AgentSkillsCN

eigenquestion-product-assessment

评估候选人能否精准识别出决定产品战略的1至2个核心变量（即“本征问题”）。在产品经理、工程师或领导岗位的面试过程中运用此技能，可有效检验其基于第一性原理的思考能力、战略思维的清晰度，以及在模糊且低风险的环境中化繁为简的能力。

SKILL.md
--- frontmatter
name: eigenquestion-product-assessment
description: Evaluates a candidate's ability to identify the 1-2 fundamental variables (Eigenquestions) that dictate a product's strategy. Use this during interviews for PM, engineering, or leadership roles to test first-principles thinking, strategic clarity, and the ability to reduce complexity in ambiguous, low-stakes environments.

The Eigenquestion Assessment is a "coded" test that separates candidates who get bogged down in details from those who can identify the core drivers of a business. By using a hypothetical scenario, you remove domain bias and force the candidate to demonstrate their mental model for decision-making.

The Process

1. The Setup

Present a low-stakes, high-impact hypothetical scenario. The goal is to provide a blank slate with infinite possibilities.

  • •The Prompt: "A group of scientists has invented a teleportation device. They’ve hired you to be their business counterpart to bring this to market. What do you do?"
  • •The Observation Phase: Allow the candidate to ask initial clarifying questions. They will likely ask about range, safety, size, or power requirements. Do not provide detailed answers; keep them brief.

2. The Constraint (The Pivot)

Once the candidate has generated a list of questions, introduce a hard constraint to force prioritization.

  • •The Pivot: "The scientists are annoyed by your questions. They will only answer two of your questions. After that, they expect a full strategic plan. What two questions do you ask?"

3. Identifying the Eigenquestions

An Eigenquestion is a question where the answer provides the answer to many other questions. Evaluate if the candidate's chosen questions allow them to create a decision matrix or quadrants.

  • •Good Eigenquestions: Address existential risks (safety) or business models (cost structure).
  • •Weak Questions: Address cosmetic or incremental details (color, size, branding) that don't fundamentally change the product's "class" or target market.

4. The Strategy Map

Require the candidate to explain how the answers to their two questions would change their go-to-market plan. They should be able to describe at least two distinct paths based on the variables they chose.

Examples of Application

Example 1: The Teleportation Device

Context: Evaluating a Senior PM for strategic depth. The Two Questions:

  1. •Is it safe for humans?
  2. •Is it expensive to build (CapEx) or expensive to run (OpEx)? The Application:
  • •Scenario A (Safe + Cheap to build/Expensive to run): Deploy "human fax machines" everywhere. High accessibility, pay-per-use model.
  • •Scenario B (Safe + Expensive to build/Cheap to run): Build major hubs. Replace airports and long-distance rail.
  • •Scenario C (Not safe for humans): Pivot to high-value cargo, organ transport, or hazardous waste disposal.

Example 2: The Physical Retail Placement

Context: Evaluating a Growth Lead's understanding of market saturation. The Scenario: "There are three gas stations on the same street corner. You are tasked with opening a fourth. What do you do?" The Eigenquestion: "Is the customer's primary driver convenience of location or price sensitivity for this specific route?" The Application:

  • •If Location: Analyze traffic flow patterns (right-turn vs. left-turn access) to capture a different "segment" of the same intersection.
  • •If Price: Analyze the supply chain/loyalty program depth to see if you can undercut the existing three on margin.

Common Pitfalls

  • •Accepting "Interesting" over "Useful": Candidates often ask "interesting" questions (e.g., "How does it work?") that don't help them make a business decision. The Fix: Always ask, "What decision would the answer to that question allow you to make?"
  • •Providing Too Much Data Early: If you answer every initial question, the candidate never has to prioritize. The Fix: Remain vague until you move to the "Two Question" constraint.
  • •Ignoring the Strategy Map: A candidate might pick two good questions but fail to explain how the answers dictate different paths. The Fix: Force them to build a 2x2 matrix or a set of "if/then" scenarios based on their questions.
  • •Focusing on "Right" Answers: There is no single correct Eigenquestion. The goal is to see if the candidate's questions are structurally sound and lead to a structured path forward.