Lean Startup – Eric Ries
Eric Ries – Lean Startup, Kirjaessee / Aleksi Laakkonen
In the book Lean Startup (Ries, 2011), Eric Ries describes a model for running a startup business. The model has its roots in the lean production methods developed at Toyota Motor Corporation with its car manufacturing process. As with Toyota Production System, the main goal of The Lean Startup is to eliminate waste. With a startup the elimination of waste is achieved first and foremost by exposing the product-ideas of startups to the public as soon as possible. Instead of spending copious amounts of time and money developing the brainchild of either a single entrepreneur or a team in anticipation of a great launch when the product finally is ready, the goal is to test the idea from the beginning with real customers. If the business doesn’t interest any customers at this early stage, it should either pivot to another direction, or be terminated immediately. This way, time and money won’t be wasted developing a product that has no demand in the market.
At the core of Ries’s model is the build – measure – learn loop, which I will try to summarize below:
Build – in a nutshell: instead of developing a product for years and then launching it for the masses, you build the core experience for even a single customer, and start testing your idea in the real world from day one. This early version of the product is called the MVP (Minimum Viable Product). In an MVP you should include only the most essential functionalities of what you see as the final product. This version of the product is not and should not be perceived as serving your whole customer base, but rather should be targeted towards the “early adopter”. Early adopters are the people in your target segment that are interested in the latest and greatest and are willing to sacrifice stability and polish for the privilege of being there first. Usually the early adopter is also more committed to the product, thus willing to actively participate in its development by giving feedback and testing new features.
Measure – with the product now exposed to the public, you must start gathering data about people’s reaction towards it. Data from this experimental phase should be used to advise further development and steer your vision towards a product that has real demand among real customers. It is important to decide on what and how to measure before beginning this testing.
Learn – data gathered from the early release should be used to make decisions in developing the product further. By analyzing key metrics and customer feedback, one should be able to draw conclusions about what is working and what is not. Things learned should be implemented to the MVP for the next round of testing.
Actionable metrics – with data in such a critical role in the process of a lean startup, the quality of it is of huge importance. Such metrics should be chosen that are meaningful to the development and success of the final product. In the case of an online publication relying on ad revenue, it might be useful to measure gross visits on their website over time, but for a physical business, simple pageviews are of little importance. In the case of a software platform it is important to keep track of customer retention over time, and maybe the value of average purchases.
There are no universally viable metrics but rather they should be chosen based on what is the desired business model for the product or company.
Split testing – another key tool in developing a lean startup is the testing of new ideas in an effective way. The concept of split testing, or A/B-testing is rather simple. Two (or more) ideas are tested against each other within different groups of customers. Maybe you want to test a new pricing strategy to get less, more valuable customers? Send two different advertisements to your customers, half with the current price, half with the new. Now see if the new strategy resulted in more of what you wanted. It’s a simple but effective way to see if your idea has potential in it. With certain types of products, this kind of testing is possible to do even with multiple versions of the same product simultaneously.
The roots of Ries’s model are in software development, but that is not to say that the methods of lean startup cannot be implemented in other startups. When it comes to projects at Proakatemia, I see a lot of potential in using these methods to develop those dream-projects that seem take forever or are just too big to try the traditional way. A good example of a non-software use-case would be a restaurant-popup during the Finnish Restaurant Day to see if what you are planning to offer in your future restaurant really draws a crowd.