by Jeff Dyer
So Levchin and Thiel developed software that could do such beaming. This business idea caught the attention of some top venture-capital companies in Silicon Valley, leading to PayPal’s first round of financing at Buck’s of Woodside, a favorite restaurant for many venture capitalists. PayPal’s investors showed up with $4.5 million preloaded on a PalmPilot that they beamed to Levchin and Thiel’s PalmPilot. PayPal seemed to be on its way.
PayPal’s initial growth was rapid, but the market leveled off rather quickly because it was limited to the roughly 3 million handheld (PDA) users in the United States. It didn’t take long before Levchin and Thiel realized another problem with the business model. “The initial idea of beaming money between PalmPilots was basically a bad idea,” Thiel told us. “I mean, if you have to meet face to face to exchange money, which you had to with the PalmPilot idea, you could just hand the other guy a check. But in the course of building out this idea, we made midcourse changes that were really interesting.” These midcourse changes were prompted in part by customers who wanted to sync their PalmPilots to their computers and send money through the internet to someone else with a computer and PalmPilot. “We came up with the idea of attaching money to an email,” Thiel recalled. “Since there were 120 million email users in the United States, this made it much more viral. You didn’t have to meet face to face.”
Today, PayPal is the world’s largest processor of email payments, but this never would have happened if its founders hadn’t been willing to constantly experiment and launch earlier versions of the product. Just as the security-wallet experiment was a “failure,” the original PalmPilot experiment also fell on its face. But these critical experiments generated the data necessary for PayPal’s ultimate success.
The PayPal experience is not atypical for innovative entrepreneurs. They realize the importance of experimenting with prototypes and pilots to see what they can learn. Because of their bias for action, they tend to launch products or businesses as quickly as possible, almost as an experiment, to see what the market’s response will be. They like to throw new product, process, and business ideas against the wall to see what will stick. PayPal’s experiments were essentially launched as products to the market, and they generated important data when the products failed to gain traction.
While some innovators seem prone to quickly launch their prototypes directly to the market, others more carefully test and compare competing prototypes to see what works best. Jennifer Hyman and Jennifer Fleiss did this before launching Rent the Runway, a Netflix-type business model for renting designer dresses. During a trip home to New York City, Hyman noticed her sister, Becky— an accessories buyer at Bloomingdale’s—struggling over what to wear for an upcoming wedding. Her sister wanted something stunning, but even though she had a decent salary, every designer dress was too expensive and out of reach. As Hyman watched her sister agonize over what to do, she wondered, “If the Beckys of this world can’t wear a designer dress, what hope is there for the rest of us?” She also thought that designers had a problem as well. “If designers can’t get their pieces into the hands of young, fashionable women,” she thought, “they are going to have a difficult time building their brands.” Hyman’s simple observation of a common ritual (finding a dress for a special occasion) in a familiar place (home) with a familiar person (her sister) produced an uncommon insight. Why not modify the Netflix mail-order business model and apply it to high-end fashion? Instead of purchasing designer dresses, women could rent the designer dresses online for that special occasion, for only one-tenth the cost.
So Hyman and Fleiss set up some experiments to test their idea. They bought a hundred dresses from designers like Diane von Furstenberg, Calvin Klein, and Halston and ran three experiments. The first was on the Harvard University campus; they rented dresses to Harvard undergrads, letting young women try on the dresses first. The pilot was an unqualified success. Women not only rented the dresses but returned them in good condition. This experiment demonstrated that there was a market for renting dresses and that renters would return them in good shape. But would women rent dresses they couldn’t try on? To answer that question, they set up another experiment, this time on the Yale campus, allowing women to see the dresses before renting, but not allowing them to try them on. Although fewer women rented, the pilot proved successful. Finally, they took photos of dresses and ran a test in New York City where women rented a dress only from PDF photos and descriptions of how the dress would fit. This experiment would tell Hyman and Fleiss whether they could truly use a Netflix model of renting over the web, or whether they must open stores where women could see and try on dresses. The final experiment showed that roughly 5 percent of women looking for special-occasion dresses were willing to try the service, enough to demonstrate the viability of renting over the web. And that’s how Rent the Runway launched. It has proved very successful, with over six hundred thousand members and roughly fifty thousand clients trying the service in the first year. Trying different experiments was critical to designing a successful business model. As Hyman told us, “Our revenue growth is amazing. This is a dream come true.”
