I remembered a specific moment from my very first startup. It was the moment I realized my company was going to fail. My cofounder and I were at our wits’ end. The dot-com bubble had crashed, and we had spent all of our money. We were trying desperately to raise more, and we could not. The scene was perfect: it was raining, we were arguing in the street. We literally couldn’t agree on where to walk next, and so we parted, in anger, heading in opposite directions. (...)
It remains a painful memory. We had begun as friends, and ended as enemies. The company limped along for months after, but our situation was hopeless. Looking back, I know our failure was inevitable, because we had no clue. It seemed we were doing everything right: we had a great product, a brilliant team, amazing technology, and the right idea at the right time. (...) But despite a promising idea, we were nonetheless doomed from day one, because we did not know the process we would need to use to turn our product insights into a great company.
When I first read about the concept of the Lean Startup in Eric's blog, it was a fragmented set of ideas, case studies and practices – most of them useful, many of them thought-provoking. Little did I know that it would take this collection of thoughts only a few years to turn into a global movement and a coherent theory of entrepreneurship.
And while the orignal story of the Lean Startup yet
remains to be written, let me try to give a practioner's overview of both the theoretical foundations of the Lean Startup as well as the specific tools and techniques it employs to achieve maximum efficiency. The
Lean Startup in a Nutshell series will walk you through each of the different building blocks of the Lean Startup, and hopefully turn you into a Lean Startup expert in no time.
None of the ideas presented in this series are original ones. They have all been extracted from the works of
Eric Ries,
Steve Blank, and other evangelists of the Lean Startup movement.
The startup as an experiment
A startup is
not a small version of a larger company. It is en entirely different kind of organization, driven by different goals and different needs. In other words: The
dollhouse theory of startups is wrong. Neither does a startup need the same departments – engineering, marketing, QA, finance, support, etc. –, nor should it follow the same product development methodology as its hypothetical larger counterpart.
While an established business focuses on executing and scaling a proven model, a startup is essentially an experiment. It is a human institution designed to create something new under conditions of extreme uncertainty – meaning that both the solution to the supposed problem as well as the problem itself are unknown. The purpose of the startup is to find a way of transforming the vision of its founders into a working and sustainable business model before it runs out of money.
The product development methodologies
Every organization strives to solve a problem. In an established business, the problem is well defined. For such a business, both the waterfall and the agile product development methodology may be appropriate, depending on whether the solution to the problem is known in advance or not.
In a startup, all that is certain is the vision of the founders. A startup is thus an organization in search of the right problem to solve. It must develop partial solutions to hypothetical problems while constantly gathering feedback until it is able to figure out what customers really want. The right methodology for a startup is the build-measure-learn feedback loop.
Let us review the three different product development methodologies.
- The waterfall methodology (known problem, known solution)specifies a linear plan of product or service milestones. It goes from requirements to design to implementation to testing. In the waterfall model, progress is achieved by advancing the plan. Using this definition of progress, whether product development results in success or failure entirely depends on whether the solution to the problem is accurately known in advance and whether the problem is actually a problem customers care about.
- The agile development methodology (known problem, unknown solution) is an iterative and incremental approach to software development. It emphasises frequent interactions over extensive documentation and the ability to respond to changes in specification over the linearity of advancing a pre-negotiated plan. It is perfectly suitable for situations where it is known what customers ultimately want, but where the team is unable to predict the best way to get there due to a lack of past experience solving that problem. Progress in the agile methodology is measured by lines of working code.
- The build-measure-learn feedback loop (unknown problem) is the underlying development methodology of the Lean Startup. In essence, it combines agile product development and customer development in a company-wide feedback loop of learning and discovery. In a Lean Startup, progress is measured by validated learning – by the number of full-turns through the entire feedback loop. It is important to understand that unless the whole feedback loop is completed quickly, work carried out in a single stage of the feedback loop (building, measuring, or learning) does not count as actual progress and is very probably a waste of time.
The remaining parts of the Lean Startup in a Nutshell series will discuss the specific techniques to accelerate each stage of the feedback loop in detail.
Pivots and innovation accounting
A startup consists of a
vision and a
strategy (a business model, i.e. a collection of hypotheses) designed to turn that vision into a real-world business that is creating sustainable value. What I particularly like about the Lean Startup framework is that within it, the vision is the only thing that remains untouched. It is thus in perfect alignment with Simon Sinek's
model of the power of purpose and the
built-to-last concept.
While the startup's vision functions as the guiding light, its strategy is temporary and fluid until it has been validated by subsequent iteration of the build-measure-learn feedback loop. In the Lean Startup model, a change in strategy is called a pivot, and it represents the most fundamental concept of the Lean Startup:
If you look at the real story, you'll discover this weird zig-zaggy path between the initial idea and the successful idea. (...) It's just that successful entrepreneurs, when they run into difficulty, they don't just give up and go home, but neither do they persevere the plane straight into the ground. They do this thing called the pivot. They change some elements of the business while keeping others constant. They keep a new strategy for achieving the same vision. So you pivot the strategy, you stay true to the vision.
And the premise of the Lean Startup is this: If we can reduce the time between pivots, we can increase our odds of success before we run out of money.
In order to decide whether to pivot or not after an iteration of the feedback loop, we use innovation accounting. Innovation accounting consists of formulating and testing a set of key metrics – quantitative assumptions – by working backwards through the feedback loop:
- What are we trying to learn next?
- What do we need to measure in order to learn that?
- What do we have to build to be able to measure that?
It is time to pivot whenever further iterations of the feedback loop lead to diminishing returns, whenever it becomes unlikely that the quantitative assumptions needed to validate the current strategy can be met at all.
Innovation accounting brings accountability to entrepreneurship: It enables us to know whether we are actually making progress and whether we are getting closer to realizing our vision or not. It empowers us to set goals and decide when to pivot in advance, and act accordingly once we collect the necessary validated learning.
The remaining parts of the Lean Startup in a Nutshell series will discuss each of the three stages of the feedback loop: Build, Measure, Learn.
The Lean Startup in a Nutshell series