In the Think → Ship → Repeat cycle, the "Think" phase culminates in a critical decision: What is the smallest thing we can build that still matters?

Many product teams mistake an MVP (Minimum Viable Product) for a "v1.0" or a "subset of features." In reality, identifying an MVP is an exercise in ruthless subtraction. It is about stripping away everything except the Core Value Loop required to prove—or disprove—your hypothesis.

1. Beyond the Feature List: The Value-First Mindset

An MVP is not a collection of "must-have" features; it is a learning vehicle. When you are in the "Think" phase, your goal is to identify the single path through your product that solves the user's primary problem.

• The "One-Job" Rule: If your product were allowed to do only one thing, what would it be? Everything else is a distraction for the first "Ship."

• Viability over Completeness: "Viable" doesn't mean "basic." It means it provides enough value that a user is willing to change their behavior, give you their data, or pay you money—even if the UI is unpolished.

2. The MVP Identifier Framework

To identify your MVP, move from the abstract "Think" to a concrete "Ship" plan by asking three questions:

1. What is the "Atomic Unit" of Value? (e.g., For Uber, it’s a car arriving at a location; for Slack, it’s a message being read by a teammate.)

2. What is the "Critical Path"? Identify the minimum steps a user must take to experience that atomic unit.

3. What can be handled manually? (The "Concierge" approach). If you can perform a task behind the scenes using a spreadsheet or a manual process instead of building an automated system, do it.

3. AI-Augmented Scoping: Cutting the Fat

Generative AI is a powerful tool for preventing "scope creep" during MVP identification. Use it as a sounding board to challenge your assumptions:

• The "Inversion" Prompt: Feed your feature list to an AI and ask: "Which of these features can be removed while still allowing the user to achieve [Primary Outcome]?"

• Synthetic User Testing: Ask the AI to simulate a skeptical user: "If I only gave you [Feature A] and [Feature B], would you still find this valuable enough to use daily? Why or why not?"

• Edge Case Identification: Use AI to find the 20% of edge cases that are taking up 80% of your development time, and then consciously decide to not solve them in the MVP.

4. The "Definition of Done" for an MVP

A successful MVP identification in the "Think" phase results in a scope that feels uncomfortably small. If you don't feel a slight pang of anxiety that you’re "missing something," you’re likely building too much. Remember: the goal of shipping an MVP isn't to win a design award; it’s to start the Repeat cycle with real-world data.

Why Subtraction is the Ultimate Strategy

By identifying the true MVP, you protect your most valuable resource: Time. Shipping a bloated "version one" takes months and risks building the wrong thing. Shipping a lean MVP takes weeks and gives you the map for the rest of the journey.

In the relentless pursuit of product-market fit, speed of learning is the only metric that matters.