The Behavioral Round
Engineers love to dismiss the behavioral round as “soft.” That’s a costly mistake — at Amazon it’s half the loop, at Netflix it can be the whole thing, and everywhere it’s where a hiring manager decides whether you’re someone they want in the room at 2 a.m. during an incident. The good news: unlike coding, it’s almost entirely preparable in advance. The questions are predictable, and the structure is fixed.
STAR — the only structure you need
Every “tell me about a time…” answer follows STAR:
| Aim for | ||
|---|---|---|
| S — Situation | Set the scene and the stakes, briefly. | ~15% |
| T — Task | Your specific responsibility or goal. | ~10% |
| A — Action | What you did, step by step. The heart of the answer. | ~60% |
| R — Result | The outcome, quantified, plus what you learned. | ~15% |
The number-one behavioral failure is inverting these proportions — five minutes of backstory (S) and one sentence of what you actually did (A). Interviewers are scoring your judgment and actions, so the Action is where 60% of your words belong, and they should be “I” not “we.”
Build one and grade it
Draft a real answer to a classic prompt below. The checklist grades the four things interviewers actually look for: all four parts present, Action as the largest section, “I” (not just “we”), and a quantified result.
“We” hides you; “I” reveals you. Candidates instinctively say “we decided… we built… we shipped.” But the interviewer is hiring you, not your team — they need to know what you personally contributed. Say “I proposed…”, “I wrote…”, “I convinced the team by…”. Using “we” throughout the Action is the most common reason a true story still scores poorly: it gives no evidence of your individual judgment.
The stories you must have ready
Prepare 6–8 stories in advance, each ~2–3 minutes, drawn from real projects. A good portfolio covers these themes — and most stories can be reframed to answer several prompts:
A hard technical problem you solved
The depth-and-rigor story. Be ready to go arbitrarily deep — interviewers will drill (“why that approach? what did you measure? what would you change?”).
A conflict / disagreement with a teammate or manager
How you handled friction professionally. The key beat: you sought to understand the other side and resolved it with data or compromise, not by “winning.”
A failure or mistake you owned
The single most important story, and the one people fumble. Pick a real failure, own it without blaming others, and emphasize what you learned and changed. “I can’t think of one” is a red flag.
Leadership / driving something without authority
A time you influenced an outcome, mentored someone, or pushed a project forward when it wasn’t strictly your job.
Ambiguity / a tight deadline
Operating when the goal was unclear or time was short — how you scoped, prioritized, and shipped.
Going above and beyond for a user/customer
Especially for Amazon (Customer Obsession). A time you dug into what the user actually needed.
Amazon’s Leadership Principles — the special case
Amazon evaluates its 16 Leadership Principles in nearly every round (coding included), and a Bar Raiser can veto an offer on culture signal alone. If you’re interviewing there, pre-tag each of your stories to the LPs it demonstrates. The most-probed ones:
| Leadership Principle | A story that demonstrates it |
|---|---|
| Customer Obsession | dug into a user pain point others dismissed |
| Ownership | fixed something outside your lane; thought long-term |
| Dive Deep | found a root cause through data others missed |
| Bias for Action | made a reversible call fast instead of waiting |
| Invent and Simplify | removed complexity / found a simpler solution |
| Have Backbone; Disagree and Commit | respectfully pushed back, then committed once decided |
| Deliver Results | shipped under constraints with a measurable outcome |
One story can serve many principles — your “shipped a migration under a tight deadline” story might hit Ownership, Bias for Action, and Deliver Results. Build a small matrix: stories down the side, LPs across the top, check the boxes. Then whatever LP the interviewer probes, you can reach for the right story in seconds instead of fishing. This prep is what separates Amazon passes from Amazon rejections.
Questions to ask them
Every round ends with “any questions for me?” — always have 2–3, and make them specific, not generic. Strong questions signal genuine interest and seniority:
- “What does success look like for this role in the first 6 months?”
- “What’s the hardest technical problem the team is facing right now?”
- “How does the team balance shipping speed against tech debt?”
- “What do you personally enjoy / find frustrating about working here?”
Avoid questions easily answered by the careers page, and never lead with comp/vacation in a technical or hiring-manager round (that’s a recruiter conversation).
Quick check
Next: The System Design Round — the design interview at speed, built on the Day 28 toolkit.