· Valenx Press · 13 min read
First-Time Manager Feedback Framework Template for Amazon Teams
First-Time Manager Feedback Framework Template for Amazon Teams
The candidates who prepare the most often perform the worst because they mistake process compliance for leadership judgment. In a Q3 hiring debrief at Amazon, a director rejected a candidate with flawless STAR answers because the candidate could not articulate a single moment where they had to withhold feedback to protect a peer’s psychological safety. The framework is not a checklist; it is a signal of your ability to navigate the specific, often brutal, cultural currents of Amazon’s “Have Backbone; Disagree and Commit” principle. Most first-time managers treat feedback as a transaction to be completed, whereas Amazon leaders treat it as a strategic lever to raise the bar. If you walk into a loop presenting a generic template, you are signaling that you cannot distinguish between managing a task and leading a person. The problem isn’t your lack of a script; it is your failure to demonstrate the judgment required to adapt that script when the data is ambiguous or the stakes are personal.
What is the actual feedback framework Amazon leaders use for new managers?
Amazon does not use a universal, static template for first-time managers; instead, it relies on a dynamic synthesis of the “Start, Stop, Continue” model filtered through the rigid constraints of the Leadership Principles. In a calibration session I attended for a Level 6 manager promotion, the hiring manager tore apart a candidate’s feedback log because it focused entirely on output metrics while ignoring the “Earn Trust” mechanic required to deliver that feedback. The core judgment here is that the framework is secondary to the mechanism of delivery; Amazon cares less about the structure of your words and more about whether your feedback loop demonstrates “Insist on the Highest Standards” without violating “Are Right, A Lot.” A generic template fails because it assumes a linear relationship between observation and correction, whereas Amazon’s environment requires a non-linear approach where context, tenure, and the specific LP violation dictate the method. You are not building a form; you are building a case study of how you raise the bar.
The first counter-intuitive truth is that the best feedback frameworks at Amazon often omit specific action items in the initial conversation. I watched a senior director reject a high-potential IC’s promotion to management because their feedback examples were too prescriptive, telling reports exactly what to do rather than guiding them to the root cause. At Amazon, prescribing the solution signals a lack of trust in the report’s ability to problem-solve, which directly contradicts the “Hire and Develop the Best” principle. Your framework must prioritize diagnostic questioning over directive instruction. If your template includes a field for “Action Plan,” it should be reserved for the third conversation, not the first. The goal is to force the report to own the gap between their current state and the leadership principle they are missing. This shifts the dynamic from a manager fixing a worker to a leader developing a peer.
The second insight is that effective feedback at Amazon is rarely about performance; it is almost always about mechanism. During a loop for a product lead, we debated a candidate who gave excellent feedback on missed deadlines but failed to address the missing written narrative that caused the delay. The candidate focused on the symptom (lateness) rather than the mechanism (lack of written clarity). A robust framework for Amazon must explicitly link the behavioral observation to a specific Leadership Principle and then trace it back to a broken process or missing document. If you cannot connect a behavioral issue to a written artifact or a meeting mechanism, your feedback is just opinion, and opinion carries no weight in a culture obsessed with data and written narratives. Your template must force you to identify the missing memo, the skipped review, or the undefined metric before you ever speak to the employee.
How do you structure a feedback conversation using Amazon Leadership Principles?
You structure the conversation by anchoring every observation to a specific Leadership Principle, ensuring the feedback is objective rather than personal. In a tense debrief regarding a manager who was perceived as too aggressive, the committee saved the candidate only because they could demonstrate how their “disagree and commit” moments were documented in writing prior to the decision. The structure must follow a rigid sequence: state the observed behavior, map it to the specific LP, provide the data or written evidence, and then invite the report to diagnose the gap. This is not a soft-skills exercise; it is a forensic analysis of behavior against a codified standard. If you start with “I feel” or “It seems,” you have already failed the “Are Right, A Lot” test because you are leading with emotion rather than evidence. The conversation is a review of facts against standards, not a negotiation of feelings.
