What does this Lean Startup thing have to do with Lean? What can a startup learn from an established giant like Toyota, or vice versa? Where is my value-stream map, I need a value-stream map, right? The ground between established Lean practice and the Lean Startup movement is full of confusion, but things are far from hopeless – it is possible to form a relatively clear picture of this whole, and that is what this post is all about.

Lean is a value system

All the usual tools aside, at its core Lean is an application of the Confucian ideal of seeking perfection in a business environment. To get to the core of Lean it is vital to think beyond value-stream mapping, A3 reports, and even the PDSA cycle, as all of those are merely means to an end. The core of Lean consists of three principles:

Create value for the customer: Customer determines what is valuable. If the customer does not appreciate (and thus, by extension, be willing to pay for) some feature, that feature is irrelevant. A company should strive to create maximal value for its customers.

Eliminate waste: All activities that are not value-adding, are waste. Even parts of activities that are value-adding may be waste, if there is a better way to create the valuable outcome. A company should strive to eliminate waste in all of its activities.

Respect people: All human beings are improvable and perfectible through personal and communal endeavor. Continuous improvement towards perfection (which Toyota calls the True North) is a desirable goal in all things, even though full perfection can never be achieved. Therefore, all people, all employees, deserve the opportunity to improve themselves and their work and full support from the organization in doing so.

Each of these three principles are needed to find perfection: customer value for perfection of the product, waste elimination for perfection of the system, and respect for people for perfection of the individual.

Scope of the Lean Startup

In the Lean Startup: How Today’s Entrepreneurs Use Continuous Innovation to Create Radically Successful Businesses, Eric Ries sets out to answer the question how can companies, large and small, overcome the hurdles to repeatedly building successful products and remaining innovative.

It is important to note that Lean Startup is actually not about startups at all, even though all startups are within its scope.

Ries defines a startup as follows: “A startup is a human institution designed to create a new product or service under conditions of extreme uncertainty.”

This definition does not say anything about the age or size of a company. It consists of three parts:

• A startup is a human institution (such as a company).
• A startup creates a new product or service.
• A startup operates under conditions of extreme uncertainty. This is of particular importance, because hidden in this part of the definition is the premise that customer needs are not discoverable without experimentation. If this premise is not accepted, the whole Lean Startup thinking collapses.

This is the context in which Ries applies Lean thinking.

The five principles of Lean Startup

According to Ries, there are five principles in Lean Startup, on which everything in it is built:

• Entrepreneurs are everywhere. Consistent with the definition of a startup, any company can be a startup.
• Entrepreneurship is management. A startup is an institution that needs to be managed. This management needs to place special emphasis on the context of extreme uncertainty.
• Validated learning. Startups exist to learn how to build a sustainable business. This learning has to be validated through experimentation.
• Build-Measure-Learn. A startup turns ideas into products, measures how customers respond, and learns whether to pivot or to persevere. Rapid execution of this feedback loop is crucial for success.
• Innovation accounting. The follow-up of startups should focus on untraditional metrics that show whether the startup is moving towards a sustainable business or not.

The first principle is actually just defining the scope of the Lean Startup, and not the content. In this sense, it is not similar to the other principles.

The other principles fall quite neatly under the Lean value system:

• Entrepreneurship is management is about eliminating waste. Without management, it is not possible to create the basic stability that is the basis of improvement.
• Validated learning, Build-Measure-Learn, and innovation accounting are about eliminating waste (by validating which things are unnecessary so as not to spend time on them) and creating value for the customer (by validating what the customer actually appreciates).

What about respect for people? None of the principles of the Lean Startup include respect for people! Respect for people can be interpreted as being part of several principles: it is definitely an integral part of the management system at Toyota (yet it is not explicitly part of the entrepreneurship is management) and validating what is valuable for the customer is part of respecting the customer.

However, regardless of the principles, respect for people is built in throughout the more detailed descriptions of what the Lean Startup is about.

