The Lean Startup: How Today's Entrepreneurs Use Continuous Innovation to Create Radically Successful Businesses

The first chapter introduces the idea that entrepreneurship and management need to be interchangeably practiced in a startup. The “just do it” (15) approach of entrepreneurship fails because it lacks discipline, checks, and balances. US Bureau of Labor Statistics confirm that manufacturing capability has increased because modern management and technology have made manufacturers more productive. The Lean Startup movement develops a coherent management paradigm for new business ventures to reduce the loss of valuable time, skills, and capacity.

The Lean Startup method uses the principles of lean manufacturing, developed by Taiichi Ohno and Shigeo Shingo of Toyota: reducing batch size, controlling just-in-time production and inventory, and accelerating cycle times, all the while enhancing the knowledge and creativity of individual workers. This comprehensive theory of entrepreneurship addresses the functions of an early-stage venture:

• Vision and concept
• Product development
• Marketing and sales
• Scaling up
• Partnerships and distribution
• Structure and organizational design (19)

Instead of organizing a company into functional departments, individuals are held accountable through learning milestones (see Chapter 7: “Measure”). Rather than measure a day’s success by the number of tangible items completed, learning milestones focus on the true goal of a startup, building the right thing “as quickly as possible” (20).

Henry Ford is an example of a successful entrepreneur who implemented management strategies to increase production. Before he was a CEO, Ford experimented with tuning and refining engines. He learned while he worked how the ignition, mechanics, and engine functioned in a precisely timed feedback loop. Here, the automobile is a metaphor for the startup as an institution. The “engine of growth” (21) relies on the development of new products, new features, and new marketing strategies. Ries also compares the relationship between the driver and the steering wheel to a CEO and a startup. The moves required when driving to work can be completed automatically, but writing these moves down in detail is nearly impossible. Similarly, too many startups try to build their product and business strategy like it’s “a rocket ship” (21). Startups should forget about planning, investing, and executing a plan based on the assumptions of success and customer appeal. Instead, they should adjust with a “steering wheel” called the Build-Measure-Learn feedback loop. Ries presents the image of a pyramid with product on top, strategy in the middle, and vision at the base. Two principles of change—optimization and pivot—occur between each level. These terms foreshadow their definitions in subsequent chapters. Lastly, the chapter recalls the importance of management in entrepreneurship. Conventional wisdom on management claims that failure is a result of inadequate planning or inefficient execution. Ries challenges this notion by asserting that new product development in the modern economy “routinely requires exactly this kind of failure on the way to greatness” (24).

The second chapter defines what an entrepreneur is, what they should focus on, and what they should develop first. Ries begins with an anecdote about a manager, named Mark, who approached him for advice at one of Ries’s lectures. Ries gave Mark his standard advice: create innovation inside big companies, leverage cool, new product development technologies, and navigate corporate policies, personnel, and processes to be more efficient. But Mark interrupted Ries. He already knew and had accomplished these steps but lacked the process for converting them into success. This anecdote introduces the idea of “intrapreneurs,” innovators inside of established organizations who can use the Lean Startup method to create new, successful products.

Ries then defines exactly what a startup is: “A startup is a human institution designed to create a new product or service under conditions of extreme uncertainty” (27). Ries emphasizes the “human” aspect of this definition. Keywords, like institution, product, and innovation all need the human quality to bring this into successful form. Traditional management theories do not account for the human element and thus face difficulties navigating the terrain of “extreme uncertainty.” The Lean Startup method embraces the uncertainty, a point that Ries proves using SnapTax as an example.

SnapTax was a startup in 2009 that wanted to automate the process of collecting information from W-2 forms. After initial setbacks with technological availability and useability, the end goal became an effort to create a product that allowed consumers to file their entire tax return on their phone. Instead of waiting to develop the perfect product, the SnapTax team delivered an early version that could do much less: it only worked for consumers with a simple return and who lived in California. The early version of the product allowed users to take a picture of their W-2 forms with their phone’s camera which would then automatically fill in the form for them. This, according to Ries, was an amazing innovation that eventually grew into a successful startup story.

SnapTax did not reflect the conventional image of startup founders living in a garage and eating ramen noodles. Instead, SnapTax started under the auspices of Intuit, America’s largest producer of tax, financing, and accounting tools. Usually, big corporations like Intuit are good at “sustaining innovation,” in the words of The Innovator’s Dilemma author Clayton Christensen, but not “disrupting innovation” (31). What allowed SnapTax to succeed within a larger corporation was an incrementally implemented management discipline that cultivated “intrapreneurial” practices within the company.

