Introduction
But in a TPM or PgM interview at Amazon, Google, or Netflix, this "perfect planning" might actually cost you the job.
Why? Because modern tech companies value Bias for Action over perfect planning. They know that in software, you can never know everything.
The interviewer will ask: "Tell me about a time you had to make a high-stakes decision with incomplete data."
If your answer is, "I waited until I gathered more data," you fail.If your answer is, "I used the 70% rule to make a calculated bet," you’re hired.
In this post, we’ll explore how to stop overthinking and start executing—and how to frame this critical skill in your next interview.
The Trap: The "Perfect Plan" Fallacy
Junior Program Managers think their job is to prevent failure through planning.Senior Program Managers know their job is to navigate failure through execution.
Spending weeks on a "perfect" plan for a software project is wasted time because the requirements will change. The "Art of Program Execution" isn't about avoiding the fog; it's about learning to drive through it.
The Framework: How to Act When You Don't Know Everything
When answering the "incomplete data" interview question, use this 3-part framework to show you have strategic judgment, not just recklessness.
1. The "Two-Way Door" Assessment (The Amazon Principle)
First, determine the reversibility of the decision.
- One-Way Door: If we do this, we can't go back (e.g., publicly launching a hardware product, deleting a database). These require 90% information.
- Two-Way Door: If we screw up, we can roll back (e.g., launching a beta feature, changing a UI color, internal process change). These only require 70% information.
Your Interview Soundbite: "I realized this was a 'Two-Way Door' decision. If the new API failed, we could roll back to the old one in 15 minutes. Therefore, the cost of waiting for perfect data was higher than the cost of trying and failing."
2. The 70% Rule
Jeff Bezos famously said most decisions should be made with around 70% of the information you wish you had. If you wait for 90%, you’re too slow.
- Action: In your story, explain what the "missing 30%" was (e.g., "We didn't know exactly how many users would click...") and why you moved forward anyway (e.g., "...but we knew the server load would hold, so we launched to 5% of traffic to get that data").
3. The "MVP Process"
Don't just apply MVP (Minimum Viable Product) to code; apply it to your Program Management.
- Instead of spending a month designing the perfect Jira workflow, implement a "Good Enough" spreadsheet today.
- The Lesson: Value iterations over perfection.
Why "Bias for Action" Matters for TPMs
Technical programs often stall because engineers are perfectionists. They want to refactor the code one more time.
As a TPM, your job is to be the forcing function for delivery
Shutterstock
. You need to be the one who says, "This architecture isn't perfect, but it scales to 100k users. Let's ship it, learn, and refactor later."
Stop Thinking. Start Preparing.
"Bias for Action" is one of the hardest competencies to prove because it requires confidence.
Our Art of Program Execution (TPM) Kit and Mastering Product Management Guide provide:
- Behavioral Story Banks: Specific examples of "Bias for Action" stories that have worked in FAANG interviews.
- Risk Management Frameworks: How to calculate when it's safe to move fast.
- The Amazon Leadership Principles: A deep dive into how to answer "Have Backbone; Disagree and Commit" and "Bias for Action."
Don't let overthinking kill your interview chances.
👉 Get the TPM Prep Kit or Get the PM Prep Guide today.
FAQs
Q1: Isn't "Bias for Action" just being reckless?No. Recklessness is acting without considering risk. Bias for Action is acting despite the risk because you have calculated that the cost of delay is higher. You must show the interviewer you did the risk calculation (Step 1 of the framework).
Q2: How do I answer "Tell me about a time you made a mistake?" using this?This is the perfect pairing! "I used the 70% rule to launch a feature quickly. We made a mistake (the UI was confusing), but because I treated it as a Two-Way Door (a small beta), we fixed it in 24 hours. The speed of our learning outweighed the cost of the error."
Q3: Does this apply to Waterfall projects?Even in Waterfall/Hardware, there are moments for action. You can't change the hardware mold once it's cast (One-Way Door), but you can make fast decisions on the packaging design or the marketing plan (Two-Way Doors).
.jpg)
