FROM ALCHEMY TO SCIENCE

 

The Lean Startup methodology reconceives a startup’s efforts as experiments that test its strategy to see which parts are brilliant and which are crazy. A true experiment follows the scientific method. It begins with a clear hypothesis that makes predictions about what is supposed to happen. It then tests those predictions empirically. Just as scientific experimentation is informed by theory, startup experimentation is guided by the startup’s vision. The goal of every startup experiment is to discover how to build a sustainable business around that vision.

Think Big, Start Small

 

Zappos is the world’s largest online shoe store, with annual gross sales in excess of $1 billion. It is known as one of the most successful, customer-friendly e-commerce businesses in the world, but it did not start that way.

Founder Nick Swinmurn was frustrated because there was no central online site with a great selection of shoes. He envisioned a new and superior retail experience. Swinmurn could have waited a long time, insisting on testing his complete vision complete with warehouses, distribution partners, and the promise of significant sales. Many early e-commerce pioneers did just that, including infamous dot-com failures such as Webvan and Pets.com.

广告：个人专属 VPN，独立 IP，无限流量，多机房切换，还可以屏蔽广告和恶意软件，每月最低仅 5 美元

Instead, he started by running an experiment. His hypothesis was that customers were ready and willing to buy shoes online. To test it, he began by asking local shoe stores if he could take pictures of their inventory. In exchange for permission to take the pictures, he would post the pictures online and come back to buy the shoes at full price if a customer bought them online.

Zappos began with a tiny, simple product. It was designed to answer one question above all: is there already sufficient demand for a superior online shopping experience for shoes? However, a well-designed startup experiment like the one Zappos began with does more than test a single aspect of a business plan. In the course of testing this first assumption, many other assumptions were tested as well. To sell the shoes, Zappos had to interact with customers: taking payment, handling returns, and dealing with customer support. This is decidedly different from market research. If Zappos had relied on existing market research or conducted a survey, it could have asked what customers thought they wanted. By building a product instead, albeit a simple one, the company learned much more:

1. It had more accurate data about customer demand because it was observing real customer behavior, not asking hypothetical questions.

2. It put itself in a position to interact with real customers and learn about their needs. For example, the business plan might call for discounted pricing, but how are customer perceptions of the product affected by the discounting strategy?

3. It allowed itself to be surprised when customers behaved in unexpected ways, revealing information Zappos might not have known to ask about. For example, what if customers returned the shoes?

 

Zappos’ initial experiment provided a clear, quantifiable outcome: either a sufficient number of customers would buy the shoes or they would not. It also put the company in a position to observe, interact with, and learn from real customers and partners. This qualitative learning is a necessary companion to quantitative testing. Although the early efforts were decidedly small-scale, that did not prevent the huge Zappos vision from being realized. In fact, in 2009 Zappos was acquired by the e-commerce giant Amazon.com for a reported $1.2 billion.2

For Long-Term Change, Experiment Immediately

 

Caroline Barlerin is a director in the global social innovation division at Hewlett-Packard (HP), a multinational company with more than three hundred thousand employees and more than $100 billion in annual sales. Caroline, who leads global community involvement, is a social entrepreneur working to get more of HP’s employees to take advantage of the company’s policy on volunteering.

Corporate guidelines encourage every employee to spend up to four hours a month of company time volunteering in his or her community; that volunteer work could take the form of any philanthropic effort: painting fences, building houses, or even using pro bono or work-based skills outside the company. Encouraging the latter type of volunteering was Caroline’s priority. Because of its talent and values, HP’s combined workforce has the potential to have a monumental positive impact. A designer could help a nonprofit with a new website design. A team of engineers could wire a school for Internet access.

Caroline’s project is just beginning, and most employees do not know that this volunteering policy exists, and only a tiny fraction take advantage of it. Most of the volunteering has been of the low-impact variety, involving manual labor, even when the volunteers were highly trained experts. Barlerin’s vision is to take the hundreds of thousands of employees in the company and transform them into a force for social good.

This is the kind of corporate initiative undertaken every day at companies around the world. It doesn’t look like a startup by the conventional definition or what we see in the movies. On the surface it seems to be suited to traditional management and planning. However, I hope the discussion in Chapter 2 has prompted you to be a little suspicious. Here’s how we might analyze this project using the Lean Startup framework.

Caroline’s project faces extreme uncertainty: there had never been a volunteer campaign of this magnitude at HP before. How confident should she be that she knows the real reasons people aren’t volunteering? Most important, how much does she really know about how to change the behavior of hundreds of thousand people in more than 170 countries? Barlerin’s goal is to inspire her colleagues to make the world a better place. Looked at that way, her plan seems full of untested assumptions—and a lot of vision.

