees familiar with the requirements. The section does not itself mention training records; the EU guide does. Part 11's text adds controls for electronic records, among them validation, secure computer-generated audit trails of operator entries and actions, and revision and change control for systems documentation. Since 2003, however, the FDA has said it will exercise enforcement discretion over those Part 11 validation and audit-trail provisions, relying instead on the underlying predicate rules, while continuing to enforce the training determination.

Annex 11 of the EU GMP guide, on computerized systems, in operation since June 30, 2011, asks that personnel have "appropriate qualifications, level of access and defined responsibilities," that audit trails be considered on the basis of risk and regularly reviewed, and that changes follow a defined procedure. In our reading, its statement of principle reads like a brief for a change team: where a computerized system replaces a manual operation, product quality, process control and quality assurance should not decrease. A revision drafted by the European Medicines Agency's inspectors' working group and PIC/S went to stakeholder consultation from July 7 to October 7, 2025. The draft says everyone involved with such systems should have "adequate system specific training," but the Commission's EudraLex listing still shows the 2011 text as current.

Two guidance documents shape how much validation evidence is enough. ISPE's GAMP 5 Second Edition, an industry guide published in July 2022, emphasizes critical thinking by experienced subject-matter experts in defining appropriate approaches. The FDA's Computer Software Assurance guidance, issued in draft in September 2022, finalized on September 24, 2025 and reissued on February 3, 2026, sets out a risk-based approach for software used in medical device production and quality management systems. Both are nonbinding.

In our reading, a new system that staff do not use as designed has not achieved its objectives.

Finance and AI

The EU's Digital Operational Resilience Act, which has applied since January 17, 2025, requires most financial entities to build ICT security awareness programs and digital operational resilience training into their staff training as compulsory modules, "applicable to all employees and to senior management staff." Members of the management body must keep their knowledge of ICT risk up to date, including through specific training on a regular basis.

Article 4 of the EU AI Act, on AI literacy, has applied since February 2, 2025. As first written, it required providers and deployers to take measures to ensure, to their best extent, a sufficient level of AI literacy among their staff and others operating AI systems on their behalf. Regulation (EU) 2026/1744, in force since July 27, 2026, replaced that with a duty to take measures to support the development of AI literacy, adding that the obligation "does not require providers or deployers to guarantee any specific level of AI literacy of any individual." The Commission's guidance says no certificate is needed and that organizations can keep an internal record of training and other initiatives. Article 99, which sets the Act's maximum fines, does not list Article 4 among the obligations it names; member states set penalties for other infringements.

In US banking, the Federal Reserve's 2011 model risk guidance, SR 11-7, was superseded on April 17, 2026 by SR 26-2, issued with the OCC and FDIC. The revised guidance says that "effective model use depends on a clear understanding of a model's limitations." It also states that it sets no enforceable standards, that non-compliance will not result in supervisory criticism, and that generative and agentic AI models fall outside its scope.

The case against, and what to do

The strongest objection is that none of these texts requires an adoption metric or a change plan. They require qualified people, validated systems, controlled changes and records. A training record shows that a session took place, not that anyone works differently, and few of the rules ask whether training worked. The EU guide's effectiveness clause is the clearest exception; the draft Annex 11 would add evaluation of security training, for example through simulated tests. In our view, the direction of travel also favors proportion over paperwork, although the draft Annex 11 is much longer and more detailed than the 2011 text it would replace. The FDA's software assurance guidance, GAMP 5's emphasis on critical thinking, the narrowing of the AI Act's literacy duty, which the amending regulation justifies partly by the compliance burden on smaller enterprises, and US bank supervisors' disclaimer of enforceable standards all point the same way.

A change team that answers regulation with more documents has misread it. The argument for treating the change plan as part of the compliance record survives in a narrower form: the plan should produce the specific evidence the rules already expect, as a by-product of doing the work well.

