When Does the Early Bird Get the Worm?
Beyond envisioning what to bring together and how to bring it together, when do you bring all the elements of your great vision together? There is a window of opportunity for every innovation. For many managers, the siren song of early-mover advantage can seem irresistible. After all, being early means unobstructed access to the market, the ability to establish an industry standard, not to mention free publicity and the resulting brand recognition that can create lasting privilege with customers. Early movers such as Amazon (online retail), Xerox (copiers), and eBay (online auctions) enjoyed all these tantalizing advantages and, to this day, lead their respective industries.
Although conventional wisdom applauds the pursuit of early-mover advantage, the complexity of innovation ecosystems can overturn this logic. The usual focus on getting products to market creates a dangerous blind spot when it comes to timing entry: the early bird may get the worm, but the second mouse gets the cheese.
In this chapter we will explore the different implications of being a pioneer or a late entrant, directly linking performance to where in the ecosystem your biggest challenges lie. If the biggest obstacle is overcoming the execution challenge, getting to market before your rivals can create great advantage. But in a world of dependencies, the benefit of preempting the competition is directly related to your co-innovators’ readiness with their offerings. If co-innovators are late (recall Nokia’s long, disastrous wait for 3G partners), striving to be first can be pointless or, worse, detrimental to your effort. You may be the driver that races ahead just to wait at the red light alongside everyone else. Given the complexities of innovation ecosystems, simply asking yourself if you should be an early or late entrant will no longer suffice. The smarter question now is: under what circumstances should I be early or late?
To embark upon our exploration, consider the evolution of the portable music player industry. At first glance, it defies logic. In 1979, Sony’s pioneering Walkman, the world’s first portable cassette player, was able to benefit from a first-mover advantage that lasted over thirty years. However, with the same customer set and value proposition, the world’s first portable MP3 player, SaeHan’s MPMan, introduced in 1998, enjoyed no early-mover advantage. It was Apple, which launched its iPod three years after the first mover, that finally dominated the industry. How did this happen?
First Mover: Advantage or Disadvantage?
In 1978, engineers at Sony successfully married a compact playback device with lightweight headphones to create the prototype for a product that would become a worldwide hit. In 1979, the Walkman was introduced in the Japanese market, selling out its entire production run within the first three months. Combining a technology based on a previous innovation—Sony’s Pressman recorder, a pricey device aimed primarily at journalists—with an aggressive marketing campaign targeting a young demographic, Sony pioneered the market. Consumers, already accustomed to playing cassettes on boom boxes and car stereos, were eager to emulate the happy youths of Sony’s ads—roller-skating, picnicking, and jogging, all the while listening to their favorite music.
Soon Sanyo, Panasonic, Sharp, Philips, and other electronics leaders joined the fray. Some rival products boasted additional features like AM/FM radio, water resistance, and recording capabilities, but none captured the popular imagination of the Walkman.
For a decade after its launch, Sony’s Walkman retained a 50 percent market share in the United States (46 percent in Japan) in a space teeming with competitors, even as it enjoyed a price premium of approximately $20 over rival offers. Showcasing the power of first-mover advantage, Sony’s pioneering technology set the benchmark for everything that came after it. And the media’s hyperfocus on the Walkman meant that followers were seen as mere imitation. Like Kleenex, Band-Aid, and Xerox, the Walkman name became the catchall term for an entire product category.
Jump ahead to the late 1990s, when the sun had set on cassettes as the preferred music delivery format in favor of compact discs and, for the technologically savvy, digital MP3 files. So too did the Walkman’s relevance begin to fade in favor of Sony’s follow-up device, the Discman. But electronics firms around the globe were betting that the compact disc would soon follow the cassette into extinction. Who was going to capitalize on the rise of the MP3? Which MP3 player would get there first and become the next Walkman?
