Lean startup

What is the lean startup method?

The lean startup is a way of working under uncertainty. Eric Ries popularized it in his 2011 book of the same name, drawing on Steve Blank's customer development work and on lean manufacturing, where waste means any effort that does not add value for the customer. In a startup, the largest waste is building something nobody wants.

The core claim is that a new product is a set of untested guesses, not a plan to execute. You are guessing who has the problem, how badly they feel it, whether your solution fixes it, and whether they will pay. The job of the early company is to test those guesses in the cheapest way that still produces a trustworthy answer.

That reframes progress. Under the old model, progress was hitting the milestones on a plan. Under this one, progress is validated learning: you know something about your customer today that you did not know last month, and you learned it from behavior rather than opinion.

The build, measure, learn loop

The method runs as a loop. You build the smallest thing that can test the riskiest assumption, measure how real users respond, and decide what the result means. Then you either continue in the same direction or change course.

The order in practice runs backwards from the name. Decide what you need to learn, choose the metric that would prove or disprove it, then build only what generates that metric. A landing page with a real signup form, a manual service behind a simple front end, or a spreadsheet you run by hand for ten customers all count. Speed through the loop beats polish on any single pass.

Ries also pushed innovation accounting: measuring conversion rates and cohort behavior rather than totals, since totals always rise and say nothing about whether the last change helped.

Why lean startup matters for startups

Time and cash are what kill small teams. Building for nine months before showing anyone burns both against an untested guess. Running the loop in two week cycles means you can be wrong five times and still have runway left.

It also settles arguments. When two founders disagree about a feature, the lean answer is not to debate harder, it is to define the smallest test that would change one of your minds, run it, and go with the data.

Lean startup in practice

Say you plan a scheduling tool for physical therapy clinics. The riskiest assumption is not that you can build software, it is that clinic owners will switch from phone booking. So instead of building, you set up a one page site describing the product with a "request early access" form, and you personally call the first fifteen signups.

Eleven say the real pain is not booking, it is patients who no-show. You build a stripped down reminder tool in three weeks, charge $29 a month from day one, and watch whether clinics keep paying past month two. Nine of the first twelve renew. That is a validated assumption bought with a month instead of a year.

Lean startup and the lean canvas

These get conflated often. The lean startup is the methodology: the loop, the experiments, the decision to pivot or persevere. The lean canvas is a one page template that captures your current assumptions about problem, customer, and revenue. The canvas is where you write the guesses down. The method is how you test them. Filling out a canvas and then building for a year is not lean, just a tidier version of the old approach.

Common mistakes

• Treating "lean" as "cheap." The goal is to reduce waste, not to underinvest. Sometimes the right experiment costs real money.
• Shipping a broken product and calling it minimal. A minimum viable product narrows scope, not quality. If the small thing works badly, you learn nothing about demand.
• Asking instead of observing. People say they would pay. Payment is the only reliable proof, which is why customer validation should involve money or a signed commitment.
• Endless pivoting. Changing direction after every disappointing week means no assumption ever gets a fair test. Define in advance what result would justify a pivot.
• Measuring totals. Cumulative signups always rise. Cohort retention tells you whether anything improved.

The method works best when the loop is tight and the feedback is real, so pair it with a functioning customer feedback loop and be honest that the finish line is product-market fit, not a completed backlog.