Model the Transformation You Expect Employees to Deliver

In so many transformations I’ve advised, the same scene plays out. The CEO opens the all-hands meeting with conviction: “We are going to become more agile, more digital, more customer-obsessed.” The town hall ends, the slides go onto the intranet, and within a week, the senior team is back in the same operating rhythm as before: the same monthly reviews, the same dashboards, the same calendar dominated by the running of the existing business. The transformation lives in the deck—it does not live in the leaders’ personal calendars.

The truism “employees resist change” is a comforting story leaders tell themselves when transformation efforts fail, but evidence points to a different cause. McKinsey’s research on more than 1,000 transformations finds that when senior leaders visibly model the behaviors they ask of employees, transformations are 5.3 times more likely to succeed—yet only 43% of leaders actually do so. Prosci’s benchmarking research, drawn from more than 25 years and 12 studies, points the same way: Active and visible executive sponsorship has been the single biggest contributor to change success in every edition since 1998, and projects with an extremely effective sponsor meet or exceed their objectives 79% of the time. For those with an extremely ineffective sponsor the number plummets to 27%. The most-cited obstacle to success is the absence of that sponsorship, not resistance from the workforce. The workforce is not, by and large, the bottleneck. The bottleneck is at the top.

This is a central argument of my book Powered by Projects: Leading Your Organization in the Transformation Age. We have entered a project economy in which the work that creates value is increasingly project work—change the business, rather than run a repeatable business—and the project-driven organization is replacing the operational machine as the dominant form. That structural shift will not deliver on its promise unless the senior team is willing to operate differently: They must model the transformation for their people.

What “Modeling the Transformation” Looks Like

Most transformations come with behavioral asks that are remarkably consistent across industries: Take more risks. Run experiments. Learn fast and change course when the evidence shifts. The leadership team typically endorses all of these.

But then, too often, leaders go back to running the company through operating forums built for a different era: quarterly business reviews, capital approvals committees, the annual planning cycle, and the performance management system. And each of those forums punishes exactly the behaviors the transformation requires.

The contradiction is not subtle, and employees register it immediately. They are being asked, in good faith, to do one thing while the systems surrounding them reward something else. Over the long run, the transformation quietly dies in the gap between the two; not because anyone sabotaged it, but because the operating system kept doing exactly what it was designed to do.

In contrast, the most successful project-driven organizations I work with share a feature: The CEO is the visible sponsor of the most ambiguous initiative in the portfolio—not the safest one. And when the rest of the executive team can see the CEO operating in that mode, they do the same.

When Piyush Gupta took over DBS Bank in Singapore in 2009, he inherited an institution that customers nicknamed “Damn, Bloody, Slow.” Over the following 15 years, he led one of the most studied transformations in financial services; Innosight eventually ranked DBS among the ten most transformative organizations of the 2010s. The strategy was familiar in outline: digitize the bank, mandate experimentation, build a startup culture inside a 27,000-person institution. What made it stick was how Gupta personally showed up.

He visited branches incognito, sitting in queues as a customer, to understand for himself why people found the bank so frustrating. He made running experiments a KPI for senior leaders, not a permission granted to junior teams; the bank ran 1,000 experiments in a single year, with most of the leadership participating. He opened an anonymous email channel called “Tell Piyush” two weeks each quarter, then spent months working through the suggestions, about a third of which led to policy changes. When DBS pushed into AI, Gupta entered the bank’s AWS DeepRacer League himself, an autonomous-car simulator used to teach machine learning, and, by his own account, finished “in the top 100 among our people.” A small gesture and an unmistakable one. He was telling 27,000 employees that he was willing to compete on a learning curve, even though he was visibly not the best person in the room.

Gupta is only one example. I have seen the same pattern in less famous places, including in transformations I have advised. The CEO of a European energy company who personally chaired the group’s most uncertain diversification bet—and deliberately let the safer programs run without him. The CFO of a global pharmaceutical company who moved her strongest controller onto an unproven AI initiative, signaling with people rather than memos where the priority sat. The CEO of a retail bank who joined the working sessions of a digital venture nobody was sure would survive, and stayed in the room when the first results came back mixed.

When Leaders Fail to Model

The failure to model the transformation is rarely deliberate, and it often arises from a series of small, defensible-looking decisions. It might look like a leader who has championed agile delivery but still requires a full business case before approving a six-week experiment. A CEO who has called for risk-taking quietly hands the most ambiguous initiative to the leader most likely to play it safe. A head of function who has signed the transformation charter declines to release any of their best people onto it, because the day job cannot afford to lose them (the same day job that the transformation was meant to disrupt).