The third critical distinction is that you must separate the “What” from the “How” using the “Have Backbone” principle as the fulcrum. I recall a hiring manager arguing that a candidate’s feedback was too soft because they praised the output while ignoring the toxic manner in which it was delivered. At Amazon, results achieved through erosion of trust are not results; they are liabilities. Your framework must explicitly carve out space to address the “How” even when the “What” is successful. A common script for this scenario is: “You delivered the feature on time, which shows Ownership. However, the way you bypassed the design review violated Earn Trust and created technical debt we will pay for in Q4.” This script works because it validates the output while isolating the behavioral defect as a separate, addressable variable. It prevents the report from feeling attacked while making it clear that the behavior is unacceptable regardless of the result.
Do not use the sandwich method; it is ineffective and viewed as disingenuous in high-performing Amazon teams. The “positive-negative-positive” structure dilutes the message and confuses the report about what actually matters. Instead, use a direct “Observation-Principle-Impact” structure. For example: “In yesterday’s meeting, you interrupted the junior engineer three times (Observation). This contradicts ‘Earn Trust’ and ‘Hire and Develop the Best’ (Principle). The impact is that the team is now hesitant to share risky ideas, which slows our innovation velocity (Impact).” This approach is cold, clear, and undeniable. It removes the manager’s ego from the equation and places the focus entirely on the alignment between behavior and company values. The report leaves the room knowing exactly what happened, why it matters, and what needs to change, without any sugarcoating to obscure the reality.
When should a new manager deliver written versus verbal feedback at Amazon?
You should deliver written feedback for any issue that involves a pattern of behavior, a significant miss on a leadership principle, or a decision that impacts the employee’s career trajectory. In a recent calibration, a manager was flagged for risk because they had only provided verbal feedback on a recurring attendance issue, leaving no paper trail when it came time for a performance improvement plan. Amazon’s culture is written; if it is not documented, it did not happen. Verbal feedback is reserved for minor course corrections, immediate tactical adjustments, or relationship-building moments that do not carry long-term HR implications. The judgment call here is about risk management: if you would not want to defend your feedback in front of an HR business partner six months from now, you must write it down immediately. Written feedback creates the “single-threaded” truth that Amazon operates on.
The nuance lies in the timing of the written document relative to the conversation. A common mistake is sending the written summary before the talk, which allows the report to defensive-prepare rather than listen. The correct sequence is the verbal conversation first, followed immediately by a written summary sent within 24 hours. This summary should not be a transcript but a distilled version of the agreement reached. I have seen managers fail their loops because their written follow-ups were vague, using phrases like “we discussed communication” instead of “we agreed that you will lead the weekly status update to practice clarity.” The written record must be specific enough to serve as a baseline for the next review cycle. If your written feedback cannot be measured against a future state, it is useless noise.
There is a specific exception for “real-time” feedback during live events, which should always be verbal and immediate. If a manager interrupts a report during a customer call, the correction must happen the moment the call ends, not in a weekly one-on-one. Waiting dilutes the context and reduces the learning opportunity. However, this immediate verbal correction must still be logged in a written format later if it represents a recurring theme. The framework here is “Verbal for Immediacy, Written for Pattern.” If you find yourself having the same verbal conversation three times, you have failed to escalate to a written mechanism, and you are now enabling the behavior. The transition from verbal to written is the signal that the issue has moved from a tweak to a structural problem requiring formal intervention.
What specific phrases should you use to disagree and commit during feedback?
You should use phrases that explicitly acknowledge the disagreement while affirming the commitment to the decision, such as “I see the data differently, but we will move forward with your approach and review the metrics in two weeks.” In a high-stakes product launch debate, a principal engineer used this exact phrasing to allow a junior PM to own a risky feature, turning a potential conflict into a development opportunity. The language must be precise enough to show you have “Backbone” but flexible enough to show you trust the team’s execution. Avoid phrases like “I guess we can try” or “Fine, do it your way,” as these signal passive aggression rather than active commitment. The goal is to demonstrate that you can separate your ego from the outcome and that you are willing to bet on your team even when you disagree.
