The Lean Startup

How Today's Entrepreneurs Use Continuous Innovation to Create Radically Successful Businesses

A method for testing a new idea cheaply before you bet money on it, useful for any owner launching something uncertain.

Eric Ries is an American entrepreneur and writer who helped found a technology company before turning to teaching and writing about how new ventures are built. The Lean Startup, published in 2011, grew out of his own experience of building products that did not succeed at first, and from his study of how manufacturers learned to cut waste. It made the phrase lean startup common in boardrooms around the world.

The book's central claim is that a startup is a human institution designed to create something new under conditions of great uncertainty. Because the future cannot be forecast, the right approach is not to write a long plan and execute it, but to run a series of small experiments and learn from each.

This matters well beyond young technology firms. An owner of an established business faces the same uncertainty every time they launch a new service, enter a new city, try a new customer segment or open a new channel. Ries calls the people who do this entrepreneurs, whether they work in a garage or inside a large company.

The ideas are simple to state and hard to practise, because they ask us to give up the comfort of planning. We have organised this summary by ideas, not by the order of the book, and we say at the end where it fits less well. The examples with Indian owners are made-up illustrations, not real cases.

A new venture is a set of guesses

The first shift Ries asks for is in how you see a new idea. Most owners see it as a plan to be carried out. He sees it as a collection of untested beliefs, each of which could be wrong. You believe a certain customer has a certain problem, that they will pay a certain amount, that they will find you in a certain way and that you can serve them at a certain cost.

Until those beliefs are tested, he says, you are speculating. And the main danger in speculating is that you can spend a great deal of money and time building something at scale before finding out that a basic belief was false.

For an owner, this is a freeing idea. It means that being unsure is normal and that the right response to uncertainty is to find the cheapest way to reduce it. You do not need to be certain before you start. You need to be clear about what you are assuming.

Ries pays special attention to the assumptions that everything else depends on. If these are false, the whole idea collapses, however good the rest of the plan. He calls them leaps of faith. Two matter most. Does the product actually deliver value to the customer who is supposed to use it? And will the business be able to grow by finding more of those customers?

Imagine a Nagpur restaurant owner who wants to launch a cloud kitchen serving office workers. The plan includes a rented space, a team and a delivery tie-up. But the leaps of faith are plain. Will nearby offices order lunch from a new brand twice a week? Will they pay a price that leaves a margin? These can be tested with a small menu, a borrowed kitchen and a few weeks of orders before any lease is signed.

A good habit is to write down, before you start any new initiative, the three beliefs that must be true for it to work. Next to each, note how you could test it in under a month and for a small amount of money. You will often find that the test is simpler than you thought.

This habit also changes conversations in your team. Instead of arguing about whose opinion is right, you can ask what evidence would settle the question. That tends to lower the temperature of meetings and speed up decisions.

There is one more benefit to naming your assumptions, which is that it protects you from your own enthusiasm. Founders are good at seeing reasons why an idea will work and poor at seeing why it might not. Writing the beliefs down, and asking a sceptical colleague to challenge them, brings the weak points into view while they are still cheap to examine.

Build, measure, learn

The heart of the method is a loop. You build something small, you measure how real customers respond, and you learn whether your guess was right. Then you repeat, using what you learned to decide what to build next.

Ries stresses that the aim of the loop is learning, not output. A team that ships many features without learning anything is moving fast but not making progress. A team that ships little but learns a lot about what customers want is far ahead.

He suggests planning the loop in reverse. Start with what you need to learn, decide what you would need to measure to learn it, and only then decide what to build. Many teams do the opposite, building first and hoping the lessons will appear later.

For owners this is a powerful discipline. Before spending on a new product, ask what exactly you want to find out, and what result would change your mind. If no result could change your mind, you are not testing, you are only confirming a decision already made.

Speed matters here, but it is the speed of the loop and not the speed of any single step. The faster you can go around, the more you learn per rupee. This is why Ries argues for small, quick experiments over large launches.

Imagine a Pune garment maker thinking of a new line of work uniforms for hospitals. Instead of producing a large range, the owner shows samples to five hospital procurement managers, takes note of their reactions and asks for a small trial order from two. The loop takes a month, and it reveals that hospitals care more about fabric washing durability than about colour choices.

Notice what the owner learned. It was not about her factory or her designs. It was about the buyer, and that is exactly the kind of knowledge a plan on paper cannot give you.

