by Jim Collins
That said, if you feel compelled to continuously make fundamental changes to the sequence or components of the flywheel, you’ve likely failed to get your flywheel right in the first place. Rarely does a great flywheel stall because it’s run out of potential or is fundamentally broken. More often, momentum stalls due to either poor execution and/or failure to renew and extend within a fundamentally sound flywheel architecture. It is to the topic of extending the flywheel that we now turn.
EXTENDING THE FLYWHEEL
How do great companies go about extending a flywheel? The answer lies in a concept I developed with my colleague Morten Hansen in our book, Great by Choice. Morten and I systematically studied small entrepreneurial companies that became the 10X winners (beating their industries by more than ten times, in returns to investors) in highly turbulent industries in contrast to less successful comparison cases in the same environments. We found that both sets of companies made big bets but with a huge difference. The big successes tended to make big bets after they’d empirically validated that the bet would pay off, whereas the less successful comparisons tended to make big bets before having empirical validation. We coined the concept fire bullets, then cannonballs to capture the difference.19
Here’s the idea: Imagine a hostile ship bearing down on you. You have a limited amount of gunpowder. You take all your gunpowder and use it to fire a big cannonball. The cannonball flies out and splashes in the ocean, missing the oncoming ship. You turn to your stockpile and discover that you’re out of gunpowder. You’re in trouble. But suppose instead that when you see the ship bearing down, you take a little bit of gunpowder and fire a bullet. It misses by 40 degrees. You make another bullet and fire. It misses by 30 degrees. You make a third bullet and fire, missing by only 10 degrees. The next bullet hits—ping!—the hull of the oncoming ship. You have empirical validation, a calibrated line of sight. Now, you take all the remaining gunpowder and fire a big cannonball along the calibrated line of sight, which sinks the enemy ship.
In looking across the history of great companies in all our research studies, we find a frequent pattern. They usually begin life being successful in a specific business arena, making the most of their early big bets. But soon they make a conceptual shift from “running a business” to turning a flywheel. And over time, they extend that flywheel by firing bullets, then cannonballs. They crank the flywheel in their first arena of success, while simultaneously firing bullets to discover new things that might work, and as a hedge against uncertainty.
Some bullets hit nothing, but some give enough empirical validation that the company then fires a cannonball, providing a big burst of momentum. In some cases, these extensions come to generate the vast majority of momentum in the flywheel, and in a few cases (such as when Intel moved from memories to microprocessors), they entirely replace what came before.
Apple’s flywheel extension into its biggest business—smart handheld devices—followed exactly this pattern. In 2002, nearly all of Apple’s flywheel momentum came from its line of Macintosh personal computers. But it had fired a bullet on this little thing called an iPod, described in its 2001 form 10-K as simply “an important and natural extension” of Apple’s personal computer strategy. In 2002, the iPod generated less than 3 percent of Apple sales. Apple kept firing bullets on the iPod, developing an online music store along the way (iTunes). The bullets kept hitting, the iPod kept adding momentum to the flywheel, and Apple eventually fired a huge cannonball, betting big on the iPod and iTunes. And Apple kept extending the flywheel, from iPod to iPhone, from iPhone to iPad, and Apple’s flywheel extension became the largest generator of momentum.20
In the nearby table, I’ve listed a range of fabulous flywheel extensions from corporate history. In every case, the company followed the bullet-to-cannonball method to extend and accelerate an underlying flywheel that had been turning for years.
COMPANY
FIRST ARENA OF FLYWHEEL SUCCESS
NEXT BIG EXTENSION OF THE FLYWHEEL
3M
Abrasives (e.g., sandpaper)
Adhesives (e.g., Scotch Tape)
Amazon
Internet-enabled retail for consumers
Cloud-enabled web services for enterprises
Amgen
Therapeutics for low-blood-cell conditions
Therapeutics for inflammation and cancer
Apple
Personal computers
Smart handhelds (iPod, iPhone, iPad)
Boeing
Military aircraft
Commercial jetliners
IBM
Accounting tabulating machines
Computers
Intel
Memory chips
Microprocessors
Johnson & Johnson
Medical and surgical products
Consumer health-care products
Kroger
Small-scale grocery stores
Large-scale superstores
Marriott
Restaurants
Hotels
Merck
Chemicals
Pharmaceuticals
Microsoft
Computer languages
Operating systems and applications
Nordstrom
Shoe stores
Department stores
Nucor
Steel joists
Manufactured steel
Progressive
Non-standard (high-risk) car insurance
Standard car insurance
Southwest Airlines
Low-cost intrastate airline (Texas only)
Low-cost interstate airline (coast to coast)
Stryker
Hospital beds
Surgical products
Walt Disney
Animated films
Theme parks
When does a new activity become a second flywheel, as distinct from an extension? Most seeming “second flywheels” come about organically, as bullet-to-cannonball extensions of a primary flywheel. Amazon showed this precise pattern with its Amazon Web Services, which enables organizations big and small to efficiently buy computing power, store data, host websites, and avail themselves of other technology services. Amazon Web Services began as an internal system to provide backend technology support for Amazon’s own e-commerce efforts. In 2006, the company fired a bullet on offering these very same services to outside companies. The bullet hit its target and Amazon had enough calibration to fire a cannonball. A decade later, Amazon Web Services (while still contributing less than 10 percent of Amazon’s overall net sales) generated a substantial portion of Amazon’s operating income.21
Even though Amazon Web Services first appears to be a very different activity than the consumer-retail business, it has substantial similarities. As Bezos wrote in his 2015 annual letter to shareholders, “Superficially, the two could hardly be more different. One serves consumers and the other serves enterprises . . . Under the surface, the two are not so different after all.” Amazon Web Services aims to lower prices and expand offerings to an ever-growing cadre of customers, leading to increasing revenues per fixed costs, which then drives the flywheel around again. The whole idea is to make it as easy and cost-effective for enterprises to meet their technology needs as it is for consumers buying personal stuff at the Amazon marketplace. Sure, there are differences in how the two businesses operate, but the two are more like fraternal twins than being from an entirely different family lineage.
