The Lean Startup

Most products fail

Most of us are familiar with the “hero’s journey” in entrepreneurship—the notion that an unwavering resolve and a fantastic idea will somehow drag a startup into stardom. But in The Lean Startup, Eric Ries tells us that this is the “mythmaking industry hard at work.” It feeds on selection bias, hindsight bias, and our desire to believe in stories of rags-to-riches.

The reality is that most products and companies fail. Henry Ford, Ries reminds, competed with almost five hundred fledgling automakers in the early 1900s. Nearly all of them crashed and burned. Perseverance does not guarantee survival in a Darwinian market-ecology. If a self-help book tells you otherwise, you best not believe it.

A “theory of entrepreneurship should address all the functions of an early-stage venture”, from “vision and concept… [through to] “structure and organizational design”, Ries writes. Following his experience at Silicon Valley, and his frustration with how business schools teach entrepreneurship, Ries developed his own approach: the lean startup model.^[1]

The lean startup, however, is not a prescriptive set of actions. “Too many startup business plans look more like they are planning to launch a rocket ship than drive a car.” The future is complex, uncertain, and forever changing. In these environments, detailed forecasts, milestones, and planning are not always helpful.

[1] As you can guess, the lean startup gets its name and inspiration from the lean manufacturing methodologies practiced in Japan. “Among its tenets”, Ries says, “are drawing on the knowledge and creativity of individual workers, the shrinking of batch sizes, just-in-time production and inventory control, and an acceleration of cycle times”. These innovations enabled the Toyota Production System and other Japanese automakers to thrive following the Second World War.

Validated learning

To begin, the lean startup method recognises the value of learning well and fast. Learning “is the essential unit of progress for startups”, Ries writes. If you view entrepreneurship as “an exercise in institution building”, the entrepreneur’s chief responsibility then is to establish the systems and processes for “validated learning”.

Indeed, the scariest thing in entrepreneurship is building something that nobody wants. The company cannot afford to wait a year or two or three to determine whether its core products and activities are wasteful or not.

Startups are great at feeling productive. We busy ourselves, developing features, fixing bugs, and meeting milestones. But many of us fail to learn about our customers until launch. The lean startup knows, however, that learning about the customers is valuable. Generating scores of code that nobody wants is wasteful.

This is easier said than done, of course. In venture capital, Ries laments that it is “easier to raise money… when you have zero revenue… and zero traction than when you have a small amount.” Where “small numbers invite questions”, “zero invites imagination.” Some entrepreneurs will avoid disconfirming evidence and critical feedback until it’s too late.

“The big question of our time is not can it be built? but should it be built? This places us in an unusual historical moment: our future prosperity depends on the quality of our collective imaginations.”
Eric Ries. (2011). The Lean Startup.

The scientific credo

To learn effectively, the lean startup employs the scientific method. After all, “if you cannot fail, you cannot learn.” The first product is about hypothesis-making, early experimentation and empirical feedback. It teaches us in ways that ivory tower planning and tinkering cannot.

Nick Swinmurn, for example, did not launch Zappos—the online retailing giant—with plenty of inventory under his feet. He started out instead by snapping photos of shoes and uploading them onto his website. If shoppers purchased a pair online, he’d then return to his local brick-and-mortar store to buy the same pair and ship them off. While the experiment was horribly inefficient, early sales gave him the validation he needed to go all in.

Growth and value hypotheses

The lean startup, Ries adds, must validate two hypotheses. First is the value hypothesis: Does your product create value for your users? Second is the growth hypothesis, which asks: Why will early adopters switch to your product? It goes without saying that if the startup hopes to scale, it must satisfy both conditions in one form or another.

Facebook, for example, passed both tests swimmingly. Shortly after its restricted launch in early 2004, more than half of Harvard’s undergraduate students were on the platform. All this too without a cent of marketing or advertising. Now that’s growth. Moreover, “more than half of [its] users came back to the site every single day,” Ries notes. Now that’s value.

