How Amazon behavioral interviews actually work
Amazon's behavioral interview process is more structured than most companies. Every interviewer is assigned a set of Leadership Principles to evaluate, and they're trained on the “bar raiser” framework — which means they're looking for evidence that you set a high bar, not just that you did your job.
The single most important thing to know:
Use “I”, not “we”. Amazon interviewers are trained to ask follow-up questions if you use “we”: “What was your specific role?” “What decision did you make?” Front-load your individual contribution in every answer.
Leadership Principles — questions & signals by LP
Customer Obsession
Starting with the customer, working backwards. Sacrificing short-term metrics for long-term customer trust.
Ownership
Acting beyond your role. Not saying 'that's not my job'. Taking responsibility for outcomes.
Invent & Simplify
Finding simpler solutions. Challenging processes. Building things that scale.
Bias for Action
Speed matters. Calculated risk-taking. Not waiting for perfect information.
Deliver Results
Hitting targets despite obstacles. Prioritising ruthlessly. Closing.
Dive Deep
Detail orientation. Investigating root causes. Data over assumption.
The STAR framework — Amazon's scoring lens
Set the scene. 1–2 sentences max. What was the context, what was at stake? Include a metric if possible.
What was your specific responsibility in this situation? What outcome were you responsible for?
The most important part. What specific steps did YOU take? 3–5 concrete actions. Use 'I', not 'we'.
What was the measurable outcome? Revenue impact, time saved, error rate reduced. Percentages and dollar amounts are gold.
Example answer — Customer Obsession
Question: “Tell me about a time you went above and beyond for a customer.”
I was the on-call engineer when a payment processing bug caused 3,400 customers to be double-charged during Black Friday peak. This was a P0 incident with $180K in erroneous charges.
My responsibility was to identify the root cause, implement the fix, and ensure affected customers were refunded — while the system was still processing 40K transactions per hour.
I immediately isolated the affected transaction queue, identified a race condition in our idempotency check introduced in the previous deploy, rolled back that specific service without a full rollback, then worked with the finance team to build a refund batch script. I drafted the customer communication email myself rather than waiting for the team lead, because I knew customers would see bank statements before morning.
All 3,400 refunds were processed within 4 hours. The root cause fix was deployed same night. Zero customers escalated to our bank dispute process. I documented the post-mortem and proposed a new pre-deploy check for idempotency validation that we shipped the following sprint.
Build your story library
The most efficient way to prepare is to build 8–10 strong stories that each cover multiple Leadership Principles. Tag each story against the LPs it demonstrates. If a question targets a LP you're weak on, find the closest story in your library and bridge to it.