Decisions with Incomplete Information

Show that you can make a sound call when the data is missing: gather what you can fast, state your assumptions, keep the decision reversible and own the outcome.

Overview

Engineering decisions rarely wait for perfect data. A launch date, an outage or a schema freeze forces a choice before every question is answered. Interviewers ask this question to learn two things: do you freeze until you feel certain, which never happens, and when you do act, is it a calculated risk or a guess?

A strong answer shows a process. You set a time limit on research, gathered the facts you could get quickly, wrote down what you were assuming, chose the option that was cheapest to undo, decided in advance what would make you change course, and then owned the result. A weak answer describes a gut call that happened to work.

Some companies say this out loud. Amazon's Bias for Action principle says that many decisions are reversible, do not need extensive study, and that it values calculated risk taking. Jeff Bezos's 2016 shareholder letter calls reversible decisions 'two-way doors' that can use a light process, and suggests most decisions should be made with about 70% of the information you wish you had.

  • Speed — Research was time-boxed — Waited for certainty, or did no research
  • Assumptions — Written down so others could check them — Hidden or unknown
  • Reversibility — Chose an easy-to-undo path when possible — Bet everything at once
  • Guardrails — Rollback trigger set before acting — Decided later whether it was working
  • Ownership — Owned the result, good or bad — Blamed the missing data

What you will be able to do

  • Tell a decision story that shows a process, not a lucky guess.
  • Tell a reversible decision from one that is hard to undo, and match your speed to it.
  • Describe the guardrails you set: flags, staged rollouts and rollback triggers.
  • Talk about a wrong call calmly, and show what you changed afterward.

Interview questions

  • Tell me about a time you had to make a decision without all the information.
  • How do you reduce risk when you have to decide without enough data?
  • Tell me about a time a decision you made with limited data turned out to be wrong.

Sign up free to read the full topic

More Behavioral topics

Start free on Cariara · Coding problems · System design studio