Build-measure-learn feedback loops

Most companies, however, are not as viral as Facebook. Most startups will have to go through an iterative process to test its “leap-of-faith assumptions”, to throw away what doesn’t work, and to focus on what does. Ries calls this the build-measure-learn feedback loop.

To do this well, “firsthand understanding” is critical. Founders and managers must not allow their egos to stop them from talking to customers, walking through stores, and experiencing their products. If you’re deferring to consultants and other hired guns, the game is probably over.

Learning loops and minimum-viable-products (MVPs) come in many forms. Consider, for instance, the early days of the file hosting company Dropbox. While their founder Drew Houston believed in the future of file synchronization and cloud storage, it was difficult to ascertain demand because there was nothing like it at the time. However, before risking everything on development, Drew made a waitlist and three-minute video about Dropbox for potential early-adopters. What happened next? Well, Dropbox’s beta waitlist reached 75,000 people overnight. The signups validated Houston’s hypothesis.

Learning loops and MVPs are not without risk, of course. Early releases may give competitors ideas, while imperfections may hurt the brand. But the benefits of validated learning, Ries suggests, usually exceeds the risk. After all, if a startup is to succeed, it will have to face the customer and the competition sooner or later. The only option is a commitment to learning, feedback, and iteration.

Why startups fail

Entrepreneurs tend to fall into one of two camps of failure. The first group rushes out of the gates, building furiously after a few cursory conversations or brainstorms. Without knowing their customers well enough, they wrongly believe that they’re progressing well. The second group, by contrast, lives with “analysis paralysis”. They are forever tinkering with their mission statements, strawman models, and market research. Indeed, many large companies stagnate for this reason too. Endless meetings and planning stifles trial-and-error and learning.

Innovation accounting

As the name suggests, the build-learn-measure feedback loop depends mightily on the quality of measurement. Entrepreneurs need to know whether their actions are translating into sustainable results for the company.

Ries says it is important to distinguish between vanity metrics and innovation accounting. The former tends to focus on gross results like total sales. They do not tell us which initiatives are working and which are not. They also incentivize “theater tricks” like “last-minute ad buys, channel stuffing, and whiz-bang demos” to pretty-up the numbers.

Innovation accounting, however, employs “actionable, accessible, and auditable” metrics that help entrepreneurs to separate cause from effect. This is also easier said than done, of course. After all, it is never easy to isolate company actions from economic conditions and parallel initiatives. But a startup that is serious about learning will find ways to measure their actions effectively.

“Vanity metrics wreak havoc because they prey on a weakness of the human mind. In my experience, when the numbers go up, people think the improvement was caused by their actions… Unfortunately, when the numbers go down, it results in a very different reaction: now it’s somebody else’s fault… Is it any wonder these departments develop their own distinct language, jargon, culture, and defense mechanisms against the bozos working down the hall?”
Eric Ries. (2011). The Lean Startup.

Growth engines

Moreover, the lean startup focuses on enduring results, not once-off gimmicks or marketing stunts. When it comes to sustainable growth, Ries has a “simple rule: new customers come from the actions of past customers.” And past customers can drive growth by: (1) word of mouth; (2) repeat purchases and addons; (3) funded advertising (i.e., profits to pay for future customer acquisition); and (4) byproducts of customer use (e.g., network effects).

The startup must therefore select and monitor its “growth engine” with care. Service providers like retail banks and insurers, for example, depend on a “sticky engine” and high switching costs. So it is most worrisome for them when existing customers are leaving in droves. (Topline vanity metrics, by the way, will not reveal this sort of deterioration.)

Other companies, like social media networks and payment systems, depend on “viral engines”. Their value grows exponentially with more users. So product awareness and usage needs to spread like a disease. Facebook, for example, would not survive if it charged users to sign up. Another free-platform would probably win out on network effects and scale economies.

