The Opposable Mind

by Roger L. Martin

  How A. G. Lafley Reinvented Innovation at P&G

  When A. G. Lafley took over as CEO of Procter & Gamble in 2000, the company had just warned Wall Street for the second time in two quarters that it would miss its profit targets. Growth was slowing, profits were sliding, and many of its biggest brands were losing market share. P&G’s stock price had fallen nearly 50 percent in six months, and the business press was openly questioning whether P&G had lost its long-standing eminence in consumer goods marketing. The front-page headline for industry bible Advertising Age read: “Does P&G Still Matter?”5

  By P&G’s conservative standards, the fifty-three-year-old Lafley was young for the top job, although he was then a twenty-three-year veteran of the company—a “lifer”—with an impressive record of innovation. He had overseen the launch of liquid Tide, Tide with bleach, and compact Tide—all successful extensions of P&G’s flagship brand. As head of P&G’s Asia operation, he and his small team transformed SKII, a tiny skin care brand originally distributed only in Japan and Hong Kong, into a global prestige brand. In the process, sales grew by a factor of twenty. But because Lafley’s predecessor had just earned the dubious distinction of being the first CEO in the company’s one hundred seventy-year history to be fired, Lafley was thrust into the CEO post without the formal grooming that was customary at P&G.

  He faced a challenge that would have tested the most experienced CEO. P&G needed nothing less than a radical overhaul of its entire approach to innovation. His predecessor saw P&G as first and foremost a research and development company, and spent accordingly. Only heavy investment in research and development, he maintained, would reignite sales and drive growth. In his eighteen-month tenure as CEO, he tripled innovation spending.

  Whatever the merits of that theory, it yielded disappointing results in practice. Only 15 percent of innovation projects met internal sales and profit targets. Give it time, said the theory’s defenders. The strategy will work, but it will take more than a quarter or two.

  Other senior executives saw P&G as primarily a marketing and brand-building enterprise. The company, they said, should focus on brands and marketing and allow R&D spending to drift back down to its historical level. That was the P&G way, said those executives, and it had worked just fine until Lafley’s predecessor tampered with it.

  Reconciling the Irreconcilable

  Lafley entered the CEO’s office, then, under pressure to opt for one of two sharply opposed models. Both sides argued that their model was, in fact, reality—“the way things are.” Lafley didn’t see it that way. He treated the opposing models as hypotheses rather than the truth, which gave him the intellectual and emotional latitude to weigh the merits and drawbacks of both models without needing to defend one and declare the other false.

  From his disinterested vantage point, Lafley was able to see that returning to historical levels of R&D spending was not going to produce the growth P&G needed. The market had changed. P&G was surrounded by fierce competitors, including store brands and private-label products that were eating into P&G’s market share. The company needed to counter with innovative new products of its own, which called for a higher level of spending than was customary at P&G.

  But Lafley could also see that the sharply increased innovation spending his predecessor favored wasn’t the entire answer either. Yes, there were some promising new products on the drawing board, but the company simply couldn’t afford to wait to see if the promised return on its massive R&D investment would materialize.

  Lafley set his sights on a resolution that would be superior to either existing alternative. He had no intention of settling for half measures. Like an artist, Lafley considered a piece of work unacceptable until it reached a certain quality standard, and an option with unpleasant trade-offs was not going to meet that standard.

  Lafley also took personal responsibility for his dilemma. Rather than blame the world at large for presenting him with a difficult choice, Lafley assumed the real problem lay within himself. “I haven’t yet,” he said to himself, “found a creative resolution that meets my standards. That’s not the world’s fault. I just haven’t thought hard enough yet.”

  Lafley took a step back and began to question assumptions that other P&G executives had taken for granted. He questioned the causal reasoning that said useful innovation output is directly proportional to the dollars invested in it. He asked himself if there were other, better ways to produce innovations than to pour money into P&G’s own research and development labs. He looked outside P&G to see how other organizations solved their innovation problems.

  He discovered that innovation was broadly distributed. Solo inventors, academics, and small companies were disproportionately represented, while big corporations had a lower share than would be expected, given their size and resources. But big corporations, with manufacturing, marketing, and distribution capabilities far beyond the reach of the small players, did have an advantage in developing, testing, and commercializing breakthrough ideas.

  Free of preconceptions about the “right” way to innovate, Lafley saw a way to hitch small-company inventiveness to P&G’s vast network of resources. He set a target for P&G to obtain 50 percent of its innovations from outside the company by connecting with a wide array of outside innovators. P&G would then exploit its huge resource advantage to develop and commercialize the innovation. Lafley believed this strategy, dubbed “Connect & Develop,” would enable P&G to parlay a relatively modest investment in innovation into above-average growth.