The script for delivering negative feedback on a missed commitment must be direct and devoid of emotional qualifiers. Use: “The data shows we missed the launch date by four days. This impacts our Q3 goals. Let’s walk through the mechanism that failed so we can ensure it doesn’t happen again.” Notice the absence of “I’m disappointed” or “You let us down.” These emotional additions shift the focus from the system to the person, which triggers defensiveness. At Amazon, the system is always the culprit until proven otherwise. By focusing on the mechanism, you invite the report to collaborate on a solution rather than defend their character. This is the essence of “Insist on the Highest Standards”: you hold the line on the result without attacking the human.
When addressing a violation of “Customer Obsession,” the phrasing must tie the internal behavior directly to an external customer impact. Say: “By skipping the user testing phase, we launched a feature that increased support tickets by 15%. This violates Customer Obsession because we prioritized speed over customer experience.” This links the internal action to an external metric, making the feedback undeniable. It removes the manager’s opinion from the equation and replaces it with customer data. If you cannot find a customer metric to anchor your feedback, you likely haven’t dug deep enough into the problem. The phrase “This impacts the customer because…” is your most powerful tool for grounding abstract leadership principles in concrete business reality. It forces the report to see beyond their immediate task and understand their role in the broader ecosystem.
Preparation Checklist
- Draft a specific “Observation-Principle-Impact” script for your next difficult conversation, ensuring you cite a specific data point or written artifact rather than a general feeling.
- Review your last three one-on-one notes; if they lack a direct link to a Leadership Principle, rewrite them to explicitly map behaviors to “Earn Trust” or “Deliver Results.”
- Create a standardized follow-up template that summarizes verbal agreements within 24 hours, focusing on measurable next steps rather than vague intentions.
- Work through a structured preparation system (the PM Interview Playbook covers Amazon Leadership Principle behavioral mapping with real debrief examples) to ensure your feedback scenarios are rigorous enough for a Level 6 calibration.
- Identify one recurring team issue and design a written mechanism (a new doc template, a meeting agenda change) to address it, rather than planning another verbal reminder.
- Practice the “Disagree and Commit” script with a peer, focusing on maintaining a neutral tone while clearly stating your opposing view.
- Audit your feedback log for the “Sandwich Method”; replace any instance of buffered criticism with direct “Observation-Principle-Impact” statements.
Mistakes to Avoid
BAD: Starting a feedback session with “I feel like you aren’t communicating well.” GOOD: “In the last three sprint reviews, you did not share the status of the backend dependencies, which caused the frontend team to idle for six hours. This violates ‘Earn Trust’ and impacts our delivery speed.” Judgment: Vague feelings invite debate; specific data points close the loop. Amazon leaders do not deal in vibes; they deal in facts.
BAD: Giving verbal feedback on a serious performance issue and failing to document it until the quarterly review. GOOD: Holding the verbal conversation immediately, then sending a written summary within 24 hours that outlines the specific gap and the agreed-upon remediation plan. Judgment: Undocumented feedback is a liability. If it isn’t written, it didn’t happen, and you are failing your duty to develop the employee.
BAD: Using the “Sandwich Method” to soften a blow about a missed deadline. GOOD: Stating the miss, linking it to the Leadership Principle, explaining the customer impact, and asking the report to propose a fix. Judgment: Softening the message dilutes the standard. High performers want clarity, not comfort. Ambiguity is the enemy of excellence.
FAQ
Can I give feedback without citing a specific Leadership Principle? No. At Amazon, feedback without an LP anchor is just personal opinion and lacks the authority to drive change. Every piece of corrective feedback must map to a specific principle like “Insist on the Highest Standards” or “Earn Trust” to be considered valid in a performance review. Without this link, you are managing based on preference, not culture.
How soon after an incident should I deliver written feedback? You must send the written summary within 24 hours of the verbal conversation. Waiting longer allows the details to fade and signals that the issue is not a priority. Immediate documentation ensures the “single source of truth” is established while the context is fresh for both you and the report.
What if my report disagrees with my feedback during the conversation? Listen to their data, but hold the line on the standard. If they provide new information that changes the context, acknowledge it and adjust. If they are simply defending their behavior, reiterate the impact on the customer or the team and remind them of the “Disagree and Commit” expectation. The conversation is about alignment, not winning an argument.amazon.com/dp/B0GWWJQ2S3).