As we studied innovators and their experiments, one thing we noticed was that the amount of experimenting required to gain new insights was almost the inverse of the amount of prior questioning, observing, and networking they had done. In other words, if you haven’t done much questioning, observing, or networking (or haven’t done them well), then you will have to run more experiments to gain the insights required to move forward. For example, Rent the Runway’s experiments were able to be carefully crafted to generate the right data because of years of observations that Hyman, in particular, had made of the needs of young women attending special events. (Hyman had worked for years at Starwood Hotels, where she launched programs to meet the needs of wedding parties and honeymooners. She also worked at WeddingChannel and IMG, one of the world’s top firms for female models.) As a result, she had a deep knowledge of the needs of fashion-oriented young women, special events, and designers and designer clothing. This allowed her and Fleiss to design better experiments to test their ideas.
The bottom line is that if you ask salient questions, observe salient situations, and talk to more diverse people, you will likely need to run fewer experiments. And the experiments you do run will be better designed to generate the data you need to take the next step. Random experimentation occurs when you know very little from your questions, observations, and networking conversations.
In the end, we’ve learned that even when you’ve effectively questioned, observed, and networked, persistent experimentation is likely to be important for generating disruptive insights. Virtually every disruptive business that we studied evolved over time—through a series of experiments—into a business model that changed an industry. Some experiments were accidental. For example, Southwest Airlines cofounder Herb Kelleher told us that the original low-cost-airline entrant stumbled onto its quick-turnaround capability when financial pressures forced the company to service its routes with three planes instead of the four it had originally planned to use. It had to either cancel flights or figure out a way to fly a four-plane schedule with three planes. This led management to develop a new set of practices for turning planes around as quickly as possible, eventually leading to a fifteen-minute plane turnaround. This innovation completely changed Southwest’s strategy and business model, as well as its bottom line.
Similarly, IKEA never intended to have knockdown kit furniture (disassembled furniture in flat parcel boxes) as a central feature of its low-cost-furniture retailing model. A serendipitous experiment early on in the company’s history yielded an important insight. After completing a photo shoot for a furniture catalog, a marketing manager found not all the furniture fit back into the trucks. When a photographer suggested that they take the legs off the table and then slide the table into the truck, the lights went on: IKEA could knock down almost all its furniture to reduce shipping costs and make the custom
er the final assembler. This small experiment was critical to IKEA’s business model as a global furniture retailer.
Innovators engage in three types of experimenting to generate data and spark new insights: trying out new experiences, taking things apart, and testing ideas by creating prototypes and pilots. Although questioning, observing, and networking are excellent for providing data about the past and present, experimenting is the best technique for generating data on what might work in the future. In other words, it’s the best way to answer what-if questions. Innovators also understand that when you ask salient questions, observe salient situations, and talk to the right people, you will likely need to run fewer experiments. This reduces the cost and time associated with experimenting. Finally, innovators understand—and accept—that the majority of their experiments will not turn out as planned (and indeed may turn out to be a colossal waste of time), but they know that experimenting is often the only way to generate the data required to ultimately achieve success.
Tips for Developing Experimenting Skills
To strengthen your experimenting skills, you will need to consciously approach your work and life with a hypothesis-testing mind-set. We recommend the following activities to practice and strengthen your experimenting skills.
Tip #1: Cross physical borders
Visit (or even better, take up occupancy in) a new country or some other new environment, such as a different functional area within your company or a new company in a different industry. Acquire a passport mind-set to break free of common routines. Explore the world by engaging in new activities. Join new social or professional activities beyond your normal sphere, attend a lecture by someone whose work you’re unfamiliar with, or visit an unusual museum exhibit. When you try out these new activities, ask yourself questions to help produce new insights from the experience, such as: “If my work team were here, what could we learn from this experience that would lead us to do something new? If I were going to replicate one thing (product, process, and so on) from this environment in my everyday environment, what would it be?” Work to cross one border at least once every month.
Tip #2: Cross intellectual borders
Take out a new annual subscription to a newspaper, newsletter, or magazine from an entirely different context (or to help save trees, intentionally and regularly search the web for country, industry, or profession information about areas distant from your own). If you live in the United States or France, for instance, consider reading a publication from China, India, Russia, or Brazil. If you work in the oil and gas industry, read a publication from the hospitality industry. If you are trained in marketing, read a publication related to engineering or operations.
Tip #3: Develop a new skill
To gain new perspectives, create a plan to develop some new skills or acquire new knowledge. Look for opportunities in your community to take classes in acting or photography, or get some basic training in mechanics, electronics, or home building. Try out new physical activities like yoga, gymnastics, snowboarding, scuba diving, or even sky diving (if you are brave enough). Check out the menu of courses at your local university and sign up for classes that sound interesting to you, ranging from history to chemistry to calligraphy. Or closer to home, identify another function in your company, whether it be marketing, operations, or finance, and do a deep dive on how that function works in your company.
Tip #4: Disassemble a product
Look through your house for something that no longer works, or go to a junkyard or flea market to buy a few things that you can easily take apart. (This is especially fun to do with your kids.) Search for something that you’ve always been interested in but have never taken the time to explore. Set aside a block of time to take the objects apart piece by piece and search for new insights into how they were designed, engineered, and produced. Draw or write about your observations in a journal or notebook.
