PM Engineering Estimates
(2026 Edition)
5 questions to ask about estimates, 5 things not to do, 5 moves when things slip, and 5 ways to build estimation trust.
Build PM Execution Skills Daily — Free →5 Questions to Ask About Estimates
What are the 3 biggest unknowns that could make this longer?
What's the smallest version you could ship in half the time?
What's blocking you — info from me, design, dependencies elsewhere?
If we had to ship in X weeks, what would we cut?
Where's the buffer in this estimate — how much risk are you factoring in?
5 Things NOT to Do
Pressure for a shorter estimate without cutting scope — erodes trust
Compare against other teams' estimates — context is different
Reopen estimates mid-sprint — undermines commitment culture
Use estimates as deadlines publicly — private estimates, public commitments
Estimate engineering work yourself — you're not qualified
5 Moves When Estimates Slip
Surface slip early — 'we're on track to be 1 week late' at week 1, not week 4
Separate 'unexpected' slip from 'bad estimate' — different responses
Work with eng to identify what can be cut to meet timeline (vs extending)
Communicate to stakeholders before they ask — proactive beats reactive
Post-mortem every major slip — extract learnings for next estimate cycle
5 Ways to Build Estimation Trust
Respect their process — if they estimate 6 weeks, plan for 6 weeks
Add buffer in your plans, not in asking for more — 20% buffer in PM plans
Celebrate predictability as much as speed — 'shipped on time' is a wins story
Take the blame when scope expands mid-flight — you didn't scope enough upfront
Never ask for unrealistic timelines to look ambitious — backfires within 1 quarter
FAQ
How should PMs react when engineering says '6 weeks' for something they thought was 2 weeks?
Ask, don't argue. 'Help me understand what's making this 6 weeks — where's the complexity?' Often you'll learn something that changes scope. Sometimes you'll validate the estimate. Either way, probing through questions > pushing back on the number. PMs who always argue estimates lose engineering trust fast.
Should PMs learn to estimate engineering work themselves?
Conceptually yes (helps with scope decisions), authoritatively no. Develop intuition for 'this feels like a 2-week vs 2-month thing' but never overrule engineering on actual estimates. The failure mode: PMs who've coded a bit thinking their estimate > senior engineer's. Always respect the specialist.
Build PM Execution Muscle Daily
Daily scenarios on scoping, trade-offs, and working with engineering.
Start Free Trial →