The lean startup
Innovating for success
Description
The myth is that a lone entrepreneur can develop a great product through sheer grit, and it will automatically succeed. But most startups fail because they don't validate customer demand. To succeed, manage your startup differently - accelerate learning, not planning. Build a minimal prototype and sell it to early adopters.
Get their feedback and iterate daily to improve it. Keep looping build-measure-learn until you have a product customers love. If you focus on validated learning with extremely fast cycle times, taking a scientific approach to decisions - the core of the Lean Startup method - you stand a better chance of startup success. Essence: focus on customer needs, rapid iteration, and data-based decisions.
Table of contents
01The lean methodology - purpose
The conventional approach to launching a startup typically begins with someone envisioning a successful business that creates products consumers love. To achieve that vision, they develop a strategy: defining the business model, product lineup, partners, competitors, and target customers. Then they build and sell the product per that strategy. With this model, entrepreneurs are seen as determined, focused people who guide startups to market and drive adoption. Setbacks are mere learning opportunities on the path to success.
In reality, entrepreneurship demands general management skills more than anything. Startups juggle multiple concurrent activities: acquiring new customers, supporting existing ones, innovating, tuning marketing, reevaluating strategy, and optimizing the product. Most emphasize immediate results over learning. They aim simply to build and launch products, considering it successful if they hit targets and come in under budget. Thus, many accidentally build something no one wants. Their financial performance matters little if the product doesn’t meet customer needs.
A startup’s primary job is to quickly and cost-effectively identify what customers will pay for. Defining startups may be helpful here. A reasonable definition is: a human institution designed to create a new product or service under conditions of extreme uncertainty. Some key implications:
Startups are institutions with processes for hiring, coordination, product development, and more. Successful ones master these processes. Startups focus on creating innovations, whether via new technology, repurposing existing technology, or developing new business models that unlock value. Uncertainty is guaranteed, rendering most traditional management tools ineffective. Startups instead need tools suited to operating amidst uncertainty.
02The lean methodology - guide
A startup serves as a catalyst to transform ideas into products. As customers interact with these products, data and feedback are generated to facilitate learning.
In other words, a quality startup provides answers to four key questions:
1. First, do consumers recognize the problem the startup aims to solve? 2. Second, if a solution already exists, do consumers experience enough frustration with the status quo that they would purchase an alternative? 3. Third, would consumers specifically buy the startup's proposed solution? 4. Finally, can the startup build its solution profitably and sustainably over time?
The Build-Measure-Learn loop offers a systematic methodology to address these questions. This framework applies the scientific method to startups.
The first step entails identifying the core hypotheses requiring validation. For most startups, two "leap-of-faith" assumptions prove most critical: The value creation hypothesis questions whether the startup's offering actually adds value for target customers as envisioned. The growth hypothesis considers whether the startup can profitably deliver its solution at scale.
These leap-of-faith assumptions represent make-or-break issues for startups; all other factors flow from correctly optimizing these two hypotheses. Each iteration of the Build-Measure-Learn loop focuses on enhancing the accuracy of the value creation and growth hypotheses.
Toyota's famed lean manufacturing system advocates genchi genbutsu or "go and see for yourself" as the optimal mechanism for testing assumptions. Directly observing real customers enables startups to validate whether target users actually face the problem in question.
For example, Toyota manager Yuji Yokoya oversaw revamping the Sienna minivan for the 2004 model year. To determine required upgrades, he embarked on a 53,000 mile trip across all 50 U.S. states and 13 Canadian provinces to interact firsthand with Sienna owners. From these conversations, Yokoya learned kids exert substantial influence on minivan purchases and incorporated child-friendly enhancements into the redesigned vehicle. The 2004 Sienna proved extremely popular, with sales expanding 60% over the 2003 version.
Similarly, when conceiving software startup Intuit in 1982, founder Scott Cook hypothesized personal computers would someday facilitate routine financial tasks like bill payment. He tested this assumption by cold calling Palo Alto, California and Winnetka, Illinois residents to gauge their bill management frustrations. Only after confirming the widespread desire for simplification did Cook commence work on Intuit's accounting software products.
Two common obstacles impede startups from directly testing their leap-of-faith assumptions: impatience and overanalysis. Some entrepreneurs rush ahead to product development rather than diligently vetting their underlying hypotheses. At the other extreme, other founders endlessly plan without progressing to real-world validation. The Build-Measure-Learn framework mitigates both pitfalls by emphasizing quick creation of a minimum viable product (MVP).
03The lean methodology - speed up
In theory, the Lean Startup methodology seems slow, clinical, and simple. However, in reality, startups face decisions that are rarely clear-cut, with circumstances changing rapidly.
Startups often cycle through the Build-Measure-Learn loop at a quick pace before making decisions on the fly. So how does Lean Startup come together in practice?
The methodology works best when you accelerate through the loop, while keeping batch sizes small. As you grow, you adapt processes without sacrificing speed. And you constantly innovate new products and features.
Taking each phase in turn, Lean Startup shines when applied to small rather than large batches. The goal is to test hypotheses through validated learning, not optimize efficiency. If a hypothesis proves wrong, you don’t want the mistake to be so costly that it destroys the company. The approach focuses on learning to build a sustainable business, not merely produce more stuff. If customers don’t want what you’re building, it’s better to find out now rather than after investing millions in finished goods that sit unused in warehouses.
The smaller the batch, the faster you cycle through the feedback loop. If you can loop faster than competitors, you will learn more. With today’s rapid prototyping tools, you can get product iterations in front of potential users quickly to gather feedback. Small batches allow you to state a hypothesis, test the idea on a small scale, and move on, avoiding excessive cost. You piece together an experiment, gain the needed learning, then iterate. This discovery process is inexpensive and fast.
Related books