Before SnapTax, Ries introduced the definition of the Lean Startup to Intuit founder Scott Cook. Cook immediately took to the idea and began molding Intuit into an “innovation factory” (34). Cook and Intuit’s Chief Executive Officer, Brad Smith, began directing employees to run experiments with their product development. Employees were encouraged to study, experiment, implement and learn from their mistakes while developing novel products. Ries argues that the experimentation system by Intuit helped SnapTax succeed despite its modest beginnings.

Chapter 3 describes Ries’s personal experience as an entrepreneur learning the Lean Startup method. The most vital function of an organization is validated learning: “the process of demonstrating empirically that a team has discovered valuable truths about a startup’s present and future business prospects” (38).

Ries illustrates an example of validated learning by referring to his tech company IMVU. Specifically, Ries identifies how his failure as a co-founder caused him to realize the key tenets of validated learning. IMVU wanted to enter the instant messaging market, which at the time was controlled by a few high-profile companies like AOL, Microsoft, and Yahoo!. IMVU faced a huge hurdle trying to compete with these companies to bring a new IM network to market. IMVU decided to build a three-dimensional videogame-like add-on product that could interoperate with other existing networks, thus saving the time and money of marketing an entirely new IM network. IMVU’s strategy was that customers would adopt the technology, learn the user interface, and invite their friends to join.

Ries gave his team six months to launch the add-on product. Initially, Ries was worried that the fast development would lower the quality of the product. He and his team were aware of significant flaws in the product design and implementation, but they also knew they had to stay on schedule. However, when Ries and his team launched the product, no one even tried it. Despite continuing to make the product better, the enrollment numbers never rose. Ries decided that change was needed, so he and his team started to conduct qualitative research by interviewing customers.

Customer after customer said the same thing; they were reluctant to invite friends from their other networks for fear of the IMVU product would be unpopular. Ries and his team continued to develop features after listening to customers. Eventually, Ries realized that his entire model for the product was outdated and not appealing to the intended customer base. So, Ries was forced to throw out “thousands of lines of code” and six months of work (46).

Ries realized that the months of work gathering information only to eventually throw out the product could have been collected in a more efficient manner. This is a tenet of lean manufacturing: “[l]earning to see waste and then systematically eliminate it” (47-8). Ries learned to differentiate waste from value in product development. Instead of spending time debating and prioritizing features that customers would never discover, IMVU could have shipped sooner and learned firsthand what appealed to customers and what did not. Validated learning means collecting empirical data from manufacturing small batches and adjusting during product development.

For example, in one instance of a validated learning experiment, IMVU changed its “avatar chat” feature to “3D instant messaging.” Half of customers were directed to the old version; and half to the new one. Ries and his team were able to measure the differences in behavior between each group. They discovered that customers in the new, experimental group were more likely to become paying customers over the long term. In another example, IMVU changed their avatar technology from that of their competitors. Instead of using the real-life identities of their customers to build avatars, IMVU developed avatar technology that ensured the safety and identity of their customer base was secure. In sum, the hundreds of experiments that IMVU conducted systematically figured out “the right things to build” (52) through validated learning.

Ries recommends that companies don’t wait to release their products. The “audacity of zero” (52-53) is when companies hold onto the myth that if no product has been released and no data has been collected, then overnight success is possible. Ries strongly advises that companies avoid this mentality and instead incrementally release products while collecting data on product efficacy.

Ries tells the audience that he uses the story of IMVU when lecturing at business schools. Students, he says, often focus only on the tactics the story illustrates: “launching a low-quality early prototype, charging customers from day one, and using low-volume revenue targets as a way to drive accountability” (54). Although these are effective techniques, they are not the intended takeaways. The scientific method informs the Lean Startup method because it understands that every product is an experiment in validated learning rather than the result of specific tactics.

Chapter 4 discusses how failures during product development leads to learning opportunities that can help startups succeed. The Lean Startup philosophy is based in the scientific method, and the objective for a startup is to determine the best way to build sustainability in accordance with the company mission. Startup experimentation works towards the former goal using lean manufacturing principles.