A caution from the book is that a loop run without honest measurement is theatre. You must decide in advance what success looks like, so that you do not read every result as a win. Write the expected outcome down before you start.

Talk to real customers while you run the loop. Ries is firm that you should get out of the building and meet the people you hope to serve, instead of relying on reports. A conversation will show you things no figure can, such as the words customers use, what they tried before and what they were afraid of.

Keep these conversations open. Ask about their last experience with the problem, not whether they would like your idea. People are polite about ideas and honest about their own past behaviour.

The smallest thing that teaches you

Ries introduces the idea of a minimum viable product, which is the smallest version of an offer that lets you complete a turn of the loop with the least effort. The word minimum is important, and so is the word viable. It is not a poor product, but a quick route to learning.

He says that the purpose is to test a hypothesis about customers, not to impress them. The product might be a simple prototype, a manual service done by hand, a landing page describing something that does not yet exist or a short video explaining the idea. The form matters less than what it teaches you.

Owners often resist this, because their reputation is at stake and they dislike offering something rough. Ries suggests that it is better to find out early, with a small group, that you are building the wrong thing than to find out after a grand launch. Customers who volunteer for an early version are usually forgiving, and they give honest feedback.

An important point is that building anything beyond what is needed to learn is waste, even if it feels productive. Extra features, polish and scale add cost and delay, and they make you more attached to the idea.

Imagine a Bengaluru accountancy firm wanting to offer a monthly compliance subscription to small shop owners. Instead of building software, the partners offer it by hand to ten clients using spreadsheets and phone calls. They learn which tasks clients actually want help with and which they will pay for, long before any tool is built.

The risk is confusing minimum with careless. Your first version should still be honest, safe and respectful to customers. In regulated or high-stakes fields, such as health or finance, the minimum has to include whatever is needed to protect people.

A good question to ask before you build anything is what the smallest thing is that we could do to find out whether anyone wants this. Then ask whether it can be done in a week. Answers to those questions often cut months of planned work.

There is also a lesson about patience. A small test that shows weak interest is not a failure of the whole idea. It may mean the wrong audience, the wrong message or the wrong price. Change one thing at a time and test again, so that you know what made the difference.

Owners who learn this stop treating each experiment as a verdict on their judgement. They treat it as a question asked cheaply, and the answer, whichever way it goes, as an asset.

Validated learning, not vanity

Ries uses the phrase validated learning for knowledge gained by testing a specific guess with real customers and seeing real evidence. It is a more demanding standard than opinion or enthusiasm. A compliment from a friend is not validation. A customer paying, returning or recommending is.

He warns against measuring progress by activity. Counting hours worked, features built, meetings held or money spent tells you about effort, not about whether you are moving toward a business that works.

He also warns about what he calls vanity metrics. These are figures that look good and tend to rise, such as total registered users, total page views or cumulative sales. They give a comforting sense of growth even when nothing is improving, because a cumulative number can only go up.

The better numbers, in his term actionable metrics, tell you whether a change you made had an effect. They compare groups, look at behaviour over time and make it possible to say what caused what. A good test is whether a number tells you what to do next.

For owners, this applies to every dashboard you have. Look at the figures you review each week. Which of them would fall if your business were getting worse? Which ones can only rise? If most are of the second kind, you may be reassuring yourself rather than managing.

Imagine a Jaipur online handicraft seller who celebrates that total customers served has passed a thousand. A closer look shows that the number of customers who order twice has not moved in a year. The cumulative figure hid the real problem, which is repeat purchase.

Ries recommends looking at groups of customers by when they joined and seeing how each group behaves over time. If newer groups behave better than older ones, your changes are working. If not, they are not, no matter how big the total looks.

He also suggests comparing two versions of an offer shown to similar customers at the same time, so you can see which does better. This is a simple, honest form of testing that any business with enough customers can do.

Be careful about one habit here. Teams under pressure often find a measure that makes a project look good and quietly adopt it. To avoid this, fix the measures before the experiment begins and share them with someone outside the project. The aim is learning, not a flattering report.

If the numbers are poor, say so early and plainly. Your team will follow your lead, and an owner who reacts to bad news with questions rather than anger gets better information next time.

Accounting for innovation

Because new ventures cannot be judged by ordinary financial targets in the early days, Ries proposes a way of tracking progress that he calls innovation accounting. The idea is to measure learning in a disciplined way, in three steps.