  P&G’s research and development staff greeted Lafley’s proposed resolution with consternation. “I had principal scientists—PhDs—angry because they thought that we were going to outsource,” Lafley said. “They thought that we were walking away from innovation, which was the lifeblood of the company. I said no, au contraire, what we are really trying to do is double or triple the productivity of the R&D labs; we want to get more of this stuff commercialized.”6

  One of the first fruits of the Connect & Develop initiative was the Spinbrush, a low-priced, battery-powered oscillating toothbrush. P&G acquired it early in its commercialization from a tiny company that based it on the Spin-pop, a battery-powered spinning lollipop that was wildly popular among prepubescents. P&G’s contribution was to slap the Crest brand—a beloved and venerable brand name in oral care—on it and distribute it to stores across the country. The combination was irresistible. Within four years, this connection of P&G’s marketing and distribution muscle with small-enterprise innovation developed into a $160 million product line.7

  Lafley didn’t do it alone. He would be the first to give P&G’s top innovation executives—Gil Cloyd, Larry Huston, and Nabil Sakkab—credit for formulating and executing Connect & Develop.8 But they wouldn’t have brought the idea to fulfillment without Lafley’s ability to distinguish between reality and models of reality, his unyielding high standards, his assumption of personal responsibility, and his willingness—eagerness, really—to rethink the entire problem from the ground up. Lafley created a leadership environment that encouraged Cloyd, Huston, and Sakkab to push their idea as far as it would go. Lafley then used his credibility and organizational authority to put the idea into operation. Integrative thinking should not be confused with solo, heroic leadership. But it is leadership all the same.

  How Red Hat Found Profits in Free Software

  One of the most fascinating stories in the entire technology sector has been the rising challenge by Linux, the open-source operating system, to Microsoft’s once-monopolistic dominance of the market for operating system software. Richard Stallman, the MIT professor who put forth the theory of “open-source” software development, and Finnish engineer Linus Torvalds, who developed the foundational code, or kernel, of Linux, are clearly the two most important figures in this story. The next most important person may be Bob Young, cofounder of Red Hat software, the dominant Linux distributor. Young’s creative resolution of a crucial strategic dilemma was the event that put Red Hat�
�and Linux—on the path to profit and power in the marketplace.

  Young stands out, even in an industry full of eccentrics. Slight and balding, always sporting red socks and a red hat, he’s given to almost feverish disquisitions about the industry he loves. He insists he’s not “one of the smart guys” in his industry, but he has changed it profoundly.9

  To understand how, a little history is in order. In the 1980s, a movement had taken shape, spurred by Richard Stallman’s theorizing, to develop software based on UNIX, an operating system invented in the 1970s at AT&T Bell Labs and made available at no cost to anyone who requested a copy.

  Torvalds was one of thousands of programmers and enthusiasts who obtained a copy of UNIX and started tinkering with it. In 1991, he posted a message on a UNIX users’ bulletin board, modestly announcing that he’d developed an operating system from the UNIX kernel and offering his fellow enthusiasts a look. Before long, suggested improvements to Torvalds’ program, dubbed Linux, were pouring in. To capture the improvements and integrate them into the program, Torvalds and a few colleagues set up a loosely organized committee to vet all the suggestions and incorporate the most meritorious into the core program.

  By 1993, Linux had evolved to the point where it was robust enough to handle heavy-duty corporate applications. But corporations weren’t buying because the market was fragmented and confusing, with many competing versions of the uncopyrighted software floating around cyberspace. Enterprises like Yggdrasil and Slackware Linux tried to bring some order to the chaos, selling their own versions of Linux to interested buyers. Young ran an outfit called ACC Corp. that distributed their “free software.”

  The term was something of a misnomer. Linux was actually just extremely cheap. Purchasers would pay a small fee for a CDROM of the software, either directly from a developer like Yggdrasil or Slackware or from a distributor such as ACC. But unlike conventional companies such as Microsoft, Linux vendors didn’t charge a license fee scaled to the number of users of the operating system. Instead, Linux purchasers were free to mount their system on as many computers as they wished, at no additional cost.

  One of the companies that sold its version of Linux through ACC was an operation called Red Hat Linux. Impressed by the product, Young combined his company with Red Hat, becoming CEO of what was now called Red Hat Software and shifting the company’s focus from distribution of several flavors of Linux to direct sales of Red Hat’s Linux product.

  From his experience as a distributor, Young knew that the still-tiny market for Linux software was growing rapidly. But the business was going to hit a ceiling unless it could find a new business model.

  Young could see that the two dominant models then in existence were profoundly flawed. On one hand, there was the classical proprietary software model the big players such as Microsoft and Oracle employed. They sold their clients only the operating software, not the source code. All enhancements and modifications were in the hands of the software maker. As Young likes to say: “Buying proprietary software is like buying a car with the hood welded shut. If something goes wrong, you are not able to even try to fix it.”

  The providers sold their proprietary product at a high price—for example, $209 for Microsoft’s important release of that period, Windows 95, with gross margins above 90 percent—serviced it at a high price, and released regular upgrades that customers had to buy if they wanted to capture the software company’s incremental improvements.