In 1998, South Korea’s SaeHan Information Systems created the world’s first portable digital audio player, MPMan. That summer, the handheld MPMan, boasting 32 MB of memory and initially priced at $350 (a 64 MB version retailed for $500), was launched in the United States. MPMan was soon joined by a panoply of MP3 players from established firms as well as start-ups. Diamond Multimedia’s Rio PMP300, less pricey at $199, led the way a few months after MPMan’s launch, followed by Compaq/HanGo’s Personal Jukebox. The next year, Creative Technology introduced its Nomad, and Sensory Science launched Rave. In 2000, Samsung and Sony both joined the race. The offerings varied in both design and capacity. (Most notably, the Personal Jukebox, launched just months after MPMan, was the first MP3 player to use a hard drive rather than a flash drive, enabling it to increase capacity to a then unheard-of 6 GB.) By the launch of the iPod in 2001, there were approximately fifty portable MP3 players available in the United States—and no firm had achieved anywhere near the dominance that the Walkman had enjoyed twenty years earlier. Although IDC predicted sales of MP3 players to reach 9 million by 2006, unit sales in 2001 were only 248,000.
SaeHan saw the market, delivered the product, embraced the right file format, and did it all before the competition. But unlike Sony, it did not enjoy a first-mover advantage. Why was first-mover advantage so large in the portable music player industry in one technology generation but completely absent in another?
First, we must ask: what does it take for a portable music player to create value? By design, these devices are basic; a customer need only add batteries and the music itself. For the Walkman, this was simple. Cassettes had been in wide use in home and car stereos since 1972. And by 1979, retailers in every town and every city sold them alongside vinyl records.
But in 1998, the story was very different for MP3s. You couldn’t purchase them in traditional retail settings, and although the press had begun sounding the recording industry’s copyright fears, the reality was that few people were actually illegally downloading songs off the Web. Remember 28.8 Kbps dial-up modems? Downloading an album—legally or not—could be a multihour affair, punctuated by the frustration of dropped connections.
In 1998, the world was on the verge of the Internet age, but it wasn’t quite there yet. At the time, Christopher Mines of Forrester Research predicted: “Once consumers get a taste of high-speed Internet access at home, they’ll never go back to dial-up.” He was right, but most people hadn’t tasted broadband yet, meaning illegally downloading music was a time-consuming process mainly embarked upon by tech-savvy college students. Few adults who could actually afford the $200–$300 MP3 players were going to spend hours painstakingly building their MP3 collection through downloads or transferring songs from their well-stocked CD racks. According to the digital rights management team a2b, “Currently the only protection mechanism in place for CDs is the sheer bulk of the data. Until the bandwidth available to most consumers increases, it will be cumbersome to move even compressed CDs around.”
While portable cassette players could create value as soon as they were launched due to the widespread availability of cassettes, MP3 players were not able to truly create value until the content was broadly available. The MPMan and its kin may have gotten there early, but they faced crucial co-innovation challenges that stymied mass adoption. It didn’t matter that MPMan was first—it wouldn’t have mattered if it was sixth, twenty-third, or forty-second. Without the extensive access to MP3s and broadband, the value proposition could not come together.
The iPod Wins, Three Years Late
The MP3 player market did eventually consolidate around a dominant product, Apple’s iPod. But the iPod, launched in late 2001—three years after the MPMan—was anything but a first mover. How can we understand the iPod’s success despite its delayed entry?
In 1997, the late Steve Jobs returned to Apple, the company he had co-founded as a college dropout, as interim CEO. As the Internet bubble grew, Apple was hungry for growth. Only a sliver of computer users had embraced its Mac offering. In 2001, Jobs noted: “Apple has about 5 percent market share today. Most of the other 95 percent of computer buyers don’t even consider us.”
Jobs was a pioneer of the convergence of digital and media. It is inconceivable that digital music was not on his radar. Still, in Jobs’s early tenure as CEO, and despite his company’s need to create a new avenue for growth, he refused to jump on the MP3 bandwagon. Asked in 1998 about his growth strategy, “Jobs just smiled and said, ‘I am going to wait for the next big thing.’” In 1998, as MPMan and others launched, he didn’t budge. In 1999, as Shawn Fanning introduced Napster (illegally), unleashing a vast cata-log of free MP3 content to the world, Jobs still didn’t move. In 2000, Jobs finally decided to start the process of developing an MP3 player.
In 2001, the bubble burst, and in the bleak outlook for all things Internet related the MP3 player was not spared. As a spokes-woman at Intel sadly noted, “With some of the consumers we have talked to, it sure seems like things are slowing down in the MP3 player market.” It was at this dreary moment that Steve Jobs finally launched his player in the market. Why? What had he been waiting for? What did he know?