First, you establish a baseline, using your earliest product to find out where you stand on the numbers that matter. Second, you try to improve those numbers through a series of experiments, each tuned to a particular figure. Third, you decide whether the numbers are moving toward the level a viable business would need. If they are not, it is time to change direction.

The merit of this approach is that it brings honesty into a stage that is usually run on hope. You say in advance what you expect, you check what happened and you hold yourself to the result. It also protects teams from the slow drift of projects that never succeed or fail but simply continue.

For an owner, the use is clear. Give every new initiative a small set of numbers and a review date. Agree beforehand what level means keep going and what level means stop or change. Without these, new projects tend to become permanent costs.

Imagine a Surat textile trader who starts selling directly to consumers through a website. He sets a target of a small number of repeat buyers within four months at an acceptable cost of acquiring each one. At the review, repeat buyers are well under the target and costs are above. He decides to change the approach, not to double the spending.

It also helps to separate learning milestones from financial ones. A new venture may have little revenue for a while, yet it can show clear progress in what you know. Report on both to your team and your investors or family partners.

This is a quiet but important discipline. It makes it possible to be patient with real progress and impatient with false progress.

Change direction or keep going

At regular intervals, Ries says, a team must face the question of whether to stay on its present course or to make a significant change. He calls the change a pivot. It is a structured shift in strategy, based on what you have learned, while keeping what you know to be true.

A pivot can take several forms. You might serve a different customer than you first imagined, solve a different problem for the same customer, focus on one feature that people actually use, change how you charge or change the channel through which you reach buyers. In each case, the vision stays and the route changes.

The alternative to a pivot is to persevere, which is the right choice when the evidence shows you are making progress. The hard part, says Ries, is telling which situation you are in. That is why he stresses having numbers and review points, and why he suggests holding a meeting for this purpose at regular intervals.

For owners, there is great value in making the decision deliberately. Many businesses linger in a state of drift, neither committing nor changing. A scheduled review gives you permission to decide, and to do so on evidence rather than on mood.

Imagine a Coimbatore engineering firm that started making a small machine for farmers. After two seasons, sales are weak among farmers but keep coming from small food processors who use it differently. The owner pivots to serving processors, adjusts the design and finds demand. The original idea was not abandoned, only redirected.

Pivots are painful because they feel like admitting failure. Ries argues that this is the wrong way to see them. They are the result of learning, and the earlier you do them, the less money and time you lose.

The key is to avoid two traps. One is changing direction at every setback, which prevents learning. The other is clinging to an idea because of the effort already spent. Evidence and a pre-agreed review point protect you from both.

It also helps to write down what you will keep, whatever happens. A pivot does not erase your skills, your relationships with customers or your knowledge of the market. Those carry over. Naming them makes a change of direction feel like a step forward and less like a loss.

Growth that matches your business

Ries also writes about how a young business grows, and describes three main engines. The first is the sticky engine, in which customers stay and keep using the product, so growth comes from keeping more than you lose. The second is the viral engine, in which customers bring others as a natural result of using the product. The third is the paid engine, in which you spend money to win customers and grow as long as each one returns more than they cost.

He says a business should usually focus on one engine at a time, and measure the figures that belong to it. Each has its own key number. For stickiness it is how many customers stay. For virality it is how many new customers each existing customer brings. For paid growth it is the gap between what a customer is worth and what it costs to win them.

This is a clarifying lens for owners of established businesses. Ask which engine actually drives your growth. A clinic that grows by referrals has a viral-like engine. A subscription service that keeps most of its members has a sticky one. A company that advertises heavily has a paid one. Knowing which helps you see where to invest.

Imagine a Nashik coaching institute that grows mainly by word of mouth among parents. The owner assumes more advertising is needed. But the numbers show that each satisfied family brings in more than one new family. Investing in results tracking and parent communication would strengthen the engine that already works.

Another benefit is that the framework tells you what to watch. If your paid engine depends on a cost per customer that keeps rising, you may soon be unable to grow profitably. If your sticky engine begins to leak, no amount of new sales will make up for it.

The point is to understand how your growth really happens, and to avoid fooling yourself with a story about how you would like it to happen.

Small steps and fast learning

Another theme is working in small batches. Ries borrows from manufacturing the idea that moving small amounts of work through each stage reveals problems sooner and costs less to fix. Large batches hide mistakes until the end, when they are expensive.

He applies this to product development, marketing and even decision-making. Release small changes often, observe the effects and correct course quickly. This also reduces the stress of each release, since no single step is so large that failure is a disaster.