As an example, let’s take a look at Ries’s listing of the prerequisites for nurturing disruptive innovation:

• Scarce but secure resources. Disruptive innovation is not built by legions, but by small groups of individuals. However, these individuals need to stick with the venture in order to succeed.
• Independent development authority. Without authority to experiment and pivot, disruptive innovation is not possible.
• A personal stake in the outcome. With authority comes responsibility. In a startup, this is tangible, in a corporation, perhaps less so, but even if there is no strong financial pressure, the pressure on reputation can be enough.

These three prerequisites describe fairly completely the respect for people approach in any company that practices kaizen in a traditional Lean environment: people have the authority to improve their environment themselves, but the resources are always limited, which forces them to come up with innovative solutions instead of easy but suboptimal ones.

The Build-Measure-Learn loop and the PDSA cycle

One thing that I found rather troubling on the surface is the Build-Measure-Learn loop and its relationship with the PDSA cycle as an application of the scientific method, which is at the core of Lean right after the value level. Both Build-Measure-Learn and PDSA are supposed to be the application of the scientific method, and if this is so, what are the differences in the premises that lead the Lean Startup to choose Build-Measure-Learn rather than PDSA?

However, looking deeper into Build-Measure-Learn, it becomes apparent that it is actually the PDSA cycle applied under extreme uncertainty: because there are so many unknowable things, it is crucially important to build a minimum viable product to test the hypotheses as fast as possible, and then reiterate through the loop as quickly as possible. In PDSA language, the phases look like this:

Plan. In the Plan phase, the startup determines the minimum viable product (or products) it is going to build. Whatever it is, it has to test one or more of the hypotheses that are behind the business idea. The two most important hypotheses are the value hypothesis (does the product deliver value to a customer) and growth hypothesis (how can the product conquer new markets).

It is important to plan the metrics in a way that clearly shows whether progress is being made: most products succeed in some small way, and accumulative metrics could show progress (vanity metrics) whereas cohort metrics could show that no real improvement is taking place (actionable metrics). Without a clear view on whether performance is improving, it is easy to stick with a plan that is ultimately unsustainable.

Do. In the Do phase, the minimum viable product that enables testing the set hypotheses is built. It does not even have to be a real product, except for the fact that it is already tested by customers. Zappo’s started as a website with pictures taken from inventories of brick-and-mortar shoe shops and bought any pairs of shoes sold online for full price from the shops in question. That was the minimum viable product that enabled testing of the business model – that people are willing to buy shoes online.

Study. In the Study phase, the results of the minimum viable product are gathered and analyzed with a focus on actionable metrics that show whether the product is improving.

Adjust. In the Adjust phase, a decision is made whether to persevere (carry on with the same goals) or pivot (change some aspect of the product strategy). This is a vital step, as it can be difficult to change course and possibly abandon large parts of an existing product, but it is often necessary, because as more and more information about the market becomes available, the relative importance of various features changes.

On compatibility and completeness

Based on the analysis above, the ground between Lean proper and Lean Startup is becoming more clear. At the very least, Lean Startup is fully compatible with traditional Lean in the context of the empirical premises it makes. However, there are many things that are compatible with Lean, and can be used by Lean enterprises and non-Lean enterprises alike, but that does not yet make them Lean as such.

Here we come to the real difficulty: the principles of Lean Startup have not been completely described, and parts of the Lean Startup practice do not follow from the explicit principles. The actual practice is even closer to Lean than the explicit principles indicate.

Therefore, the Lean Startup needs further development, and this development can either take it towards Lean or away from it. It has already become significant enough for traditional Lean practitioners to take part in developing it further, for examples see this blog post by Jim Womack and this video of a presentation by two Toyota product development engineers who started applying Lean Startup at Toyota.

So how Lean is the Lean Startup? It is Lean enough to fit into the toolkit of a Lean enterprise, but not yet Lean enough to make an enterprise Lean by itself without further development.

Picture: Lean Startup Machine by John Fischer (CC)