Lean startup is a methodology for developing business introduced by Eric Ries and presented in the book The Lean Startup.
Eric Ries developed the lean startup methodology following his different experiences as an employee, consultant and company founder. His first company did not succeed because he failed to understand the needs of the targeted customers and spend to much time on the first version of his product. Then, as an employee of a startup up that spend 5 years in stealth mode, he assisted at the failure of the There.inc as the company did not succeed in gaining early adopters. His analysis was that the company vision was “too concrete”. Therefore, the company developed a product that could not match customer demand.
Lean Startup Methodology
Lean startup is inspired by lean manufacturing, a methodology developed and applied first in the Toyota production system.
The key principles of the lean startup are:
- Minimum valuable product
- Split testing and actionable metrics
Minimum Valuable Product
A Minimum Valuable Product (MVP) is the version of a new product which allows collecting the maximum amount of validated learning about customers with the least effort. A frequently seen illustration of MVP is the opposition of the following processes for a company creating a car.
In the classical process, the car manufacturer delivers a full car by:
- Creating the frame
- Adding wheels
- Adding the car body
- Delivering the final product to the customer
Proceeding this way implies a large investment in R&D and production before customer feedback and knowing if the product can find its market. In the MVP approach, a startup would:
- Build a skateboard
- Make a scooter
- Build a motorbike
- Finally, build the car
The advantage of the MVP approach is customer involvement at each step of the process. This allows the startup to learn what is the customer expectation at a lower cost and even start customer acquisition before product finalization.
Split testing and actionable metrics
Split testing is an experiment in which variations of a product are offered at customers at the same time. The goal of this test is to observe the differences in customer behaviour between groups. Thanks to the results of the test, it is possible to improve each version of the product.
Actionable metrics are enabling to make informed business decisions and actions. These metrics shall be well selected to enable actions, in particular metrics linked to business revenue. An example of an actionable metric for a web-based business would be the number of paid subscriptions. This metric would be opposed to a non-actionable metric of the total number of visitors if they do not create value to the company.
During the process of learning threw tests, a startup can discover that its product is not in line with client needs. However, it can also discover another need that is not addressed or that present a better business opportunity. Then, in that case, the startup can change the product to match this new need. This change of model or product is a “pivot”.
Build/Learn/Measure has to be a continuous learning process in the startup. Threw rapid iterations of the MVP, this learning loop should enhance startup product or market fit. The phases of the cycle are the following ones:
- Start with ideas
- Build a version of the product
- Confront this version to the market
- Measure the market response
- Make use of the data to learn what has to is to change or to improve within the product
This process should be continuously maintained in the startup to ensure the best performances of the startup.
The lean startup is a very interesting book that can provide solutions to some startup problems. However, people either praise or reject the book and its methodology. For Enhanced innovation, the lean startup methodology is relevant, but of course, you still have to adapt it to your company.
The MVP approach applies very well when the market knowledge is not good, funding is low and fast iterations are easy.
- Market/product maturity: MVP applies very well to innovative ideas and to not addressed markets. If you are planning to enter an existing market by duplicating an existing model, the MVP approach will not apply as is (but there is also the risk that your company does not succeed either). However, the MVP approach can apply to any new product development in an established company or to test the relevance of product enhancement.
- Low funding: the application of the MVP is relevant even with large funding, but provides a good mean to reduce spendings due to limited resources. If founding is larger, it is possible to develop more achieved MVP or with a higher frequency to gain more knowledge. In the case of extensive funding and a market with a lot of potential concurrent, it can also be interesting to have the opportunity to be the first on the market by a very fast MVP cycle.
- Fast iterations: if the product is based on software, it is much easier to perform fast iterations than for a hardware-based product. Therefore, the perimeter MVP is important to maximize the ROI on gained knowledge.
The build/measure/learn methodology is a classical process. All companies, not only startups, should apply this methodology to maintain an advance on the competitors.
In conclusion, the lean startup is not a miracle method to ensure that a startup will not fail (of course, if such a method existed, everyone would apply it). Yet, it provides some clues and questions to that a startup or company should ask to succeed. Moreover, it can apply to any project within an existing and/or established company.
Want to learn more or make your own opinon? Buy the book from on the the following link!