Of course, most startups are neither sticky nor viral. They have to rely on a “paid engine”. These startups must pay attention to acquisition costs, customer lifetime value, and related measures of per-sale return-on-investment. If free cash flows cannot pay for long run growth, the company will run into problems.

Pivots and runways

Now, after multiple rounds of the build-measure-learn feedback loop, the entrepreneur must decide whether to pivot or persevere. In this light, the startup’s runway depends not only on its burn rate but on the number of pivots it can make in its struggle for survival. Misleading metrics, vague hypotheses and fear can diminish the number of chances for change the company gets.

“You have to commit to a locked-in agreement—ahead of time—that no matter what comes of testing the MVP, you will not give up hope. Successful entrepreneurs do not give up at the first sign of trouble, nor do they persevere the plane right into the ground.”
Eric Ries. (2011). The Lean Startup.

It goes without saying that pivots are hard. In Only the Paranoid Survive, former Intel CEO Andy Grove recalls the anguish, inertia, and identity crisis that gripped Intel as they abandoned the memory business to refocus on microprocessors. Indeed, Intel was hemorrhaging and no longer competitive with Japanese memory makers. And had Intel delayed any longer, Grove believes “[they] would have been relegated to an immensely tough economic existence.”

Similarly, Tiny Speck co-founder Stewart Butterfield recalls his failed venture with the multiplayer online game Glitch. Despite their best efforts, the game just could not build traction. But Butterfield wasn’t done. During development, his team had built an inhouse communication tool to support their collaboration. Butterfield noticed its potential and made the painful but fateful pivot. Indeed, “the office messaging platform Slack rose from the ashes of [Glitch].”

Innovation requisite structures

For large organizations, learning and pivoting may seem impossible in the face of bureaucracy, corporate inertia, and dysfunction. But it need not be this way. Innovation, Ries suggests, is about getting three “structural attributes” correct (most of which tend to come naturally to smaller, bootstrapped enterprises).

The first principle is “scarce but secure resources”. It reflects the fact that to sustain innovation, accountability and discipline are necessary factors. Indeed, teams tend to mispend and misprioritize when they’re inundated with excess resources. But when budgets are volatile and uncertain, teams tend to get risk averse. Why commit to anything long-term if there’s a chance that your project is scrapped before completion? Companies need to get the balance right.

The second and third principles, Ries says, are “independent authority to develop their business”, and “a personal stake in the outcome”. As we’ve heard Nassim Taleb say many times over, skin-in-the-game matters. What’s more, an innovative team needs “autonomy to develop… conceive and execute experiments” without an endless bog of approvals, meetings, and consultations.

Many universities, I think, continue to struggle with research commercialization for these reasons. Their resources are abundant but their budgets are insecure. And teams rarely have the independence or agility to experiment within the enterprise. What’s more, citations follow the author, but patents stay with the institution. Where’s the incentive? Where’s the skin-in-the-game?

The boring stuff matters most

“Entrepreneurship is a kind of management”, Ries writes. “It’s the boring stuff that matters the most.” The lean startup must find the right structures, incentives, and institutions if it wants to cultivate something worthwhile. Most firms today, however, are bloated. They live under the bane of uncoordinated focus. Their managers extract countless hours from their employees and squander it on wasteful pursuits. One cannot help but wonder at the possibilities “if we stopped wasting people’s time”.

“As Peter Drucker said, “there is surely nothing quite so useless as doing with great efficiency what should not be done at all.” And yet we are doing the wrong things efficiently all the time… It is insufficient to exhort workers to try harder. Our current problems are caused by trying too hard—at the wrong things… What is needed is a massive project to discover how to unlock the vast stores of potential that are hidden in plain sight in our modern workforce.”
Eric Ries. (2011). The Lean Startup.

Sources and further reading

Ries, Eric. (2011). The Lean Startup.
Taleb, Nassim N. (2017). Skin in the Game.
Grove, Andrew. (1988). Only the Paranoid Survive.
Ohno, Taiichi. (1978). Toyota Production System.