Every large organization will eventually have multiple sub-flywheels spinning about, each with its own nuance. But to achieve greatest momentum, they should be held together by an underlying logic. And each sub-flywheel should clearly fit within and contribute to the whole.
The most important thing is to keep turning the overall flywheel—and every component and sub-flywheel—with creative intensity and relentless discipline. Even with the early growth and profitability of Amazon Web Services, Bezos remained obsessed with keeping Amazon’s consumer
-retail business as vibrant and energized as when the company first began. After all, even as Amazon approached $200 billion in annual revenues, it had less than 1 percent of the global retail market.22
STAY ON THE FLYWHEEL . . . AND OUT OF HOW THE MIGHTY FALL
In studying the horrifying fall of once-great companies, we see them abandoning the key principles that made them great in the first place. They vest the wrong leaders with power. They veer from the First Who principle and cease to get the right people on the bus. They fail to confront the brutal facts. They stray far beyond the three circles of their Hedgehog Concept, throwing themselves into activities at which they could never become best in the world. They subvert discipline with bureaucracy. They corrupt their core values and lose their purpose. And one of the biggest patterns exhibited by once-great companies that bring about their own senseless self-destruction is failure to adhere to the flywheel principle.
In our research for How the Mighty Fall, we found that the demise of once-great companies happens in five stages: (1) Hubris Born of Success, (2) Undisciplined Pursuit of More, (3) Denial of Risk and Peril, (4) Grasping for Salvation, and (5) Capitulation to Irrelevance or Death. Take special note of Stage 4, Grasping for Salvation. When companies fall into Stage 4, they succumb to the doom loop, the exact opposite of building flywheel momentum. They grasp for charismatic saviors or untested strategies or big uncalibrated cannonballs or cultural revolutions or “game-changing” acquisitions or transformative technologies or radical restructurings (then another and another) or . . . well, you get the idea.
In Stage 4, each grasp for salvation creates a burst of hope and momentary momentum. But if there’s no underlying flywheel, the momentum doesn’t last. And with each grasp, the enterprise erodes capital—financial capital, cultural capital, stakeholder capital—and weakens. If the company never gets back to the discipline of the flywheel, it will likely continue to spiral downward until it enters Stage 5. No enterprise comes back from Stage 5. Game over.
Circuit City, which we studied in the original research for Good to Great, later “earned” a spot in How the Mighty Fall, and its demise teaches important lessons about the flywheel. During its good-to-great years, Circuit City made the leap from dismal mediocrity to superstar success under the inspired Level 5 leadership of Alan Wurtzel, transforming a hodgepodge of hi-fi stores into a sophisticated system of consumer electronics superstores, generating total returns to investors of more than eighteen times the general stock market over fifteen years. But after the Wurtzel era, Circuit City began to decline, slowly at first, almost imperceptibly, as usually happens with companies moving through the early stages of decline, then it plummeted precipitously through Stage 4, and right into Stage 5 capitulation and death.
