I Stopped Playing Firefighter And Taught My Team To Fight Fires Themselves

I swapped quick fixes for three guiding questions and watched a dependent group grow into a self-reliant one — a small change that multiplied my leadership reach and made the organization more resilient.

Early on, I treated problem-solving like my badge of honor. When systems flared up or deadlines tightened, I stepped in and closed the gap. At first it felt efficient and, frankly, necessary. But it also sent a message: I was the default solver. The moment I stopped being the instant answer, things shifted. People started to own issues, offer solutions and take the kinds of risks that build capability. That single shift — from doing to asking — had a bigger payoff than any quick fix I’d ever pulled off.

The lesson hit home after a critical outage during a product launch. I ran a quick extremes-first diagnostic and patched the system in minutes. Later, a senior engineer told me, without heat, that she would have found the root cause if given the chance. Her disappointment was subtle but real. In the weeks that followed she stopped volunteering for tricky tasks and checked out during problem brainstorms. By rescuing the situation I’d also rescued her from learning. I realized I’d been depriving the team of practice and eroding their willingness to try.

I saw the same pattern across organizations. When I bridged gaps between design and data teams to accelerate a rollout, things moved faster with me in the middle. But when I was unavailable, momentum stalled. Making yourself the go-to person creates short-term gains and long-term fragility. Organizational psychologists I spoke with say leaders who habitually take over work end up limiting decision-making capacity across the group. Put another way, every time you solve rather than guide, you add one more task the team won’t learn to do without you.

So I adopted a discipline I now call the three-question practice. First I ask what the person has already tried. Next I ask what assumptions underlie their plan. Finally I ask what the smallest sensible next step would be. Usually by the third question the person sketches a clear course. And when they do, they treat the outcome as their responsibility. I also use a simple litmus test before intervening: if doing nothing leads to a minor delay or an easy recovery, I let it play out. If the gap is a true skill shortage, I teach and model. If it’s a confidence gap, stepping in only reinforces doubt.

The transition wasn’t tidy. Some teammates worried I’d withdrawn support. Others felt abandoned. That uncomfortable stretch was part of the work. Over time, lateral problem-solving increased; people began bringing well-formed proposals instead of loose questions. Meetings got shorter and decisions moved faster. And my workload shifted from firefighting to coaching, which scaled my impact far more than any single solved incident ever had.

As automated tools get better at handling routine issues, the temptation will be to hoard insight or lean on software for every answer. But machines don’t learn judgment or build professional courage. The useful work of leadership, it turns out, is less about rescuing and more about creating conditions where people learn to rescue themselves. It takes restraint and the occasional uncomfortable pause, but the payoff is a team that can actually handle the next crisis without you.