The Lean Startup, in bytes

The idea

Ries defines a startup as a human institution designed to create something new under extreme uncertainty. That includes teams inside large companies.

In real life

A new product team in a big firm is a startup in this sense.

The idea

Startups need management, but a different kind, built for uncertainty rather than for executing known plans.

In real life

Detailed five-year plans rarely survive contact with real customers.

The idea

The real progress of a startup is learning what customers actually want, proven with evidence, not just opinions.

In real life

Ries describes how his company IMVU learned customers didn't want a feature it had worked hard to build.

The idea

Treat each product idea as an experiment to test a hypothesis.

In real life

Zappos's founder tested demand by photographing shoes in local stores and posting them online before holding any stock.

The idea

Every plan rests on two big assumptions: the value hypothesis (does it deliver value?) and the growth hypothesis (how will it spread?).

In real life

Assuming finance teams will adopt a new tool because it saves time; testing whether they actually do.

The idea

Founders should talk to real customers early, rather than relying on assumptions in the office.

In real life

Visiting users at work reveals problems no survey would.

The idea

An MVP is the simplest version that lets you start learning with the least effort. It isn't the smallest product, but the fastest route to learning.

In real life

Dropbox tested demand with a demo video before the product was finished, and sign-ups surged.

The idea

Sometimes the best MVP is doing the service by hand for a few customers.

In real life

A meal-planning startup served its first customers personally before automating anything.

The idea

The core loop: build something, measure how customers respond, learn, and repeat. Speed through the loop is the key advantage.

In real life

Releasing small updates and studying usage each week.

The idea

Startups need a way to measure progress before revenue: set a baseline, tune the engine, then decide whether to pivot or persevere.

In real life

Tracking whether each change improves customer retention.

The idea

Vanity metrics, like total sign-ups, look good but don't guide decisions. Actionable metrics show cause and effect.

In real life

Rising total users can hide the fact that few new users stay.

The idea

Look at groups of customers who joined at the same time, to see whether changes actually improve behaviour.

In real life

Comparing this month's new users with last month's shows whether the product is improving.

The idea

Test two versions with different groups to learn which works better.

In real life

Testing two onboarding emails to see which leads more users to complete setup.

The idea

When experiments show the strategy isn't working, change course while keeping what you've learned. Ries describes many types of pivot.

In real life

A product built for one customer group finds its real market in another.

The idea

A startup's real runway is how many pivots it can still make, not just how many months of money it has.

In real life

Faster learning means more chances to find what works.

The idea

Small batches reveal problems sooner and allow faster learning. Ries's example: stuffing envelopes one at a time can be faster than doing each step in bulk.

In real life

Releasing small updates often instead of one big release a year.

The idea

Ries describes three engines of growth: sticky (keeping customers), viral (users bringing others) and paid (buying customers profitably).

In real life

A subscription business focuses on retention; a social app on referrals.

The idea

When something goes wrong, ask “why?” five times to find the root cause, and invest in a fix proportional to the problem.

In real life

A failed release traced back to missing training for a new team member.

The idea

Large companies can innovate by giving teams a protected space with clear limits, their own metrics and freedom to experiment.

In real life

A small team allowed to test new ideas without disrupting the core business.

The idea

The lean startup aims to cut the biggest waste of all: building things nobody wants.

In real life

Testing demand before building saves months of wasted effort.