Mistakes you shouldn't let your reports make

Leaders often hear that they should let reports make mistakes to build trust and encourage risk-taking. The article flips that narrative, arguing that while some mistakes are useful, certain high-cost errors must be caught early. Examples include starting a project that will never ship, chasing pet projects for personal recognition, committing to a solution that misfits the long-term strategy, over-emphasizing abstract benefits at the expense of shipping, and diving into system changes without considering downstream impact. Each of these wastes time and drags the organization away from its goals.

The first guardrail is manager engagement. When managers stay involved enough to spot emerging risks-through regular 1:1s, code reviews, and forward-looking planning-they can halt doomed initiatives before momentum makes them hard to stop. The author admits to spending too much time on immediate status updates and not enough on upcoming work, which led to late-stage interventions. The second guardrail is crystal-clear company and team goals. If every engineer understands how their work ties to the broader mission, they can self-correct and avoid large missteps. Frequent all-hands or better communication of strategy are practical ways to reinforce this alignment.

Even with strong processes, some engineers will drift toward shiny-new ideas or over-engineer solutions with minimal practical value. The piece recommends coaching these individuals rather than stifling curiosity, striking a balance between exploration and impact. When mistakes do occur, leaders should debrief to ensure the team grasps the impact and the path to correction, while reaffirming trust. Ultimately, the article shows that preventing big mistakes is a core leadership responsibility that protects team momentum and aligns effort with business outcomes.