  • Map each change deliverable to the requirement it supports: role-based training to Part 11 and, in draft, to the revised Annex 11, staff qualifications to the current Annex 11, effectiveness checks to EU GMP Chapter 2, post-implementation review to ICH Q10, management training to DORA, literacy measures to the AI Act.
  • Grant system access only after training on the current version of the procedure is recorded.
  • Judge training effectiveness with operational data, such as deviations, data corrections and audit-trail findings in the months after go-live, aggregated by role rather than used to score individuals.
  • Keep the change artifacts that serve as evidence in the controlled document system, under version control.
  • Bring quality assurance into the change team before cutover.

Sources

]]> The Chief Transformation Officer moves into the C-suite https://www.thelangfordinstitute.org/articles/chief-transformation-officer-c-suite/ https://www.thelangfordinstitute.org/articles/chief-transformation-officer-c-suite/ Sun, 04 Oct 2026 08:00:00 GMT Leadership and Sponsorship BCG found the number of companies hiring a Chief Transformation Officer rose more than 140 percent between 2019 and 2021. Monitor Deloitte's 2025 data show the role gaining seniority, while change management and communications still get the smallest share of transformation budgets. The number of companies hiring a Chief Transformation Officer rose by more than 140 percent between 2019 and 2021, according to a 2024 BCG study of 350 people in the role. Hiring was led by the consumer, industrial goods and financial institutions sectors. In BCG's usage, CTO means Chief Transformation Officer, not Chief Technology Officer, a frequent confusion when the figure is repeated.

The role has also moved up. In Monitor Deloitte's 2025 Chief Transformation Officer Study, the share of transformation officers at executive or senior management level rose to 95 percent in 2024, from 81 percent in 2022. That band is wider than the C-suite, but the direction is clear.

The budgets they run still favor technology. Deloitte's 2025 respondents gave change management and communications an average of 9 percent of transformation budgets, the smallest of six spending areas, against 26 percent for technology, and 33 percent said they would raise the change share in hindsight.

The rest of the C-suite is changing too. IBM's 2026 CEO Study describes chief executives reshaping senior roles for AI: 76 percent of organizations reported a Chief AI Officer, up from 26 percent a year earlier. In the same 2026 study, 83 percent of CEOs said AI success depends more on people's adoption than on technology, yet they estimated that only 25 percent of their workforce uses AI regularly.

A seat in the C-suite does not settle who owns adoption.

The figures

MeasureFigureSource
Growth in companies hiring a Chief Transformation Officer, 2019 to 2021More than 140%BCG, 2024 (study of 350 CTOs)
Transformation officers at executive or senior management level95% in 2024, from 81% in 2022Monitor Deloitte, 2025 (200 executives)
Share of transformation budget for change management and communications9% (technology: 26%)Monitor Deloitte, 2025
Leaders who would raise the change share in hindsight33%Monitor Deloitte, 2025
Organizations reporting a Chief AI Officer76%, from 26% a year earlierIBM, 2026 (2,000 CEOs)
CEOs who say AI success depends more on people's adoption than on technology83%IBM, 2026
Share of workforce using AI regularly, as estimated by CEOs25%IBM, 2026

What the numbers do not show

BCG's 2024 article gives no absolute counts or sampling frame. Its growth figure starts from a low base and stops in 2021, so it shows the role spreading quickly, not how common it is now. BCG also links Chief Transformation Officer hires to improved shareholder returns, but that link is correlational.

Deloitte's 2025 sample of 200 transformation executives skews toward experienced and successful leaders: more than 90 percent had overseen three or more transformations, and more than 80 percent led programs that met all targets. Its budget shares are self-reported averages for 2024, and the 2025 report puts technology at 26 percent in one chart and 28 percent in another.

IBM surveyed 2,000 CEOs and equivalent senior leaders in 33 geographies and 21 industries between February and April 2026. IBM's 2026 figure of 25 percent is the CEOs' own estimate for their total workforce, not measured usage. All three firms sell services in the areas they study: transformation work at BCG and Deloitte, AI software and consulting at IBM.

What to check

A seat in the C-suite does not settle who owns adoption. For an organization with a Chief Transformation Officer, or one weighing the role, the figures suggest four checks.