In accordance with traditional management practices, Barlerin is spending time planning, getting buy-in from various departments and other managers, and preparing a road map of initiatives for the first eighteen months of her project. She also has a strong accountability framework with metrics for the impact her project should have on the company over the next four years. Like many entrepreneurs, she has a business plan that lays out her intentions nicely. Yet despite all that work, she is—so far—creating one-off wins and no closer to knowing if her vision will be able to scale.

One assumption, for example, might be that the company’s long-standing values included a commitment to improving the community but that recent economic trouble had resulted in an increased companywide strategic focus on short-term profitability. Perhaps longtime employees would feel a desire to reaffirm their values of giving back to the community by volunteering. A second assumption could be that they would find it more satisfying and therefore more sustainable to use their actual workplace skills in a volunteer capacity, which would have a greater impact on behalf of the organizations to which they donated their time. Also lurking within Caroline’s plans are many practical assumptions about employees’ willingness to take the time to volunteer, their level of commitment and desire, and the way to best reach them with her message.

The Lean Startup model offers a way to test these hypotheses rigorously, immediately, and thoroughly. Strategic planning takes months to complete; these experiments could begin immediately. By starting small, Caroline could prevent a tremendous amount of waste down the road without compromising her overall vision. Here’s what it might look like if Caroline were to treat her project as an experiment.

Break It Down

 

The first step would be to break down the grand vision into its component parts. The two most important assumptions entrepreneurs make are what I call the value hypothesis and the growth hypothesis.

The value hypothesis tests whether a product or service really delivers value to customers once they are using it. What’s a good indicator that employees find donating their time valuable? We could survey them to get their opinion, but that would not be very accurate because most people have a hard time assessing their feelings objectively.

Experiments provide a more accurate gauge. What could we see in real time that would serve as a proxy for the value participants were gaining from volunteering? We could find opportunities for a small number of employees to volunteer and then look at the retention rate of those employees. How many of them sign up to volunteer again? When an employee voluntarily invests their time and attention in this program, that is a strong indicator that they find it valuable.

For the growth hypothesis, which tests how new customers will discover a product or service, we can do a similar analysis. Once the program is up and running, how will it spread among the employees, from initial early adopters to mass adoption throughout the company? A likely way this program could expand is through viral growth. If that is true, the most important thing to measure is behavior: would the early participants actively spread the word to other employees?

In this case, a simple experiment would involve taking a very small number—a dozen, perhaps—of existing long-term employees and providing an exceptional volunteer opportunity for them. Because Caroline’s hypothesis was that employees would be motivated by their desire to live up to HP’s historical commitment to community service, the experiment would target employees who felt the greatest sense of disconnect between their daily routine and the company’s expressed values. The point is not to find the average customer but to find early adopters: the customers who feel the need for the product most acutely. Those customers tend to be more forgiving of mistakes and are especially eager to give feedback.

Next, using a technique I call the concierge minimum viable product (described in detail in Chapter 6), Caroline could make sure the first few participants had an experience that was as good as she could make it, completely aligned with her vision. Unlike in a focus group, her goal would be to measure what the customers actually did. For example, how many of the first volunteers actually complete their volunteer assignments? How many volunteer a second time? How many are willing to recruit a colleague to participate in a subsequent volunteer activity?

Additional experiments can expand on this early feedback and learning. For example, if the growth model requires that a certain percentage of participants share their experiences with colleagues and encourage their participation, the degree to which that takes place can be tested even with a very small sample of people. If ten people complete the first experiment, how many do we expect to volunteer again? If they are asked to recruit a colleague, how many do we expect will do so? Remember that these are supposed to be the kinds of early adopters with the most to gain from the program.

Put another way, what if all ten early adopters decline to volunteer again? That would be a highly significant—and very negative—result. If the numbers from such early experiments don’t look promising, there is clearly a problem with the strategy. That doesn’t mean it’s time to give up; on the contrary, it means it’s time to get some immediate qualitative feedback about how to improve the program. Here’s where this kind of experimentation has an advantage over traditional market research. We don’t have to commission a survey or find new people to interview. We already have a cohort of people to talk to as well as knowledge about their actual behavior: the participants in the initial experiment.

This entire experiment could be conducted in a matter of weeks, less than one-tenth the time of the traditional strategic planning process. Also, it can happen in parallel with strategic planning while the plan is still being formulated. Even when experiments produce a negative result, those failures prove instructive and can influence the strategy. For example, what if no volunteers can be found who are experiencing the conflict of values within the organization that was such an important assumption in the business plan? If so, congratulations: it’s time to pivot (a concept that is explored in more detail in Chapter 8).3