Tip #5: Build prototypes
Identify something that you would like to improve. What would it look like if you changed it? Build a prototype of your new, improved invention from random materials in your house or office, or go on a shopping spree to obtain odd things that might work well in the prototype. Play-Doh (the children’s modeling clay) is a great medium for creating prototypes. If you are feeling adventurous and want to splurge, you may even want to buy a three-dimensional printer that produces objects on demand (according to your design). Or hire an industrial designer to create a software blueprint of a product you design and get it 3-D printed at a local university.
Tip #6: Regularly pilot new ideas
Gordon Moore, the cofounder of Intel, once recalled that, “most of what I learned as an entrepreneur was by trial and error.” Engage in frequent pilot tests (small-scale experiments) to try out new ideas and to see what you learn from doing something differently than you’ve done before. You, too, can become an experimenter when you embrace learning through trial and error, but you must have the courage to fail and learn from your failures. Make up your mind to plan and carry out a pilot test of an idea you have at work during the next month.
Tip #7: Go trend spotting
Actively seek to identify emerging trends by reading books, articles, magazines, web links, blogs, and other sources that specifically focus on identifying new trends. Read material written by individuals you believe excel at identifying trends and seeing what’s next. Try reading the work of Kevin Kelly (executive editor of Wired and author of New Rules for the New Economy), Chris Anderson (former editor in chief of Wired and author of The Long Tail and Free), or another author who is looking into the future. Then think about how these trends might lead to an interesting experiment with regard to a new product or service. Figure out a way to creatively conduct that experiment.
PART TWO
The DNA of Disruptive Organizations and Teams
7
The DNA of the World’s Most Innovative Companies
“Fast-growth companies must keep innovating. Companies are like sharks. If they stop moving, they die.”
—Marc Benioff, founder and co-CEO, Salesforce
IN THIS BOOK’S first six chapters, we described how innovative people think differently and act differently to generate creative ideas for new products, services, processes, and businesses. Now we shift our attention to answering the following question: how do companies comprising many people build the code for innovation? Without a doubt, executives worldwide face this critical question as they try to build innovation capabilities within their companies to generate growth opportunities. Before addressing this question, though, let’s look at two other equally important ones. First, which companies are truly the most innovative and should serve as models of innovation? Second, does having an innovation capability (and a reputation for innovation) turbocharge a company’s market value?
In 2005, Businessweek began creating a list of the world’s one hundred most innovative companies. It based this list on a Boston Consulting Group survey of executives who voted on the companies. (See table 7-1 for the Businessweek top twenty-five innovative companies from 2005 through 2009.) A quick look at the list shows Apple at number one and Google at number two. OK, intuitively that sounds right. But because of the methodology, the list was largely a popularity contest based on past performance. From our perspective, by 2010 it no longer made sense to have General Electric, Sony, and Toyota in the top 10 of a list of most innovative companies.
TABLE 7-1
Businessweek list of most innovative companies, 2005–2009
*Five-year average rank; excludes private companies: Virgin at 16 and Tata at 25.
So we decided to develop our own list of innovative companies back in 2010 based on expectations of future innovations. We thought the best way to do this would be to see whether investors—voting with their wallets and purses—could give us insight into which companies they believed were most likely to produce new products, services, or markets.
We teamed up with HOLT (a division of Credit Su
isse that had done a similar analysis for The Innovator’s Solution) to develop a methodology for determining what percentage of a firm’s market value could be attributed to its existing products, services, and markets. If the firm’s market value was higher than the cash flows attributed to its existing businesses, then the company showed an innovation premium. This is the proportion of a company’s market value that cannot be accounted for from cash flows of its current products and businesses in its current markets. Investors give this premium because they expect companies to come up with profitable new products or markets (for details on how to calculate the premium, see the endnote).1 It is a premium that every executive and every company would like to have.
So how did the Businessweek top twenty-five from 2005 through 2009 stack up using our methodology? Our analysis revealed a very different ranking order. (See our ranking in table 7-1 based on the average innovation premium over five years.2) Our research put Amazon at number one (with a premium of 57 percent), Apple at number two (a premium of 52 percent), and Google at number three (a premium of 49 percent)—results that are reasonably similar to the Businessweek list. But take a look at the bottom five. Samsung (–29 percent), Sony (–28 percent), Honda (–27 percent), Toyota (–26 percent), and BMW (–26 percent) were generating declining cash flows from existing businesses that were producing what we call an innovation discount. In other words, investors were not anticipating growth from new innovative products or services and, worse, they were expecting that the profit levels from these firms’ existing businesses would likely shrink.