  • Name the owner of adoption. Where a transformation office and a Chief AI Officer both exist, write down which of them, or which business leaders, own adoption targets and how adoption will be measured.
  • Compare the change share. Set the share of your transformation budget going to change management and communications against Deloitte's 2025 average of 9 percent, remembering that the figure comes from mostly successful programs.
  • Ask the hindsight question now. A third of Deloitte's 2025 respondents would raise the change share in hindsight. Ask your transformation officer which line they would raise today, while the money can still move.
  • Count usage instead of estimating it. IBM's 2026 figure of 25 percent is a CEO estimate. System data on regular AI use, by team, gives a transformation leader a number to manage.

Sources

]]> Employees do not trust change, and leaders can see adoption lagging https://www.thelangfordinstitute.org/articles/employees-do-not-trust-change/ https://www.thelangfordinstitute.org/articles/employees-do-not-trust-change/ Sun, 04 Oct 2026 08:00:00 GMT Research and Data A 2025 Gartner survey finds 79 percent of employees have low trust in their organization's ability to change. Only 32 percent of leaders report healthy adoption of the last change they led. Seventy-nine percent of employees have low trust in their organization's ability to change, according to an April 2025 Gartner survey of more than 2,850 employees. Gartner calls this "change trust," meaning trust in the organization's ability to change effectively. It is not general trust in the employer.

Leaders report a related problem. In a separate survey reported in July 2025, Gartner found that only 32 percent of mid-to-senior business leaders said the last change they led achieved healthy adoption by employees. Gartner defines healthy change adoption as employees acting on a change, on time, without undue stress or harm to performance and engagement.

The same Gartner release connects trust to how well change is governed. In an April 2025 Gartner survey of 141 HR leaders, organizations facing "ungovernable change" were 1.6 times less likely to have high change trust.

Older and narrower evidence points the same way. In a 2017 survey for the American Psychological Association, US workers going through organizational change were almost three times as likely as those with no change to distrust their employer, at 34 percent against 12 percent. Across all US workers in the same 2017 survey, not only those going through change, 29 percent believed management had a hidden agenda in promoting changes, and only 43 percent were confident that changes would have the desired effects.

How the research was done

Gartner's July 2025 release gives a sample of more than 2,850 employees for the trust survey but no sample size for the leader survey behind the 32 percent, and it does not disclose the country mix.

Gallup's evidence on trust comes from a different angle. Gallup reported in 2024 that among US employees who had experienced significant disruption, those who strongly agreed they trusted their organization's leadership were 4.5 times as likely to be engaged and 62 percent less likely to feel burned out. In the same 2024 Gallup study, disrupted employees who strongly agreed their leaders communicate effectively were 4.3 times as likely to be engaged and 65 percent less likely to feel frequently burned out. Both findings come from Gallup Panel web surveys of US employees conducted in the first three quarters of 2023, and Gallup does not say which of those surveys produced them.

The APA figures come from an online Harris Poll of 1,512 employed US adults, conducted from February 16 to March 8, 2017.

Adoption that is not defined before launch cannot be measured after it.

What it does not show

These studies measure different kinds of trust. Gartner's change trust concerns the organization's ability to change effectively. Gallup measured trust in leadership, and the APA survey measured distrust of the employer. They should not be merged into one claim about trust.

Gartner's 2025 figures of 79 percent and 32 percent come from different surveys of different people, and the 32 percent is leaders' own assessment of the last change they led. Together they describe a gap, but they cannot show that low trust caused weak adoption.

Gallup's 2024 results are correlations and relative measures, so they do not show that building trust would cut burnout by 62 percent. Trust and engagement may also share causes. The APA data are from 2017 and cover only the United States. They show associations, not causes.

Gartner also said in 2025 that routinizing change makes healthy adoption three times more likely. That is a prediction from Gartner's model, not an observed result.

What to do with it

Adoption that is not defined before launch cannot be measured after it. Gartner's definition gives leaders a usable test: whether people act on the change, on time, without undue stress or harm to performance and engagement.

