handbooksThe five principles
Entrepreneurs are everywhere, entrepreneurship is management, validated learning, innovation accounting, and build-measure-learn. What each one changes about how you work.
- Five principles sit under everything else. Skip them and the tools become theatre.
- Two are about who does this work. Three are about how progress gets measured.
- The hardest one is validated learning, because it means treating your plan as a set of guesses.
The five
| Principle | What it means |
|---|---|
| Entrepreneurs are everywhere | This is a skill set, not a personality. It can be learned, and it applies inside big companies too |
| Entrepreneurship is management | A startup needs managing, not just building. The scarce skill is deciding what not to build |
| Validated learning | Progress is what you have proved with real customers, not what you have shipped |
| Build, measure, learn | The loop. Turn ideas into products, measure the response, decide whether to pivot |
| Innovation accounting | Boring but necessary: measure progress, set learning milestones, hold people to them |
1. Entrepreneurs are everywhere
You do not need a garage or a funding round. Anyone working on something new under real uncertainty is doing this job, including a team of three inside a large company.
2. Entrepreneurship is management
The instinct in a new product is to focus on the technical work. Ries argues the opposite. The building is rarely the constraint. Deciding what to build, and having the discipline to build only that, is the constraint.
A startup is an institution, so it needs management. Just not the kind that produces a two-year roadmap.
3. Validated learning
Every startup plan rests on guesses. Validated learning means turning those guesses into experiments and letting real customer behavior settle them.
The test is simple: what do you now know that you did not know before, and what evidence backs it? "We shipped four features" is not an answer.
It is easy to call a failed release "a learning experience" afterward. That is not validated learning. Validated learning means you decided in advance what the experiment would prove, and it proved it.
4. Build, measure, learn
The loop that connects the other four. Each pass should end with a decision, not just a release.
5. Innovation accounting
You cannot use normal financial measures on a product with no revenue yet, so a startup needs its own way to show progress. Innovation accounting covers how.
Ask your team what they learned in the last month that changed a decision. If the answers are all features and dates, you are running a delivery process, not a discovery one.