Stop Treating Emergencies as a Delivery Plan

Emergency overtime should be like antibiotics: rare, targeted, and followed by recovery. When a 9-9-6 sprint becomes the answer to every planning miss, the shortcut calcifies into culture and talent leaves.

How Crisis Hours Become Default

Repetition quickly converts a "one-time" request into tradition, effectively setting a burnout template for every new hire. As this norm solidifies, reciprocity pressure mounts, when one person stays late, others feel an unspoken obligation to follow suit. Consequently, roadmaps begin to drift because this hidden overtime masks the true scope and dependency risks. Ultimately, the most employable people exit first, taking their predictability and institutional knowledge with them.

Planning Is the Real Job

Planning means sequencing uncertainty into staged bets and saying no when the plan cannot fit. Teams dodge this work because it feels slow or political, then reach for overtime to plug the gap. Better planning looks like:

Better planning involves tackling risky dependencies early and sizing scope into thin, releasable slices. It requires forecasting with historical throughput rather than optimism, and reserving genuine time for integration and unknowns, not just feature coding.

Why PMs Miss the Estimate

Many PMs lack hands-on shipping reps. Without intuition for testing and integration time, optimism fills the gaps - especially when incentives reward aggressive dates. Calibrate plans to the team you have, not the team you wish you still had.

Define Emergencies in Writing

Treat sustainable pace as a nonfunctional requirement. If a plan only works with overtime, change the plan. Define true emergencies narrowly:

True emergencies should be defined narrowly: a production outage with revenue impact, a P1 security issue with an exploit in the wild, or a regulatory deadline with genuine legal exposure.

Internal demos, aspirational OKRs, or uncontracted sales promises do not qualify.

Plan for Sustainability, Not Heroics

Commit to outcomes and thin slices, not heroic hours. Protect your integration buffers, and increase them if you find yourself constantly using them. Map dependencies with upstream and downstream owners before committing, and keep PRs small, ship behind flags, and swarm on blockers.

Define emergencies tightly, plan with data, and honor recovery. You will ship more, lose less sleep, and keep your best people longer.