  • Write that test into the plan before launch, with a date for checking it and an owner for the result.
  • Measure change trust on its own. Ask employees whether they trust the organization to carry out change well, and keep that question apart from engagement and general trust items.
  • Test communication in teams under heavy change. Gallup's 2024 finding is a correlation, but a single survey item on whether leaders communicate effectively can show whether the pattern holds in your own organization.
  • Keep one register of all planned changes, so someone can see when the load is becoming hard to govern. Gartner's 2025 finding on "ungovernable change" rests on 141 HR leaders, which makes it a signal to watch rather than a benchmark.

Sources

]]> Frontline AI users save a workday a week. Two-thirds get little guidance on using it. https://www.thelangfordinstitute.org/articles/frontline-ai-users-save-a-workday/ https://www.thelangfordinstitute.org/articles/frontline-ai-users-save-a-workday/ Sun, 04 Oct 2026 08:00:00 GMT AI Adoption BCG's 2026 AI at Work survey finds 42 percent of regular frontline users, non-managerial office staff in its definition, save at least a full workday a week. Sixty-six percent say they get limited or no guidance on what to do with that time. BCG's 2026 AI at Work survey found that 42 percent of frontline employees who use AI regularly save at least a full workday a week. In the same 2026 survey, 66 percent said they get limited or no guidance on what to do with that time, and over half do not redirect it to strategic work.

Frontline has a narrow meaning here. BCG defines frontline employees as individual white-collar employees without managerial responsibilities, not deskless or shop-floor workers. Among that group, BCG's 2026 survey puts regular AI use at 74 percent. BCG's release gives two inconsistent accounts of how fast that share has grown, one over a year and one over two years, so the size of the increase is left out here.

AI agents show a related gap. In BCG's 2026 survey, 30 percent said AI agents are already integrated into workflows, up from 13 percent, while 52 percent have a limited understanding of what agents are.

How the research was done

BCG surveyed 11,749 workers across 14 markets for its fourth annual AI at Work report, released on June 3, 2026. The time savings are respondents' own estimates, not measured hours. BCG's 2026 release does not state the base for the 66 percent figure; it appears to be regular frontline users. The full methodology sits in BCG's report rather than the release. BCG sells transformation services.

Many employees are saving time with little direction on how to use it.

The same 2026 research compares groups of respondents. BCG associates a clear AI strategy with a 25 percentage point lift in measurable business impact, against about 5 points for better tools alone. In BCG's 2026 data, respondents at companies pursuing workflow redesign were 24 points more likely to see measurable improvement and 22 points more likely to save a full day a week. In the same 2026 comparison, they were also 20 points more likely to report higher job satisfaction.

Chief executives point the same way. In IBM's 2026 CEO Study, 83 percent of CEOs said AI success depends more on people's adoption than on technology, and the CEOs estimated that only 25 percent of their workforce uses AI regularly. IBM's 2026 estimate and BCG's 74 percent are not comparable: IBM's is a CEO estimate for the whole workforce, while BCG's covers white-collar staff without managerial duties. IBM sells AI software and consulting.

What it does not show

Nothing in these figures shows what the saved time produces. Hours saved are self-estimated, and business impact is as perceived by employee respondents, not audited results. The lifts for strategy and redesign are BCG's comparisons across respondent groups, not controlled experiments, so companies with a clear strategy may differ in other ways that also matter.

Nor does the survey describe deskless work. Leaders of shop-floor and other deskless workforces should not read these results as a picture of their people.

What to do with it

Many employees are saving time with little direction on how to use it. Deciding where that time goes is a management task, and it can start this week.

  • Decide what the time is for. Ask each manager to agree with their team where reclaimed hours should go, and write it down.
  • Measure the hours. BCG's 2026 figures are self-estimates. Time a few common tasks before and after AI use, and replace estimates with counts.
  • Redesign before buying. In BCG's 2026 comparisons, workflow redesign was associated with a 24-point lift and better tools alone with about 5 points. Fund the redesign of a few high-volume workflows before adding licenses.
  • Spend some of the day on skills. CEOs in IBM's 2026 study expect 29 percent of employees to need reskilling for a different role and 53 percent to need upskilling between 2026 and 2028. Reclaimed time is a natural place to schedule it.
  • Ask your own users. Survey regular AI users each quarter on time saved and on whether they know what it is for. Track the share with clear guidance as closely as the share using the tools.

Sources

]]>