Writing Great Questions
The best interview questions don't test whether a candidate can answer your specific scenario. They invite the candidate to show you the best version of themselves — and then you measure that.
The goal of a great question
Here's the counterintuitive truth about behavioral interview questions: you want the candidate to succeed.
Not because you're being generous — but because the best signal comes from a candidate performing at their peak. If your question is so narrow or obscure that even a strong candidate struggles to find a relevant example, you've wasted your interview time collecting noise.
The best interview questions let candidates talk about what they're best at. Then you evaluate the ceiling, not the floor.
What makes questions fail
| Problem | Example | Why it fails |
|---|---|---|
| Too specific | "Tell me about a time you migrated a PostgreSQL database to DynamoDB under regulatory constraints" | Most candidates won't have this exact experience. You're testing memory, not capability. |
| Too vague | "Tell me about a challenge you faced" | So broad that the candidate wastes time picking a story instead of telling one. The answer could go anywhere. |
| Leading | "Tell me about a time you showed great leadership" | You've told them what to demonstrate. They'll frame any story as "leadership" whether it was or not. |
| Hypothetical | "What would you do if your team disagreed with your approach?" | You get rehearsed theory, not evidence of actual behavior. |
| Compound | "Tell me about a time you dealt with conflict on your team and had to deliver under a tight deadline while managing stakeholders" | Three questions in one. The candidate can only address part of it, and you can't tell which parts they skipped vs. forgot. |
The anatomy of a great question
A great behavioral question has three qualities:
1. Open enough for the candidate to choose their best story
Don't prescribe the technology, the team size, or the industry. Let the candidate pick the situation where they shone brightest. Their choice of story is itself data — what do they consider their best work?
2. Focused enough to target a specific principle
The question should naturally surface evidence for one or two of your evaluation principles. "Tell me about a time you identified a problem no one else saw and drove the solution" naturally targets Ownership and Bias for Action.
3. Invites depth, not breadth
The best questions lead to 15-20 minute conversations, not 3-minute summaries. They should have natural probing paths: decisions, tradeoffs, collaboration points, results, and learnings.
Good questions by principle
| Principle | Great Question | Why it works |
|---|---|---|
| Ownership | "Tell me about a time you took on something important that wasn't part of your job description." | Open enough for any role. Naturally reveals initiative, scope expansion, and outcome ownership. |
| Customer Focus | "Tell me about a time you changed your approach because of something you learned from a customer or end user." | Invites stories where user feedback actually changed behavior — not just "I care about users." |
| Collaboration | "Tell me about a time you needed to get buy-in from people who didn't report to you." | Targets influencing without authority — a universal collaboration challenge. |
| Bias for Action | "Tell me about a time you had to make a significant decision without complete information." | Reveals how they handle ambiguity, what "enough information" means to them, and how they course-correct. |
The "tell me about your best work" principle
When in doubt, zoom out. Some of the highest-signal questions are simply:
- "What's the project you're most proud of in the last two years?"
- "Tell me about the most complex problem you've solved recently."
- "What's the best example of you making something significantly better?"
These questions let the candidate bring their A-game. And when someone shows you the best version of themselves — that is the ceiling. You can evaluate: Is this ceiling high enough? Does this best-case scenario demonstrate the principles we care about?
If their best story doesn't show ownership, depth, or impact — that's definitive signal. You're not seeing a nervous candidate on a bad day. You're seeing their highlight reel.
Exercise: Rewrite weak questions
Each question below has a problem and an assigned principle. Rewrite it as a strong behavioral question that targets that principle — open enough for any candidate's best story, specific enough to generate signal.
Question 1 — Target Principle: Bias for Action
Weak question: "Tell me about a time you used Kubernetes to solve a scaling problem."
Problem: Too specific. Only candidates with Kubernetes experience can answer. You're testing for a technology keyword, not a capability.
Rewrite this question to target Bias for Action. Remove the technology constraint but keep the focus on solving a scaling or performance problem under pressure.
Your rewrite should: (1) Ask about past behavior, not hypotheticals. (2) Be open enough for any engineer — frontend, backend, data, infra — to find a relevant story. (3) Naturally surface evidence of moving quickly with imperfect information.
This exercise supports AI-powered coaching via Claude. Enter your access code to enable it, or use the offline feedback below.
Question 2 — Target Principle: Collaboration
Weak question: "Are you good at working with difficult people?"
Problem: Leading and hypothetical. The answer is always "yes." You get a self-assessment, not evidence.
Rewrite this question to target Collaboration. Convert it from a yes/no leading question into a behavioral question that surfaces real evidence of navigating interpersonal difficulty.
Your rewrite should: (1) Ask for a specific past example, not a self-assessment. (2) Focus on the collaboration challenge without telegraphing the expected answer. (3) Invite a story with natural probing paths — decisions, tradeoffs, resolution.
This exercise supports AI-powered coaching via Claude. Enter your access code to enable it, or use the offline feedback below.
Question 3 — Target Principle: Ownership
Weak question: "Tell me about a challenge."
Problem: Too vague. The candidate could talk about anything from a hard math problem to a difficult commute. No principle is targeted, no signal is generated.
Rewrite this question to target Ownership. Add enough focus to surface evidence of taking responsibility beyond one's immediate scope, while keeping it open enough for any candidate.
Your rewrite should: (1) Focus the question on a specific type of challenge — one that reveals ownership behavior. (2) Stay open enough that candidates from any role, industry, or level can find a relevant story. (3) Naturally lead to a 15-minute conversation with probing paths.
This exercise supports AI-powered coaching via Claude. Enter your access code to enable it, or use the offline feedback below.
Why do overly specific questions ('Tell me about a time you migrated from Oracle to PostgreSQL') produce weak signal?
A candidate picks their 'proudest project' and the story is well-told but shows no evidence of collaboration or customer focus. What does this tell you?
Which question best targets the Ownership principle while staying open enough for any candidate?