In 2008, two promising startups emerged in the nascent civilian drone industry. Their beginning circumstances were remarkably similar. But a decade later, they’re in profoundly different places.
Both Dà-Jiāng Innovation Science and Technology Co. (DJI) and 3D Robotics (3DR) were founded by impassioned hobbyists. Both were bootstrapping to fund their activities, as neither had finance backing from venture capitalists. And both could be said to adhere to “lean startup” principles, which are widely used by entrepreneurs entering new or emerging markets. The lean startup methodology is characterized by hypothesis-driven experimenting with different product features and functions while staying in close touch with customers.
Yet, how these two startups applied lean principles varied widely. Today, China-based DJI is a billion dollar company, the global leader in an industry expected to top $129 billion by 2028. In contrast, California-based 3DR crashed and burned, and was forced to exit the consumer drone business in 2016.
Why? Same timeframe, same goals, following same lean startup principles of rapid prototyping and iterating to “fail fast,” and maintaining the flexibility to “pivot” business models as required by the often-chaotic dynamics of the embryonic drone market.
When entrepreneurs take the lean startup model at face value, they likely miss some key complexities and challenges of establishing and growing a startup.
“Lean has super helpful ideas for any entrepreneur, but he or she also needs to think in a more sophisticated way about how to scale and ultimately succeed,” says Kathleen Eisenhardt. Eisenhardt is the Stanford W. Ascherman M.D. Professor in Stanford University’s Department of Management Science & Engineering and Faculty Co-Director of the Stanford Technology Ventures Program. Eisenhardt’s research sits at the nexus of strategy and organization theory, focusing particularly on high-velocity markets and technology firms.
Eisenhardt’s research shows that there is more to the lean startup methodology than you might see at first glance. When entrepreneurs take the lean startup model at face value, they likely miss some key complexities and challenges of establishing and growing a startup. Eisenhardt is working to illuminate gaps in our understanding of startups (including lean) and to create insights that help entrepreneurs think more clearly about how to navigate uncertain and high-velocity new markets.
In the case of DJI and 3DR, they both followed the lean startup recipe. But DJI thought more granularly—and strategically—about the ingredients, and combined them in a way to get a much better outcome.
Identifying why some startups succeed—and others don’t
Eisenhardt pioneered methods for creating deep theoretical understanding of how startups actually win. Termed multi-case theory building, it has been widely adopted by many academics around the globe attempting to understand how organizations form, operate and succeed—or not, as the case may be.
Applied to startups, multi-case theory building involves systematically tracking and comparing similar companies right from the start. Through exhaustive external data collection (via company-generated blogs and videos, as well as analyst insights), rich open-ended interviews with founders, key employees and investors over multiple years, as well as onsite observations, researchers can gain extraordinary in-depth insights into how strategies, processes and structures evolve. By comparing the distinct paths that closely matched startups take—and their results—Eisenhardt and her colleagues can pinpoint the essentials of successful strategies and tactics.
In her research, Eisenhardt (usually with PhD students) immerses herself in early-stage companies as close to their inception as possible, and observes them over time using the multi-case method. The result: a rich body of research that has resulted in a prolific outpouring of peer-reviewed publications for Eisenhardt, who has authored two award-winning books, five Harvard Business Review articles, and more than 100 articles to date.
Three ways to build on lean principles
To help entrepreneurs, Eisenhardt has three insights for digging deeper and going further with lean concepts:
- Run nuanced experiments: Continual, iterative experimenting is at the heart of lean. But there are different ways to experiment. Choosing the right one for a given business situation is essential.
- Codify lessons into “simple rules”: Continuous learning from mistakes is another essential part of lean. But if you don’t embed that learning into guidelines or rules that you actually can use going forward, then you haven’t benefited from your failures.
- Focus on bottlenecks, not just customers: If you adhere to lean, then you focus fiercely on customers. This may work in straightforward markets with easier technologies and products. Yet as complexities multiply, you need to think in terms of what’s holding back your growth. Yes, sometimes that bottleneck is understanding customer needs, and so focus on that. But at other times, more critical bottlenecks require immediate care.
Run the right experiment
According to lean principles, you learn by doing, or, experimenting. What this means: showing prototypes to potential customers. And not just once. Over and over again, improving your product with each iteration. But Eisenhardt says that entrepreneurs would do well to adopt a more sophisticated way of looking at experimenting.
Eisenhardt and colleagues have identified three different types of experiments: passive, parallel, and sequential.
- Passive means doing nothing: just putting your product out there for customers to play with, pausing, and seeing what happens.
- Parallel means doing multiple experiments at once.
- Sequential is simply doing one experiment at a time, one after another, improving the prototype product each time.
Sequential is the method most associated with lean. But it’s not always the best. Take the civilian drone market. Entrepreneurs, investors and customers were wildly enthusiastic about drones. But there was considerable debate about how they would be used. What would be the proverbial “killer app”?
DJI went the “parallel” experimentation route. Rather than focusing on one particular market opportunity or waiting for customers to come to them, DJI sent employees out to multiple trade shows—Midwest agriculture fairs, Las Vegas real-estate conventions, German sports conventions, for instance. By seeking out and connecting with lead users in multiple fields, they experimented in parallel with how drones might be used in agriculture, real estate, motion picture and other industries.
“By engaging in parallel experimentation, DJI reduced the uncertainty of ‘where will drones really pay off?’” says Eisenhardt. And they found a killer app: the movie industry, which could replace the costly aircraft and helicopters they used for aerial shots with less-expensive drones.
