Eric Ries · 2011 · Business
Read
Eric Ries argues that most startups fail not because they cannot build their product but because they build something nobody wants. Drawing on his own experience, including a company that spent months on features customers ignored, he proposes a scientific way to create new products and businesses under extreme uncertainty.
The central idea is the build-measure-learn loop. Instead of executing a detailed plan, a team identifies its riskiest assumptions, builds a minimum viable product to test them, measures real customer behaviour and decides whether to persevere or pivot. Ries adapts ideas from lean manufacturing, such as small batches and asking why five times, and argues that progress should be measured by validated learning rather than activity or vanity metrics. The approach applies to large organisations as well as startups.
Key insights
- Identify your leap-of-faith assumptions, about whether customers value the product and how it will grow, and test the riskiest ones first.
- Build a minimum viable product, the smallest thing that lets you learn from real customers. It might be a video, a landing page or a manual service.
- Ignore vanity metrics like total sign-ups. Track actionable measures, such as how each new group of customers behaves, that show whether changes actually help.
- Hold regular pivot-or-persevere meetings. A pivot changes strategy while keeping what you have learned, so it is a course correction rather than a failure.
- When something goes wrong, ask why five times to reach the root cause, then make a proportionate fix at each level of the problem.
Read it if you are launching something new and want to stop wasting time building things nobody needs.