Amazon's leadership principles need story banks, not 16 scripts

The biggest prep miss is writing sixteen STAR scripts — one per Leadership Principle.
It feels organized. It usually folds under the first hard follow-up.

Amazon publishes all sixteen on its Leadership Principles page and says you will be evaluated on them throughout the loop (Amazon Jobs — Leadership Principles).

Former Bar Raiser Dave Anderson is blunt in his LP interview guide: recruiters can tell who memorized a post versus who can defend a decision.

For SWEs, build a story bank: 6–8 real stories, each tagged to multiple principles, each sturdy enough for a dig.

What they are listening for

Weak answers usually fail in one of these ways:

• Team did X — your ownership never appears
• No numbers — impact sounds inflated
• No trade-off — judgment looks thin
• Too perfect — sounds rehearsed
• Follow-up “why?” collapses the story

They are really asking a smaller set of questions:

• How did you decide under ambiguity?
• Did you dig past the dashboard?
• Did you push when the right move was uncomfortable?
• Did the outcome improve, or did you stay busy?
• When you missed, did you own it?

Anderson’s Three Tales from an Amazon Bar Raiser shows ordinary projects beating flashy ones when the trade-off is clear.

Principles that show up most for SWEs

Know all sixteen at recognition level. Amazon added Strive to be Earth's Best Employer and Success and Scale Bring Broad Responsibility in 2021 (Anderson’s write-up).

For software loops, our prep advice is to build stories against these first:

• Customer Obsession — you changed the plan because someone was getting hurt
• Ownership — you fixed it even outside your formal scope
• Dive Deep — you found the real failure mode
• Bias for Action — incomplete info, not reckless
• Earn Trust — disagreement without defensiveness
• Insist on the Highest Standards — you refused to ship half-correct
• Have Backbone; Disagree and Commit — push back, then support the call
• Deliver Results — pressure hit and the outcome moved

Senior / staff probes lean more on Think Big, Hire and Develop the Best, Earth’s Best Employer, Broad Responsibility.

Story bank, not script bank

Tag each story with 2–4 principles. Reuse across rounds; don’t clone sixteen openings.

Answer shape

1. Situation — one breath of context
2. Your ownership — “I owned…” not “we shipped…”
3. Decision + trade-off — what you chose and what you accepted
4. Result — numbers if you have them; engineering proxies if not
5. Reflection — what you would change or what mechanism you added

Exponent’s LP guide names the “we-trap”: collaboration without ownership gives them nothing to grade.

Follow-ups that break scripts

Expect:

• Why that trade-off?
• What data changed your mind?
• What did you miss first?
• Who disagreed?
• What if you had done nothing?
• How do you know your change caused the result?

For each story, pre-write: metric, trade-off, risk, mistake, next mechanism.
If those five answers are crisp, the story holds.

Six shapes to practice

DesignGurus lists dozens of variants; they compress to:

• You changed direction on customer evidence
• You owned something outside your role
• You disagreed with a technical call
• You moved with incomplete data
• You found a root cause others missed
• You failed and changed the system afterward

Stay operational. Passion adjectives don’t help.

Two-hour prep

Hour 1 — bank: six stories from the last 2–3 years; tag principles; write metric, trade-off, mistake.

Hour 2 — pressure: say each in ~2 minutes; add one follow-up; cut openings that take >20 seconds to reach the problem; kill “helped/supported.”

Pair with the Amazon company page and a coding pass on practice problems so behavioral prep stays tied to the real loop.

The goal is not polish. The goal is sounding like an engineer who made hard calls and can explain them.