Lean Startup: Build, Measure, Learn

Lean Startup is a framework popularized by Eric Ries, based on Lean principles (originally from Toyota), adapted to the context of creating new products or businesses, especially in high uncertainty environments such as technology startups.

Its main objective is to eliminate waste in the development process, understanding as waste any effort that does not contribute to learning what customers really want.

Instead of investing large amounts of time and resources in building a complete product based on assumptions, Lean Startup proposes a rapid cycle to validate business and product hypotheses empirically, using feedback from the market as a primary guide.

Lean Startup Principles (MVP, Continuous Deployment, Pivot)

Lean Startup is based on several key principles to accelerate learning and reduce risk:

• Minimum Viable Product (MVP - Minimum Viable Product): build the simplest possible version of the product that allows testing a fundamental hypothesis about the customer or market with minimal effort. It allows to learn quickly before investing more.
• Continuous Deployment: Deliver software (or value) to customers as often as possible, ideally several times a day. This shortens the feedback cycle and allows to validate changes quickly. Practices such as "Push on Green" are enabling.
• Split Testing / A/B Testing: Experiment with showing different versions of a feature to different user segments simultaneously to measure which one works best, basing decisions on real data.
• Pivot: Making a structured and significant course correction to the product or business strategy when data shows that the original hypothesis is incorrect. It is a fundamental change, not a simple optimization.
• Others: Include Actionable Metrics and Innovation Accounting.

The Build-Measure-Learn Cycle

The central driver of Lean Startup is the Build-Measure-Learn feedback cycle.

It works like this:

• Build: An MVP or experiment is quickly created to test a specific hypothesis about the product or market.
• Measure: You launch the MVP or experiment to users and collect quantitative and qualitative data on their behavior and feedback.
• Learn: The data collected is analyzed to validate or invalidate the initial hypothesis.
• This learning informs the next decision: persevere with the current strategy (if the hypothesis is validated) or pivot (if it is invalidated).

The goal is to go through this cycle as quickly as possible to maximize validated learning about customers with minimal waste of resources.

Actionable Metrics and Innovation Accounting

For the Build-Measure-Learn cycle to be effective, Lean Startup introduces two key concepts:

Actionable Metrics: these are metrics that demonstrate a clear cause-effect relationship and help make informed product decisions.

Unlike "vanity metrics" (e.g., total number of visits, which can be misleading), actionable metrics (e.g., conversion rate per cohort, engage