Jobs knew that, on its own, an MP3 player was useless. He understood that, in order for the device to have value, other co-innovators in the MP3 player ecosystem first needed to be aligned. And in October 2001, when Apple announced the iPod, those pieces were solidly in place: both MP3s and broadband were finally widely available.
Apple waited, and then waited some more—until it finally made its move, putting the last two pieces in place to create a winning innovation: an attractive, simple device supported by smart software. The first-generation iPod for Macintosh retailed at $399, had 5 GB capacity, and could store up to 1,000 songs. It boasted an intuitive interface design and was, for its time, lightweight. But the value of the device was cemented by its seamless integration with the iTunes music management software. Organizing and managing your MP3s was finally a breeze. But unlike other MP3 players, which used a USB cable to transfer music files from computer to device, the iPod relied on a faster, built-in FireWire port, a standard feature on all newer Macs but available on only a fraction of PCs. (FireWire was a favorite feature of Jobs’s, given that it is significantly faster than USB—users could transfer 1,000 songs in less than ten minutes.)
Despite being available only for Mac users, the iPod was the fastest-selling MP3 player to ever hit the market. That first holiday season saw 125,000 units sold. Analysts called it “the hit of the holiday season,” “revolutionary,” and “brilliant.” In March 2002, Apple doubled the iPod’s disk capacity with a 10 GB offering and in July announced the much-anticipated Windows-compatible iPod. Sales were on the rise: by the end of the year, consumers had purchased over 600,000 iPods. Still, Apple held only 15 percent of the digital player market.
In April 2003, Apple announced the iTunes Music Store, an online retail hub where customers could browse and purchase music for $0.99 per song (or $9.99 per album). Crucially, Apple’s stringent FairPlay digital rights management system meant that record labels could finally breathe easy that the intellectual property of their artists would be protected. And, iPod users could hold their heads up high knowing that their music was (at least potentially) legally acquired. “Consumers don’t want to be treated like criminals, and artists don’t want their valuable work stolen. The iTunes Music Store offers a groundbreaking solution for both,” Jobs stated. The iPod, of course, still did an excellent job of playing pirated content as well.
At launch, the iTunes Store had a collection of 200,000 songs from major labels such as BMG, EMI, Sony Music Entertainment, Universal, and Warner. In its first week, Apple sold over 1 million songs. By 2005, its library had grown to 1.5 million songs. Although Apple would make scant profit from selling songs at $0.99 (it had sold nearly 8 billion songs by the end of 2009, but with an estimated operating margin of 10 percent on song sales, that translates to only $800 million in operating profits over six years—trivial when compared to its earnings from the over 220 million iPods it had sold by 2009), the iTunes store gave the iPod legitimacy in a world of shady MP3 downloads.
Soon after the launch of the iTunes store, Apple released an iPod that was compatible with both FireWire and USB cables, opening up its best-selling device to the entire world of PC users. It was at this point that Apple cemented its dominance.
According to NPD Group, sales of portable CD players were still more than double those of MP3 players during the holiday season of 2004. But between the third quarters of 2004 and 2005, sales of the iPod had leaped 616 percent.
As the years passed, Apple kept on innovating. The iPod became sleeker, smaller; it had more and more memory. For consumers, this meant there was always a fresher, more appealing iPod to own. As the same customer base kept repurchasing new and better iPods, Apple’s profits soared: by 2008 it had captured 48 percent of the MP3 player market share. SanDisk’s Sansa MP3 player was the iPod’s closest competitor with 8 percent market share.
Few would deny that the iPod is a great product, surpassing any other MP3 player offering. But is it six times better? Why is it that this technology succeeded to such a great extent where others—established, smart others—barely made any progress at all? Apple was, after all, three years late. But perhaps this logic should be flipped: perhaps everyone else was three years too early. As we’ll see again in chapter 8 when we explore the iPhone, Jobs tended to be late for everything because he wanted everything to be ready for him. Reflecting on catching technology waves in 2008, he said, “Things happen fairly slowly, you know. They do. These waves of technology, you can see them way before they happen, and you just have to choose wisely which ones you’re going to surf. If you choose unwisely, then you can waste a lot of energy, but if you choose wisely, it actually unfolds fairly slowly. It takes years.” Jobs’s discipline paid off. In the three years between the launch of the MPMan and the iPod, each element in the MP3 player ecosystem turned from red to green. Instead of waiting at the red light with everyone else—wasting precious resources and time—Apple drove right on through a green light toward victory, becom-ing, according to the Economist, “the Walkman of the early 21st century.”
