In March, I hosted a wine tasting for a group of senior leaders from across industries—banking, manufacturing, engineering, and tech—and we had a candid conversation about the real challenges they’re navigating during this time of AI transformation. Not the polished version. The honest version.
What came out of that conversation was practical, grounded, and immediately applicable. Here’s what’s working.
___
1. Use clarity to inspire
When people feel anxious, they don’t need a rally. They need to know where they stand and what’s expected of them.
Every leader in the room came back to some version of goal clarity as the foundation of managing through change. When business conditions shift, teams instinctively return to old habits—not because they’re resistant, but because the familiar provides an anchor when everything else feels uncertain. The leader’s job is to give them a better anchor: clear objectives, a visible line of sight from individual work to organizational goals, and enough communication that people aren’t filling the silence with worry.
This isn’t about over-managing. It’s about removing the ambiguity that causes people to contract when the business needs them to stretch.
2. Build a 90-day cadence into your goal-setting
Annual goals made sense when annual planning was adequate. Right now, in the world of AI, business conditions are moving faster than that.
Several leaders described a model that’s working for them: set direction annually, but build in structured 90-day reviews—individual and team—to reassess priorities, identify what’s no longer worth chasing, and reallocate energy to what matters most right now.
The key insight from one leader in financial services: AI tools have dramatically expanded what’s possible, which means the volume of good ideas has also expanded. Without a regular forcing function to prioritize, teams end up diffuse—spreading effort across too many bets, moving fast on everything, making real progress on nothing. The 90-day sprint structure creates the discipline to focus.
3. Start with the smallest possible change
One leader from the civil engineering sector shared a principle I haven’t stopped thinking about: when a big change needs to happen, don’t lead with the big change.
Find the smallest change you can credibly ask the team to make. Get them to do it. Make a visible deal out of it—recognize it publicly, share it broadly. Then build toward larger changes once the team has evidence that they’re capable of adapting.
This approach isn’t timid. It’s psychologically sophisticated. People don’t resist change because they’re lazy. They resist it because they’re not confident it will work, or that they’ll be able to succeed at it. Early wins build the confidence that sustains harder work later.
4. Create space for AI experimentation—and for failure
The leaders making real progress on AI adoption aren’t relying on top-down mandates. They’re creating low-stakes environments to practice.
Hackathons, sandbox time blocked on team calendars, small exercises where the goal is exploration rather than output—these are the mechanisms driving genuine adoption. Not because they produce polished work, but because they produce familiarity. And with familiarity comes the reduction of fear that’s the real barrier to adoption for most teams.
One leader described building a culture where AI failures are surfaced, shared, and laughed about. When an AI model makes an obviously wrong call—and they do, often—it’s a teaching moment. It humanizes the technology, demonstrates that it requires human judgment to be useful, and gives people permission to engage without feeling like they’re supposed to already know how it works.
At ExecOnline, we did something similar during a company offsite: each small group used AI to build a version of how we should update our website based on our strategy, with someone embedded on each team who was comfortable with the tools. Nobody shipped a website. But everyone left with a hands-on sense of what AI can actually do—and the mysticism came down considerably across the organization.
What these examples share is a design principle: the experience is structured for the team, not just the individual. One person completing a course doesn’t move the team. A team experimenting together—failing together, problem-solving together—builds something more durable: shared language, shared confidence, and a shared baseline for what comes next. That gap, between individual learning and team-level change, is the most underinvested space in most organizations’ AI strategies.
5. Make the career case for building AI skills now
The most honest moment in the conversation came when someone named the elephant in the room: AI will change jobs. It already is.
According to Gallup, 18% of U.S. employees say it is very or somewhat likely their job will be eliminated within five years—a figure that rises to 23% among employees working in organizations that have already adopted AI. The leaders managing this well aren’t pretending otherwise. They’re making a different case—one grounded in the individual’s long-term interest. The skills being built right now, learning to work alongside AI, developing the judgment to prompt and guide it, are portable. They travel with the person, not the employer. And the window to build them from the ground floor is right now.
I’ve been using an analogy from my own career for this. When I was a law associate, some partners refused to adopt email. They had assistants print their messages, wrote responses by hand, and had the responses typed back up. They weren’t bad lawyers. But they opted out of the infrastructure of professional life—and that had costs.
AI isn’t email. It’s more disruptive and more consequential. But the dynamic is the same. The leaders helping their teams engage now—before the path is obvious, before adoption is mandatory—are building organizational capacity that compounds over time.
———
The common thread across all of these practices is that they treat people as adults navigating genuine uncertainty—not as obstacles to be managed or problems to be solved. That orientation matters. It’s what makes the difference between leaders who get “check-the-box” compliance and leaders who build teams that actually move through hard things.
That conversation in March clarified something I’ve been sitting with since: the gap between individual learning and team-level execution isn’t a culture problem or a motivation problem. It’s a structural one. Organizations invest heavily in developing individual leaders while leaving the team—the actual unit of execution—largely unsupported. It’s why we recently acquired Teamraderie, whose model of expert-facilitated team development is built to close exactly that gap: turning individual learning into changed workflows, stronger execution, and measurable outcomes at the team level.
None of this is simple. But the leaders doing it aren’t waiting until conditions are favorable. They’re building the systems, rhythms, and cultures that make change something teams can handle—and eventually, something they can lead.