Leadership development has no shortage of new ideas. Purpose-driven leadership. Agile leadership. Human-centered leadership. Now, artificial intelligence (AI)-augmented leadership. Each framework aims to be the model that sticks, but all too often another approach is right around the corner.

These frameworks trend in and out for good reason: Leaders, new and experienced, are under real pressure to keep up. Studies show that roughly 60% of organizations name leadership skill gaps as their top internal challenge, and more than 70% of employees say they’d stay longer at companies that invest in their growth. It’s little wonder the leadership development market keeps growing.

Years of building these programs have taught me that knowledge isn’t scarce, especially in the age of AI. Knowing what to do in a moment that’s ambiguous, emotionally charged or high stakes is not the same as being able to do it. Every decision a leader faces involves some combination of strategic tradeoffs, resource constraints and interpersonal tension, filtered through the mindsets that leader already holds.

A leader can know the textbook steps for leading change and still fail to apply them, if they believe change is something to push people through rather than lead people in. That’s why leadership development should start instead with the moments leaders will need to navigate.

Turning Leadership Knowledge Into Action

The most useful starting point for leadership development is the behavior the business needs from its leaders. What must they do differently for the strategy to succeed? Answering that means starting with leaders’ actual context — their pressures and tensions today, and where the strategy needs them to go next, rather than with theory. Leaders are performing constantly; they aren’t sitting back to study.

I saw this play out with a client at the intersection of retail, manufacturing and distribution. Their leaders don’t have one-on-ones or sit behind desks; most of their toughest moments happen in public, in front of customers and crews. So instead of starting with a framework, we mapped the situations they routinely faced: a shipment delay, a team member calling in sick, a hostile customer, the daily stand-up. Then we asked, in an ideal future state, what would be different? How would people show up? What would get discussed and how would outcomes change? That allowed us to simulate both the current state and the future state, so leaders could feel the disconnect and practice closing it.

Why Practice Changes What Frameworks Can’t

Moments-based simulations, a form of experiential learning, let leaders run a realistic version of a high-stakes moment in a risk-free environment.

The real power of this technique is not that it’s engaging, but that it lets leaders experience consequences, make tradeoffs, get feedback and try again before the real moment arrives. It’s one of the most effective tools for changing mindsets, because mindsets aren’t something you can instruct someone to change. They’re a choice, active or passive, built from experience.

If you want someone to hold a different belief, you have to give them a different experience that exposes them to it.

We recently used this strategy with a client whose leaders had to address a real and rising fear; employees worried that AI would make them obsolete. We put those leaders through an AI-driven role-play of that exact conversation and gave them feedback on how well they held two things at once — real empathy for the concern, and honesty about what was changing.

This exercise helped leaders learn not to promise that it would be fine when it might not be. They acknowledged what they knew and admitted what they didn’t. And they invited the other person into shaping a future where they were still relevant, doing new and valuable work only a person could do.

From Proving ROI to Predicting the Future

Too many learning and talent functions focus heavily on proving the return on investment (ROI) of their programs. That measurement matters, but it doesn’t tell the whole story. L&D teams also need ways to understand whether leaders are applying what they’ve learned in the moments that matter. Practice-based approaches, including AI role-play and simulation, can provide a window into how leaders respond to realistic, tension-filled situations and where they may need additional development.

The most strategic learning leaders I know use that data, both qualitative and quantitative, to build predictive insight into performance and culture, then make strategic decisions about talent.

When development is designed to be future-focused and built for that kind of data, it becomes a real-time view of the organization’s leadership capacity, and a clear indicator of where that capacity needs to grow next.

Leadership development shouldn’t be judged by whether people enjoyed it or scored well afterward. It should be judged by whether it changes what leaders can do in the moments that matter most, and by what it reveals about where the organization needs to go next.