Early-Mover Advantage in Products vs. Ecosystems
When we think about early-mover advantages we tend to think in product terms. The outsize profits of winning exemplars—Birds Eye in frozen foods, Xerox in copiers, Intel in microprocessors, Bayer in aspirin, and DuPont in nylon—are held up as proof of the prize that goes to those who “get it right first.” Early movers can capture mind share and establish standards when the field is still uncrowded; they can block followers through patents and preemptive acquisition of scarce resources; they learn from their production experience, optimizing processes to decrease costs while increasing quality; and they benefit from switching costs when shifting vendors requires customers to retry and retrain on an innovation.
But because early movers, by definition, pioneer new market space, they are exposed to greater uncertainties than laggards. There is no prize for those who get it wrong first. While early movers flail about trying to find the right product architecture to position in the right market segment, it is the latecomers that reap the benefits of pioneering failures. And so, losing exemplars—Ampex in video recorders, Raytheon in microwave ovens, Chux in disposable diapers, and MITS in personal computers—are held up as proof that pioneers are the ones with arrows in their backs.
These dueling lists of examples and counterexamples make one thing clear: the relevant question for managers is not whether early movers can be advantaged but, rather, under what conditions they will be advantaged. We can better understand the relationship between entry timing and advantage by using the First-mover Matrix (see figure 6.1).
The debate about first-mover advantage has hinged on the likelihood that the early mover enters the market with the right product. In our terminology, the focus has been on execution. And the expected size of the advantage is directly linked to the size of the execution challenge: being first to market with an easy-to-implement, and hence imitate, product (a new office chair design or a vitamin-enhanced beverage) will yield a lesser prize than pioneering a harder-to-implement product (a new engine design or a supercomputer). In a product world, this makes sense.
Figure 6.1: First-mover matrix for determining relative advantage from early entry as a function of the level of innovator execution challenge and complementor coinnovation challenge. (Adapted from Adner and Kapoor, 2010.)
This is the logic behind quadrants 1 and 2 in the first-mover matrix. Quadrant 1 (first in gets the win) is the baseline level of early-mover advantage—how big a win we can expect the early mover to enjoy. The greater the execution challenge you need to overcome to deliver the product, the harder it is for followers to ride on your coattails, and the bigger the reward for being early. This is the situation in quadrant 2 (winner takes more).
In a product-based world, this makes sense. But in an ecosystem world, this logic is incomplete. In an ecosystem world, as we have seen time and again, delivering a brilliant product that the competition cannot match is not enough. We need to make sure that all the other elements that our product requires to create its value are in place as well. When the co-innovation challenges in the ecosystem are high, this leads to the “hurry up and wait” syndrome of quadrant 3. Recall Nokia in 3G, Sony in e-books, and SaeHan in MP3 players: all raced to launch a great product into the market, and all fell flat. Why? Because being first doesn’t matter if the critical complements aren’t ready.
Over time, as firms and partners work through their challenges, the natural tendency of the ecosystem is to drift toward the upper left segment of the matrix. When execution and co-innovation challenges are both high at the time of entry (quadrant 4), the advantage of the entrant will depend on which challenge is resolved first—whether the system shifts to quadrant 2 (better) or quadrant 3 (worse).
The First-Mover Matrix at Work
How big a deal is this really? And can we actually get guidance from this approach? To answer these questions, I undertook a four-year project focused on the semiconductor lithography industry with Rahul Kapoor, my coauthor on the research studies. We used an ecosystem lens to explore early-mover advantage during the introduction of nine different technology generations between 1962 and 2005. The results, and their implications, are striking.
Semiconductor lithography is the beating heart of the digital revolution, the principle technology that has powered the improvements predicted by Moore’s law (the prediction that the number of transistors that can be placed on a chip will double approximately every two years). It is the process by which circuit designs are imprinted on a semiconductor wafer—the process that allowed Intel to fit 3 million transistors on a Pentium chip in 1993 and now, thanks to the improvements we will explore, 2.8 billion transistors on a Xeon chip in 2011.
