📖 ~20 min read | Stop building products. Start running experiments.
The Lean Startup
Eric Ries · 2011
Why this book matters
Eric Ries synthesised lean manufacturing, agile development, and Steve Blank’s customer development methodology into a complete framework for building companies under conditions of extreme uncertainty. The vocabulary it introduced — MVP, pivot, validated learning, Build-Measure-Learn — is now the standard language of product development globally.
Most valuable not for the specific techniques (many of which have been refined in subsequent work) but for the fundamental reframe: a startup is not a small version of a large company. It’s a different kind of organisation, designed for a different purpose: searching for a repeatable, scalable business model in conditions where the right answer is unknown.
“The only way to win is to learn faster than anyone else.”
— Eric Ries
The Build-Measure-Learn loop — the engine of everything
The core operational unit: form a hypothesis → build the minimum version needed to test it → measure whether the hypothesis is confirmed or refuted → learn from the result → repeat, as fast as possible.
The goal is not to build a great product. The goal is to complete the loop as fast as possible — because speed through the loop determines how much you can learn in a given time period. The startup that completes ten learning cycles in the time its competitor completes one has ten times as many opportunities to discover what actually works.
Run the loop backwards: start with what you want to learn (which assumption is most dangerous?), work backward to the minimum experiment that would validate or refute it, then work backward to the minimum product required to run that experiment. This produces experiments that are smaller, faster, and more informative than starting from ‘what should we build?’
The MVP — most misunderstood concept in product development
The Minimum Viable Product is not the smallest crappy version of the full product. It’s the minimum investment required to test the most dangerous assumption facing the business. What it needs to contain depends entirely on what you’re testing.
The Dropbox MVP was a video — not a product at all. A three-minute demonstration of what Dropbox would do if it were built. Thousands of people signed up for a product that didn’t yet exist. The assumption being tested was not ‘can we build this?’ but ‘will people want this?’ The video was sufficient to test the second question at a fraction of the cost of building the answer to the first.
Vanity metrics vs actionable metrics
Vanity metrics — page views, signups, downloads, social followers — make you feel good and tell you nothing about whether your business is working. The question is never ‘how many people signed up?’ — it’s ‘what percentage came back and did the thing that indicates the product is creating real value?’ Cohort analysis, retention curves, and activation rates reveal whether the product is actually working. Most startups optimise for vanity metrics because they’re easier to move and more comfortable to report.
Pivot or persevere — the most important recurring decision
A pivot is a structured course correction — changing one element of the business model based on evidence that the current approach isn’t working. Not giving up. Not starting over. Applying learning from failed experiments to a refined hypothesis.
Multiple types: customer segment (same problem, different customer), problem (same customer, different problem), solution, revenue model, channel, technology. The framework prevents the false binary of ‘continue exactly as we are’ versus ‘start over entirely.’
The failure to pivot when evidence requires it is as costly as pivoting unnecessarily. The startup that continues investing in a product, market, or model after the evidence has clearly shown it isn’t working is not demonstrating commitment — it’s demonstrating the inability to distinguish perseverance from denial. If experiment results look like noise — not clearly good, not clearly bad — that’s strong evidence the fundamental hypothesis needs to change.
Innovation accounting — measuring what matters
Conventional accounting tells you the state of a business at a point in time. For an early-stage startup these measures are largely meaningless — the values are too small and the trajectory matters more than the absolute level.
Innovation accounting works in three steps: establish the baseline — where are you today on the metrics that measure whether the product creates real value? Experiment to move it — what specific changes should improve those metrics? Pivot or persevere — given the actual movement from the experiments run, is the current approach working or is a fundamental change required?
Most startups never do this explicitly and end up optimising for whatever is easiest to measure rather than whatever is most important to measure.
“The only way to win is to learn faster than anyone else.”
— Eric Ries
📌 Bookmark this. Come back before you pick the book up again.
— Dan @ The Mojito Network
Summary by Dan Marques. All ideas, frameworks, and quotes belong to the original author. Buy the book — this summary is a reason to do it, not a replacement for it.