Overview
Eric Ries treats a startup as an experiment run under great uncertainty. Instead of spending months perfecting a product in secret, he says founders should put a rough version in front of real customers, measure what they do, and adjust. He draws on his own company, IMVU, and on lean manufacturing ideas from Toyota. The aim is to waste less time building things nobody wants.
In brief
Build, measure, learn
A startup turns an idea into a product, watches how customers respond, and learns from the data. Ries says the aim is to get through this loop as quickly as possible.
Minimum viable product
This is the simplest version of a product that lets you test an assumption with real customers. It can be crude. Its job is to produce learning, and polish can come later.
Pivot or persevere
After each round of testing, founders must decide whether to keep going or change one part of the strategy. Ries calls the change a pivot and recommends regular meetings to make the decision.
Validated learning
Progress is measured by what you have proved about customers, not by features shipped. Ries warns against vanity metrics, such as total sign-ups, which can rise even when the business is not working.
Who should read this
Someone planning a business, an app or a side project who is afraid of spending months and savings on something nobody wants. Also useful for people inside larger companies who are asked to launch something new.
Who might skip it
Skip it if you've already absorbed the ideas through startup culture — MVP, pivot, build-measure-learn are now so embedded that reading 300 pages of the original may feel repetitive. Also skip if you're running a mature business; the methodology is specifically designed for conditions of extreme uncertainty, not for scaling what already works.
A startup is an experiment
Ries defines a startup as a group of people creating something new under extreme uncertainty. Size and industry do not matter. A team inside a bank building a new service counts. A person selling home-cooked food on Instagram counts too. What they share is that nobody yet knows who the customer is or what the customer will pay for.
Because of that, he argues, a traditional business plan is mostly guesswork dressed up with numbers. He learned this at his own company, IMVU, where the team spent months building a product around an assumption about how customers would use it. When they finally released it, customers did not want to use it that way at all. Much of that early work was thrown away.
His conclusion is that the real output of a young company is knowledge about what customers want.
Learning you can prove
Every founder who fails says they learned a lot. Ries wants a stricter standard, which he calls validated learning. It means you can show with real customer behaviour that you have found something true about your business.
He asks you to write down the assumptions your plan depends on. Two matter most. The value hypothesis asks if customers really get value from the product once they use it. The growth hypothesis asks how new customers will find it. He calls these leap-of-faith assumptions, because everything else in the plan falls if they are wrong.
Take a student who wants to start online tuition classes. The plan assumes parents in the area will pay for online tuition, and that satisfied parents will tell others. Both can be tested in two weeks with a WhatsApp message offering a trial class for a small fee. If ten parents pay, that is evidence. If fifty say it sounds like a nice idea and nobody pays, that is evidence too, and it costs far less than renting a room and printing banners.
The minimum viable product
A minimum viable product, or MVP, is the smallest thing you can put in front of customers to test an assumption. It is not a cheap version of the final product. Its only purpose is learning.
The book gives two well-known examples. Zappos began when its founder photographed shoes in local shops and posted the pictures online. When someone ordered, he bought the pair at full price and shipped it. He found out that people would buy shoes online without trying them on, before spending anything on stock or warehouses. Dropbox faced a product that was hard to build, so its founder first made a short video showing how it would work. The waiting list grew sharply overnight, which showed demand before the hard engineering was done.
Founders fear that a rough product will damage their name or that a competitor will copy the idea. His answer is that early customers forgive rough edges if the product solves a real problem, and that most ideas are in more danger from being ignored than from being stolen.
Build, measure, learn
The working method of the book is a loop. You build something small, measure how customers respond, and learn whether your assumption held. Then you go round again. Ries says the aim is to reduce the total time through the loop, since each round is one lesson.
Measurement is where many teams fool themselves. He warns against vanity metrics, which are numbers that always go up and make you feel good, such as total downloads or total registered users. They say nothing about whether the latest change worked. He prefers actionable metrics, especially cohort analysis, where you look at each group of new customers separately. If a hundred people signed up in March and twelve paid, and a hundred signed up in April after your redesign and eleven paid, the redesign did nothing, whatever the total user count says. This kind of bookkeeping is part of what he calls innovation accounting.
Pivot or persevere
At regular intervals a founder has to decide whether to keep going in the same direction or change. Ries calls the change a pivot: a deliberate shift in one part of the strategy while keeping what you have learned. You might keep the product and aim it at a different type of customer, or take one feature that people love and make it the whole product.
He is clear that the decision is hard, mostly for emotional reasons. Vanity metrics let a team believe it is making progress. Admitting a wrong guess feels like failure. So companies drift, neither growing nor dying, using up money. He suggests scheduling a pivot-or-persevere meeting in advance, so the question gets asked regularly with data on the table.
He also changes the way a startup's remaining life is counted. The usual measure is months of cash left. He counts the number of pivots you can still afford.
Growth and fixing root causes
Ries describes three engines of growth. In the sticky engine, growth comes from keeping customers so that few leave. In the viral engine, customers bring in others simply by using the product. In the paid engine, you spend on advertising and grow as long as a customer earns you more than it cost to win them. He advises concentrating on one engine at a time, since each needs different measurements.
For problems inside the company he borrows the Five Whys from Toyota. When something breaks, ask why, then ask why of that answer, five times. A crashed website may trace back to a new employee who was never trained, and the real fix is training. He adds that the effort put into each fix should be in proportion to the size of the problem.
The whole approach comes from lean manufacturing, which is why the book keeps returning to waste. The waste Ries worries about most is a team building, on time and on budget, something nobody wants.
Try this week
- Write down the two assumptions your idea depends on most: who will pay, and how they will hear about it.
- Talk to five possible customers this week and ask what they currently do about the problem, not whether they like your idea.
- Design a test that costs almost nothing, such as a simple post or form offering the product for advance order.
- Choose one number that reflects real customer behaviour, like repeat orders, and stop counting likes and followers.
- Put a date in your calendar four weeks from now to decide, using that number, whether to continue or change direction.
Our honest take
The examples come almost entirely from software and internet companies, where changing a product is cheap. The method is harder to apply to medicine, hardware, food or anything regulated, and Ries does not fully deal with that. The term MVP has also become an excuse in some companies for shipping careless work. The evidence is mostly stories from companies that survived, several of which look dated now, and the book's own vocabulary can turn into jargon that replaces thinking.
If you liked this
Read The Mom Test by Rob Fitzpatrick for the customer-interview half of the problem. Zero to One by Peter Thiel for the counter-argument.
Questions readers ask
Is The Lean Startup worth reading?
Yes, if you are about to spend months or savings on an untested idea. The examples are mostly software companies and the vocabulary has turned into jargon, but the habit of testing your riskiest assumption cheaply before you build is worth learning even if you are not in software.
Who should read The Lean Startup?
Someone planning a business, an app or a side project who is afraid of spending months and savings on something nobody wants. Also useful for people inside larger companies who are asked to launch something new.
How long does this summary take to read?
About 6 minutes. It covers the book's 6 main ideas, who it is for, what to try this week, and an honest view of its weak points.
Read it offline, or listen to it. This summary is in the free MyByteNest app for Android, in English and Urdu, with 267 others. No sign-up.
Get the app
Share your thoughts
Found a mistake, disagree with the verdict, or have a book to recommend? Email us at support@mybytenest.com — we read everything.