None of these decisions feels like sabotage to the person making it. Each one, in isolation, looks like prudent management. But the cumulative effect is that the transformation gets staffed by part-time volunteers, governed by the same forums that already exist, and judged by the same metrics it was meant to replace. Employees throughout the organization correctly concludes that nothing fundamental has changed at the top.

Look at the appointment of António Horta-Osório as chairman of Credit Suisse in April 2021. The bank had recently endured the collapses of Greensill Capital and Archegos Capital—the Archegos default alone cost the bank about $5.5 billion—laying bare what the chair of Switzerland’s financial regulator called “a cultural problem that translated into a lack of accountability.” The board recruited Horta-Osório, who was credited with turning around Lloyds Banking Group, specifically to repair Credit Suisse’s culture. He framed the new direction in his own words: “We are committed to developing a culture of personal responsibility and accountability, where employees are, at heart, risk managers.”

Nine months later, he resigned. An internal investigation had concluded that he had personally breached Covid quarantine rules twice, including a trip to the Wimbledon tennis finals, and his use of the bank’s private jets had become a focus of the inquiry. The cultural transformation Horta-Osório had been hired to lead never recovered momentum. Within fifteen months of his resignation, Credit Suisse no longer existed as an independent institution—in March 2023 it was forced into an emergency, state-brokered takeover by UBS.

While this Credit Suisse story shows a series of governance and risk failures, at its heart is Horta-Osório’s failure to embody and model the transformation the organization needed. The chairman, explicitly tasked with fostering a culture of personal accountability, was unable to live up to the standard he demanded of everyone else. In a transformation, the behaviors that travel are the ones leaders demonstrate, not the ones they articulate.

Are You Really Modeling the Transformation?

In my work with executive teams, I have started using a new opener. Rather than “What is your transformation strategy?” I lead with, “In what visible way are you, personally, doing the thing you are asking everyone else to do?” Upon asking this, I am usually met with silence—to break it, I test them with three questions. Each is a different way of determining whether the senior team is really living inside the transformation.

The calendar test. In any given month, how much of the senior team’s time is spent on transformation rather than on running the existing business? In the teams I work with, the share rarely exceeds 10% on first measurement, even when the rhetoric suggests it should be closer to half. And within that time, is the CEO personally sponsoring a bet whose outcome is genuinely unknown, or only the initiatives that are sure bets? To truly model the transformation, the CEO’s name should be on the riskiest bet and their time should be devoted to it.

The uncertainty test. Does the senior team have any shared experience of operating under genuine uncertainty together? Most executives built their careers on mastering businesses they already understood; transformation rewards the opposite skill. A team that has never sat together inside a live, unresolved bet tends to manage the transformation from the outside rather than from within, reviewing and governing it but never inhabiting it. For a leadership team to understand what they are asking of their people, they need to do it themselves first. You can’t delegate this, and you can’t simulate it in a workshop. They must sit together on an unresolved bet, make hard calls before the data is in, and own the result either way.

The cancellation test. When was the last time the senior team publicly killed an initiative it had previously championed? Stopping a project—changing your mind when it actually costs something—is the most visible act of learning a leadership team can perform. Healthy portfolios retire two or three bets a year, with lessons learned well-understood and stated; unhealthy portfolios keep everything from years ago alive, starved but never finished. If nothing has been retired in years, that’s the place to begin. The leadership team should kill something it once championed, and say out loud what they have learned from it. A cancellation in the open proves that changing your mind costs nothing when the evidence changes—which is what you’ve been asking everyone else to do all along.

The three tests are really one question: is the senior team inside the transformation, or watching from outside? Fail them, and you’ve already told your people that the work of change is for everyone but you.

. . .

A large industrial company I had been advising spent the better part of two years “in transformation” with little to show for it: The strategy was sound, and the program office was busy, but the executive team treated the effort as something to review once a month rather than something to lead.

Then a new CEO arrived. He took personal ownership of the two hardest initiatives, moving them to the top of his own calendar, and showing up in the working sessions where the real trade-offs were being made. The rest of the team followed his example within weeks, and decisions that had been stuck for months began to clear.

DBS succeeded—and Credit Suisse failed—based on the uncomfortable truth that leaders cannot delegate a transformation. You can sponsor it, fund it, communicate it, govern it, but if you are not visibly living the change you are asking for, the organization will eventually stop pretending to do so.

Copyright 2026 Harvard Business School Publishing Corporation. Distributed by The New York Times Syndicate.

Explore AAPL Membership benefits.