Module 1: Lean Foundations

Build-Measure-Learn

Most new products are built the same way: someone has an idea, spends months and savings building it, launches, and then discovers whether anyone wants it. When the answer is no, the money and time are gone. The lean startup method, described by Eric Ries in his book The Lean Startup, offers a different approach. Instead of betting everything on one big launch, you learn in small, fast cycles. The core of the method is the Build-Measure-Learn feedback loop.

The loop explained

The loop has three steps that repeat:

1. Build: create the smallest thing that lets you test an important idea with real people. This is your minimum viable product, or MVP. It might be a simple web page, a manual service, a sample batch or even a conversation script.
2. Measure: observe how real customers behave when they meet the MVP. Do they sign up, pay, return, recommend?
3. Learn: compare what happened with what you expected. Decide whether your idea is supported, needs changing, or should be dropped.

Then you start the loop again with your new understanding. The goal is not to build a product quickly. The goal is to learn quickly what customers really want, so that what you eventually build has a much better chance of succeeding.

Plan the loop backwards

Although the loop is described as build, measure, learn, you should plan it in reverse:

1. First decide what you need to learn. Which belief, if wrong, would sink the business?
2. Then decide what you need to measure to find out. What behaviour would prove or disprove that belief?
3. Only then decide what to build. What is the least you need to create to observe that behaviour?

Planning backwards stops you from building features that teach you nothing.

Worked example: a laundry pick-up service

James wants to start an app-based laundry pick-up service for busy professionals in Kilimani, Nairobi. His original plan was to spend several months and a large share of his savings building an app with booking, tracking and payments. Using the loop in reverse, he asks:

• Learn: Will busy professionals in Kilimani pay to have laundry collected and returned, and how often?
• Measure: The number of people who book and pay in the first month, and how many book a second time.
• Build: A WhatsApp number, a one-page flyer for apartment notice boards, a price list, and an arrangement with an existing laundromat. James does the pick-ups himself on a motorbike. Payment is by M-Pesa.

Within two weeks and a very small budget, he has real data. If people book and rebook, he has evidence to invest further. If they do not, he has saved months of building an app nobody would use. Either way, he has learned something true about his market.

What counts as learning

Ries calls the goal validated learning: knowledge backed by evidence from real customer behaviour, not opinions or guesses. "Friends said they love the idea" is not validated learning. "Twelve of forty households we approached paid for a first pick-up, and seven booked again within two weeks" is. Validated learning is measured in facts that change your decisions.

How fast should the loop run?

The faster you complete each loop, the more you learn before your money and energy run out. Early on, aim for loops measured in days or a few weeks, not months. If a test will take months, look for a smaller test that answers part of the question sooner. Speed matters because every loop reduces uncertainty, and a founder who learns ten things in three months is in a much stronger position than one who learns one thing in a year.

Common mistakes to avoid

• Building first and thinking later. Without a clear learning goal, you collect data that does not answer anything.
• Building too much. If your MVP takes months, it is probably not minimal.
• Measuring opinions instead of behaviour. What people do matters more than what they say.
• Never closing the loop. Running tests without deciding anything afterwards wastes the effort.