A related practice is to look for the root cause when something goes wrong, instead of blaming whoever was nearest. Ries describes asking why repeatedly, until you reach a cause that is about the process, not the person. He suggests making a small, proportionate investment to fix each root cause, so that the same problem does not return.

For owners, this means treating mistakes as information about the system. If a customer received the wrong item, ask why, and then ask why again, until you find the weak point in your process. Fix that, and the problem often disappears for good.

Imagine a Chennai logistics company where deliveries are often late. The first answer is that drivers are careless. The next question reveals that loading is delayed. The next reveals that orders arrive at the depot late in the evening. The real cause is order cut-off timing, which can be changed in an afternoon.

This approach is also kinder to your team. People are more willing to report problems when they expect to be asked about the process rather than blamed. That brings issues to light early, which is when they cost least.

Small steps also suit the way most owners must work, with limited cash and limited attention. You do not need to bet the company on a single launch. You can make a series of modest, reversible moves and let the evidence guide you.

Innovation inside an established company

Although the book is aimed at startups, Ries argues that the same method works inside large and established firms. A company with a profitable core faces the problem that its systems are designed to execute known plans, not to explore the unknown. New ideas are often judged by the same standards as the core business, and so they are killed or forced to look good too soon.

He suggests giving innovation teams a protected space, clear limits on what they can spend, a small group with the freedom to run experiments and a way of judging them that reflects learning. The core business supplies resources, and the new team supplies a means of finding future growth.

For an owner of a business with revenue above Rs 5 crore, this is a very practical idea. You probably have a core that works and demands your attention. A new line of work will compete for that attention and will usually lose, unless you set aside time, money and people for it on purpose.

Imagine a Vadodara family-owned chemicals distributor that wants to explore selling directly to small manufacturers online. The existing team is busy and sceptical. The owner assigns one capable person, a small monthly budget and a quarterly review with specific learning goals. Over a year the experiment either grows into a line or ends without harming the main business.

It is important to protect the experiment from the demands of the core. The same measures that suit a mature division, such as monthly profit, will make a new idea look poor. Give it its own measures, and be clear about when those measures will start to resemble the rest.

This is also a way of keeping the organisation curious. A business that runs a few small experiments at any time learns about its customers and market in a way that planning meetings never achieve.

Where to push back

The Lean Startup was written mostly from the world of software, where releasing a change costs very little and results arrive within days. Many Indian owners work in physical businesses such as manufacturing, construction, food, retail and healthcare, where a test costs real money and takes real time. The principle still holds, but the speed of the loop is much slower and the minimum product is less minimal.

Some fields also carry safety, legal or reputational duties that limit how rough a first version can be. A hospital, a school or a food company cannot treat customers as test subjects. The method has to be adapted with care in such cases.

The approach works best when you have enough customers or prospects to learn from. A business that sells a few large contracts a year cannot run quick split tests. It needs different forms of evidence, such as conversations with buyers, pilots with a single client and careful reading of why deals are won or lost.

There is a risk of using the language to justify poor work. Calling an unfinished product a minimum viable product, or a failed launch a pivot, can excuse a lack of discipline. The method asks for more rigour, not less.

The book is also strongest on the early stage and thinner on what happens when you scale, manage people or deal with funding and culture. Owners at a later stage will find some of it too basic and will need other sources.

Finally, the book's examples are drawn from a mainly Western technology context, and some are told in a way that cannot easily be checked. Take the principles and test them in your own market. Check claims about other companies' success against independent sources before relying on them.

Try this week

First, pick one new idea you have been thinking about, such as a new service, product, city or customer group. Write down the three beliefs that must be true for it to work. Mark which of the three, if wrong, would end the idea.

Second, design the cheapest possible test of the riskiest belief. It might be five conversations with potential buyers, a one-page offer sent to twenty past customers or a hand-delivered pilot for one client. Set a date within three weeks and write down what result would make you continue.

Third, review the five figures you look at most often in your business. For each, ask whether it can only go up, and whether it tells you what to do next. Replace at least one with a figure that compares groups, such as repeat orders by month of first purchase.

Fourth, choose one problem that recurred in the last month. Ask why it happened five times in a row, each time answering about the process and not the person. Make one small change to fix the last cause you reach.

Fifth, if you have an existing experiment that has run for months without a decision, put a review on the calendar. Agree beforehand what result means continue, change or stop.

If these ideas are useful, the original book goes much deeper, with more cases and detail than we can give here. It is worth reading in full.