Meanwhile, 3DR took a “passive” approach to experimentation: it simply put its drone technology out into the world, and waited for the killer app to emerge. It never did for them. DJI sprinted dramatically ahead.
“Being customer-driven works in certain industries and scenarios. But if the customer’s not the biggest immediate bottleneck, then start somewhere else.”
– Kathleen Eisenhardt
How to decide between these three methods of experimenting?
- Passive works when “you don’t know what you don’t know”: that is, where there is overwhelming ambiguity.
- Parallel is best when the entrepreneur has an abundance of clear potential alternatives that can be tried at once. (This is also called a multi-armed bandit solution—named after Vegas slot machines.)
- Sequential is optimal when the entrepreneur is stepping through stages in solving a problem. As with the other two ways to experiment, the focus is on uncertainty, but it’s more roughly linear rather than undefined or multi-pronged.
Codify learning into simple, scalable rules
Entrepreneurs can learn from lean’s rapid prototyping and iteration cycles. But for this to work, entrepreneurs have to take the extra step and codify what they’ve learned into “simple rules”—a few rules of thumb—that they can use to stay focused, flexible, and agile.
Eisenhardt is the author (with Donald Sull) of Simple Rules: How to Thrive in a Complex World, a culmination of more than a decade of research into the question of how businesses balance between being flexible and efficient (and fast). What she has found: even large, complicated companies like Intel and Google rely on simple rules of thumb that guide them in making acquisitions, developing products and hiring top talent.
Eisenhardt, for example, studied “Master Chef” (a pseudonym to keep the company anonymous which improves candor and access to her research targets), a company that illustrates how simple rules can shape strategy and scale firms in new and growing markets.
Master Chef connects travelers in Asia with people who cook meals in their homes for guests. Going to Hanoi and want an authentic home-cooked meal? The idea was that the Master Chef platform would seamlessly connect a traveler with a local host.
Master Chef founders began with the hypothesis that their hosts were going to be rural women, steeped in local culture and interested in making some extra money. But they were surprised. It turns out that these women generally didn’t have internet access. Nor did they speak English. Thus the Master Chef team learned by experimenting early on that hosts would need to be relatively well-off urban women, with some Internet and language skills.
Further, they learned that in addition to wanting to be connected to great cooks—an obvious attribute—guests enjoyed hosts who were culturally engaged with their region and charismatic storytellers.
“The founders developed a “simple rules” checklist—codifying how to pick the right hosts,” says Eisenhardt. “Great food, check. Internet access, check. Speaks English, check. Tells intriguing stories, check. Codifying learning into rules makes the lessons of experiments stick.”
When Master Chef began expanding into different countries, its founders came up with simple rules for internationalizing. For example, they now have a rule that any one location needs three to five hosts for critical mass. More is helpful, but not required. Two is just not enough. They also learned rules for which cities would work well (like Hanoi, yes; Phuket, no). And they stuck rigorously to these and other simple rules.
“Experimentation per se doesn’t buy much if the entrepreneur doesn’t actually put the learning to work by codifying simple rules for the business,” says Eisenhardt.
Focus on bottlenecks—not always customers
Lean startups are also supposed to be aggressively customer-driven. Rapid and continuous feedback from customers improves product and fleshes out a business model, or suggests when it’s time to pivot. That works well for some kinds of businesses. But it’s not “one size fits all.” Especially with a highly technical product or with a complex platform or ecosystem that requires partners, the customer may not be the right initial focus.
Then, focusing on customers might distract from more pressing bottlenecks to growth.
Instead of asking, what does the customer want? Ask, what’s the major bottleneck that’s blocking growth? What’s keeping us from growing now?”
“Being customer-driven works in certain industries and scenarios. But if the customer’s not the biggest immediate bottleneck, then start somewhere else,” says Eisenhardt.
Master Chef is an example of a company that faced a series of bottlenecks to growth —and yes, those included figuring out what customers really wanted. But instead of blindly going after developing customer relationships, Master Chef’s founders took a step back. Was the lack of customers the biggest bottleneck—now? Or was it developing the technology platform so customers could seamlessly book meals with hosts? Or perhaps the biggest barrier to moving forward was finding a sufficient number of qualified hosts to make and serve meals?
The founders considered alternatives, and compared their business model to two-sided marketplace pioneers like Uber and Airbnb. They also picked up advice from savvy executives at Lyft, Expedia and other digital travel companies.
Master Chef founders ultimately decided that developing a high-quality host community was where to start—that is, it was the biggest bottleneck to creating a vibrant marketplace. Without superb hosts, there were no customers. It was as simple as that. So the founders focused on finding and training hosts first.
Eisenhardt is working to illuminate gaps in our understanding of startups (including lean) and to create insights that help entrepreneurs think more clearly about how to navigate uncertain and high-velocity new markets.
The founders moved everything else to the side. Using their savings and capital from family and friends, they hacked together a website in a few days. At first, they manually booked meals. They didn’t advertise, but depended on word of mouth to get travelers interested in their growing network of hosts. Eventually, they reached what Eisenhardt calls a “plateau” where they paused on adding hosts, and moved onto their next bottleneck, which was attracting the customers, to get the “flywheel” of a two-sided marketplace going.
More to Lean than meets the eye
Today, most entrepreneurs and students of entrepreneurship are aware of lean startup principles. They know that too much planning in uncertain and high-velocity markets is fruitless. They understand that experimentation is important. Learning is important. Customers are important.
But there is more to startups than the basics of lean. Eisenhardt and her students are unpacking the basic concepts of entrepreneurship, including lean—like constant experimentation and iterating on product design with customers—to come up with specific and often unexpected recommendations for how to survive and thrive in entrepreneurial settings.