How did this happen? A big part of the answer lies in two fundamental mistakes the post-Wurtzel leadership team made related to the flywheel principle. First, they became distracted by searching for the Next Big Thing. Circuit City sought big new ideas for growth, anticipating the day that the consumer electronics superstores would run out of great locations in which to open across the country. This in itself was a good idea, just as Amazon continually sought new ideas to propel the flywheel. But, unlike Amazon under Bezos, Circuit City neglected to keep the consumer electronics retail business robust and relevant. Meanwhile, an up-start competitor named Best Buy seized the market.23
Second—and this is the most fundamental lesson from Circuit City’s demise—the post-Wurtzel team underestimated how far a flywheel could go if seen as an underlying architecture (that can be extended) rather than as a single line of business. The great tragedy of Circuit City is that it did indeed invent a spectacular extension, called CarMax, which could have created momentum for at least another couple of decades. The idea behind CarMax was to do for the used-car business what the Wurtzel team had done for hi-fi stores, to professionalize and transform a hodgepodge industry into a sophisticated system of superstores under one trusted brand.24
Circuit City fired a bullet with its first CarMax store in Richmond, Virginia. It proved successful. So, it fired a second bullet, opening a second CarMax in Raleigh, North Carolina, which also proved successful. Next, it fired two more bullets in Atlanta, Georgia. With empirical validation in hand, Circuit City fired a cannonball, opening CarMax superstores and expanding into new regions—Florida, Texas, California, and beyond. By the early 2000s, CarMax was growing at close to 25 percent a year, generating more than $3 billion in profitable sales in 2002.25
Now, stop and think about this for a minute. How did CarMax’s success presage Circuit City’s fall? With CarMax, Circuit City had created a huge new extension of its flywheel that could generate years of additional momentum. The CarMax flywheel extension could have been analogous to Apple’s flywheel extension from personal computers to smart handhelds, Boeing’s flywheel extension from military propeller-driven bombers to commercial jet airliners, Marriott’s flywheel extension from restaurants to hotels, and Walt Disney’s flywheel extension from animated films to theme parks. And in the event that the consumer electronics superstores did become untenable as a business, the company could redeploy all its energy into CarMax (similar to Intel’s move out of memories and into microprocessors). But to do so would have required the conceptual wisdom to see CarMax as an extension of an underlying flywheel architecture.
Sadly, the post-Wurtzel team got rid of the CarMax superstores, jettisoning CarMax into its own separate company. It was as if Intel had decided in 1985 to get rid of the microprocessor business and keep the memory-chip business; the spun-out microprocessor company might have been successful, but Intel would have likely died. Fortunately for Intel, Grove and Moore had the strategic acumen to see the microprocessor business as an extension of its underlying flywheel. Circuit City failed to make this conceptual leap.
As Alan Wurtzel later wrote in his book, Good to Great to Gone (which I warmly recommend reading), “Looked at from a long-term perspective, it is disappointing that CarMax did not remain part of the Circuit City portfolio . . . The initial premise for CarMax was to create a portfolio of retail companies, so that as one matured another would be coming along to support the growth of the whole.”26 Wurtzel understood CarMax as part of a bigger flywheel, but those who ran the company after him didn’t. If Circuit City had continued to evolve and renew the consumer electronics superstores (as Best Buy did) and continued to extend its underlying flywheel into new arenas (such as with CarMax), it might have remained a great company, climbing steadily upward in the S&P 500. Instead, Circuit City lost all its flywheel momentum and careened into the later stages of decline—down, down, down, the doom loop hurtling the company toward irrelevance. The once good-to-great company died in the winter of 2008.27
THE VERDICT OF HISTORY
After conducting a quarter-century of research into the question of what makes great companies tick—more than six thousand years of combined corporate history in the research data base—we can issue a clear verdict. The big winners are those who take a flywheel from ten turns to a billion turns rather than crank through ten turns, start over with a new flywheel, push it to ten turns, only to divert energy into yet another new flywheel, then another and another. When you reach a hundred turns on a flywheel, go for a thousand turns, then ten thousand, then a million, then ten million, and keep going until (and unless) you make a conscious decision to abandon that flywheel. Exit definitively or renew obsessively, but never—ever—neglect your flywheel. Apply your creativity and discipline to each and every turn with as much intensity as when you cranked out your first turns on the flywheel, nonstop, relentlessly, ever building momentum. If you do this, your organization will be much more likely to stay out of How the Mighty Fall and earn a place amongst those rare few that not only make the leap from good to great but also become built to last.
Appendix
THE FLYWHEEL WITHIN A FRAMEWORK
A Map for the Journey from Good to Great
I wrote this mo
nograph to share practical insights about the flywheel principle that became clear in the years after first writing about the flywheel effect in chapter 8 of Good to Great. I decided to create this monograph because I’ve witnessed the power of the flywheel, when properly conceived and harnessed, in a wide range of organizations: in public corporations and private companies, in large multinationals and small family businesses, in military organizations and professional sports teams, in school systems and medical centers, in investment firms and philanthropic endeavors, in social movements and nonprofits.
Yet the flywheel effect alone does not make an organization great. The flywheel fits within a framework of principles we uncovered through more than a quarter-century of research into the question of what makes a great company tick. We derived these principles using a rigorous matched-pair research method, comparing companies that became great with companies (in similar circumstances) that did not. We’d systematically analyze the histories of the contrasting cases and ask, “What explains the difference?” (See nearby diagram, “The Good-to-Great Matched-Pair Research Method.”)