An example is Zappos founder, Nick Swinmurn, and his willingness to experiment with his product early on in his company’s existence. Swinmurn began by asking local shoe stores if he could photograph their inventories. In exchange for their permission, he would upload the photos online and then return to buy the products at full price if a customer contacted him. Swinmurn’s strategy asked a simple question to determine his product: “[I]s there already sufficient demand for a superior online shopping experience for shoes?” (57). Instead of relying on market research or surveys, Swinmurn started with a single and simple product to test his product’s success. He gained accurate data about customer demand, put his product in contact with real customers, and discovered otherwise unknown and unexpected information from his sales. He used the data from this experiment to actualize the company’s vision: to sell shoes to customers through a convenient and online format.

An intrapreneurial example is Caroline Barlerin, a director of global social innovation at Hewlett-Packard (HP), who developed a company volunteer program for employees. Barlerin’s program envisions employees using four hours of company time each month to volunteer for their communities. However, most employees do not know about the volunteer program, and most of the actual volunteering that occurs involves manual labor, not the highly trained skillsets of the employees. This company initiative acts as a startup within a large corporation, however Barlerin uses traditional management practice and theory to investigate this problem, whereas Ries recommends she use lean startup methods. Ries encourages Barlerin to start testing certain assumptions, like a company-wide focus on short-term profitability and a general avoidance of volunteering because these opportunities do not use employee skillsets. Instead of strategically planning to address these concerns, they can test the assumptions through experimentation.

The first step in the Lean Startup model is to break down the vision of the startup into component parts. This means identifying the value hypothesis and the growth hypothesis. The value hypothesis tests whether a product or service delivers value to customers. The growth hypothesis tests how new customers will discover the product or service. For Barlerin’s case, both hypotheses can be analyzed by conducting an observational study of a small group of employees who volunteer and then examine their retention rates. Then, the value and growth hypotheses can be determined by examining the number of employees who remain in the volunteering program and asking them why they perceive value in the activity. Then, Ries recommends that Barlerin creates a concierge minimum viable product, which would provide the employees who willingly volunteered with the best volunteer experience possible and then measuring what those employees accomplished. Additional experiments like this can be elaborated so that specific data on the volunteer experiment is collected. Using this method, Barlerin would be able to determine the early adopters of her product, the motivation behind volunteering, and improvements necessary for the program. If glaring issues arise, then Ries recommends a strategic pivot (explained in Chapter 8).

In the Lean Startup method, an experiment “is more than just a theoretical inquiry; it is also a first product” (63). Strategic planning and market research can be conducted alongside these experiments to continuously collect data and adjust. Ries provides three more examples, like Barlerin’s, to explain this method.

When Kodak wanted to develop a new, online “event album” feature for their customers, they built a simple prototype and allowed customers to use it. They discovered that none of the early customers were able to create an event album because it lacked useability and essential features. With this data from the first experiment, Kodak developed features according to the customer complaints. They scrapped all theoretical features that they thought customers would want and only focused on the tangible feedback from the prototype. The next beta launch taught the Kodak team that customers also wanted to be able to arrange the order of their pictures before inviting other users to contribute. So, the team redesigned the product again, repeating this process until they developed an adequate product.

The next example is the Village Laundry Service in India. Akshay Mehra, a Proctor & Gamble Singapore employee, decided to make a laundry business available for people in India who could not typically afford this type of service. He and the Village Laundry Services (VLS) began conducting experiments with their customers with prototypes of the service. The experiment used a pickup truck to take laundry from people, clean it offsite, and return it to customers for payment by the end of the day. The team learned that customers distrusted the company because it was believed that the pickup truck could drive away and steal the clothes. So, in the next experiment, the team decided to redesign the truck to wash clothes on-site. They also discovered that customers would pay double to have their laundry back in four hours, as opposed to twenty hours. In the end, VLS and Mehra launched a “three-foot by four-foot mobile kiosk that included an energy-efficient, consumer-grade washing machine, a dryer, and an extra-long extension cord” that would clean clothes with a quick turnaround for customers (68).

The last example is about a government startup during President Barack Obama’s administration. The Obama administration created a new federal agency, the Consumer Federal Protection Bureau (CFPB) that would protect American citizens from predatory lending by financial companies. Ries suggested the following to create this government startup: “treat the CFPB as an experiment, identify the elements of the plan that assumptions rather than facts, and figure out ways to test them” (69). First, the agency experimented with an automated hotline for consumer claims and questions. With this inexpensive product, the agency immediately learned the problems and questions of American citizens. They discovered the motivations behind consumers’ calls, the percentage of people in the target areas, and the baseline behavior that could provide a foundation for market research (71). This early service allowed the CFPB to growth exponentially in its early years, while helping millions of American citizens with their financial problems.