  Young has nothing but scorn for this way of doing business. “If you ran into a bug that caused your systems to crash,” he says, “you would call up the manufacturer and say, ‘My systems are crashing.’ And he’d say, ‘Oh, dear.’ What he really meant was, ‘Oh, good.’ He’d send an engineer over at several hundred dollars an hour to fix his software for you that was broken when he delivered it to you, and he called this customer service.” In opposition to this problematic way of doing business was the free software model employed by Slackware, Yggdrasil, and Red Hat itself.

  Young didn’t think either model had much of a future. The proprietary software model, in his view, was an inefficient and obsolescent way to develop software. But the “free software” model was problematic as well. “You couldn’t make any money selling [the Linux] operating system,” Young says, “because all this stuff was free, and if you started to charge money for it, someone else would come in and price it lower. It was a commodity in the truest sense of the word.”

  If Red Hat was going to be something more than a low-margin distributor of a commodity product, it would have to find some way of adding value to Linux that didn’t involve improving the code. That meant finding something salient about the Linux business that other programmers and distributors had overlooked.

  Young found that point of salience in corporate buying habits. Young realized that because big companies are making decisions that they will have to live with for ten or even twenty years, they want to buy from the industry leader. “If Merrill Lynch was going to choose someone to support their use of Linux,” Young says, “they weren’t going to choose the number-two Linux developer. They were going to choose the number-one.”

  Proceeding from that insight, Young discovered a causal relationship between sales level and sustainability that his rivals had overlooked. “If we were the number-two brand, we did not have a business of any value whatsoever. If you’re going to sell free software, you’d better be the number-one player at it if you hope to get any leverage in the marketplace.”

  How, then, could Red Hat establish itself as the Linux market leader in the eyes of corporate users? By imposing order and control on the chaotic process by which improvements to Linux are developed and captured. But let Young explain it himself:

  A typical Linux operating system, whether Red Hat or Slackware, is some 800 to 1,000 different packages all compiled together. Those packages are maintained by different teams of people, and those people update those packages once or twice a year. If you accept, for argument’s sake, that they get updated once a year, that’s closing in on three updates a day. You, as a systems administrator for Merrill Lynch, have to keep track of all those updates and deploy them across all of your Linux servers if you’re going to use Linux.

  If you use Microsoft, you don’t have that problem. Microsoft delivers you a new version once every six months and they tell you how to install it. You have this nice, safe, controlled deployment of your new software.

  You can see that serious corporate users weren’t going to use [Linux]. Although the price was right, they wouldn’t save money if they used free software where their systems administrators had to track all of the random updates.

  But they would save money—and have a more stable, reliable operating system than Windows—if there were a way to manage the flood of updates. Red Hat would make itself invaluable to customers by taking on that management task.

  Young was pleased with this resolution, which blended the best elements of the two competing business models into a new way of doing business. But there was a hitch. A big one. Corporate customers wouldn’t buy Red Hat’s Linux unless Red Hat was the clear leader in the Linux space.

  Red Hat would have to find its way onto every hard drive in corporate America. To get there, Young’s programming team rewrote the Red Hat version of Linux so that it could be distributed over the Internet instead of via CD-ROM. Then Young told his team, “We’re going to put it up on every FTP [File Transfer Protocol] server we can find on the Internet everywhere in the world, and we are going to encourage people to download it for free.”

  Young’s goal was, he says, “to give [Linux] away more efficiently” than any of his competitors. It was a risky move. Red Hat was sacrificing all the potential revenue it stood to earn from its new release of Linux. But that was the price of making Red Hat’s version of Linux the de facto standard. In a stroke, Red Hat’s Linux became legitimate in the eyes of the corporate users.

  In 1999, Red Hat went public, and Young became a billi
onaire in the first day of trading. By 2000, Linux had captured 25 percent of the server operating system market, and Red Hat held over 50 percent of the global market for Linux systems. And unlike the vast majority of the dot-com era’s start-ups, Red Hat has continued to grow.

  What made the creative resolution of Red Hat possible? Young treated the existence of unattractive alternative models exactly as we would expect an integrative thinker to do. He recognized that the existing proprietary software and free software models weren’t reality; they were simply the accepted models for coping with dynamics of the software business. Second, he didn’t rest until he found a new business model that was clearly better than the existing alternatives. Third, he took personal responsibility for figuring out a new way to compete, instead of accepting the choices offered him. Fourth, he read the existence of unpleasant trade-offs as a signal to rethink the problem from the ground up. In doing so, he found clues to what was salient to corporate software buyers and gained a key insight into the causal relationship between industry leadership and prosperity. His path to a creative resolution was remarkably similar in its structure to the paths forged by Sharp and Lafley.

  The Buzz Machine That Ate Toronto

  Piers Handling fits the role of film festival director to a tee. He has lived in the world of film from a young age, and it’s a world he tears himself away from only reluctantly. At gala parties, he appears to be wishing he was watching yet another of the festival’s most obscure films rather than mingling with the stars. But he is more than an afficionado of the art of film. He is also a keen student of the business of film.

  When Handling joined the staff of the Toronto International Film Festival (TIFF) in 1982, the festival was the very model of a struggling start-up.10 Launched in 1976, it was barely international and in its own country it trailed the Montreal World Film Festival in prominence and prestige.