Semiconductor lithography is a replication process that, in simplified terms, is a lot like old-fashioned chemical photography. At its core is the lithography tool (the camera body) into which are integrated two critical components: an energy source (flash) and a lens system (lens). For a lithography tool to create value, however, it must be used in conjunction with two key complements: a circuit mask, which holds the circuit design that is to be replicated (the object to be photographed), and a resist (chemical developer) that will react when exposed to the energy source to replicate the circuit image on the mask onto the silicon wafer (the photo paper).
Throughout its history, the driving goal of semiconductor lithography has been to increase image resolution (this is because resolution determines the extent of miniaturization and circuit density that can be achieved by a semiconductor manufacturer in its microchips). In the 1960s, resolution was in the range of 7,000 nanometers (0.007 mm). In 2011, it is 22 nanometers—a 300-fold performance improvement. Underlying this phenomenal pace of improvement have been a series of technology revolutions in the design of the tool as well as in the elements that are brought together in the lithography process. But despite all the changes to the different elements, the basic structure of the lithography ecosystem has not changed in over fifty years: the lens and the energy source are essential components that are integrated into the tool by the toolmaker (firms like Canon, Nikon, and ASML), while the mask and the resist are critical complements that are brought together by the customer (semiconductor manufacturers like Samsung, Toshiba, and Intel).
Figure 6.2: The semiconductor lithography equipment ecosystem. (Adapted from Adner and Kapoor, 2010.)
New generations of lithography tools are marked by transitions to more sophisticated architectures that provide greater control and repeatability, shifting from mechanical to electromechanical to electronic controls; from reflective to refractive light management; and incorporating digital logic throughout the device. What is interesting is that the innovation challenges at different points in the ecosystem varied quite dramatically across the different generations. Sometimes, the lens technology used in one generation could simply be reused in the next. At other times, entirely new lens materials and manufacturing processes would need to be developed to pass a new wavelength of light. In some generations, the energy source could simply be replicated, while in others, new categories of lasers needed to be invented. Similarly, some generations required major innovations in mask and resist while others did not. Thus, across the different generations, depending on how many different elements needed to be innovated and how great a change each required, we can characterize a range of co-innovation challenges that feed into the firm’s ability to deliver its own innovation (the extent of execution innovation challenges) and co-innovation challenges that affect the ability to deliver the full value proposition to the end consumer (the extent of complementor co-innovation challenges).
It is the combination of this variance in the levels of execution and co-innovation challenges in conjuction with the stable ecosystem structure that allows us to compare outcomes across the nine technology generations and draw conclusions about the industry.
When Do Early Movers Win?
The tool manufacturers and their partners have a deeply held belief in the power of first-mover advantage. And the classic product-based indicators all suggest it should be high: there is a lot of coordination with customers and industry organizations like SEMATECH about the desired technical specifications of new technology generations, so uncertainty about requirements of the “right” product is low. Early movers gain significant learning curve advantages, reducing their costs and increasing their quality with rising production volumes. And, most importantly, because the semiconductor manufacturers customize their production techniques to the specific tool they buy, there are high costs to switching tool suppliers: once a customer adopts your tool, they are very unlikely to switch to a rival.
But when we applied an ecosystem lens in conjunction with statistical tools to isolate the relationship between how early firms entered and how much of a market share advantage they enjoyed, we identified a key contingency. In generations where complementor co-innovation challenges were low, the traditional product-based logic held true. In the baseline case of relatively low execution challenges (quadrant 1) early movers enjoyed a distinct advantage: a pioneer with a three-year head start could expect 3 percent higher market share advantage over the latecomer in total sales over the life of the generation. Keeping in mind that, in 2009, a 1 percent market share advantage in the new lithography tool generation (DUV 193i) equated to $28.4 million in annual sales, and that tool generations continue to be sold for many years, it is clear that the stakes are high. And it is assuring to see that the industry belief in first-mover advantage is fully supported in the baseline scenario.
When execution challenge is high and co-innovation challenge is low (quadrant 2), early movers are at an even greater advantage: a three-year head start yields a 6 percent market share advantage to the pioneer. Here, the belief in first-mover advantage is on even sounder footing.
The most intriguing finding, however, is the one that corresponds to the low execution challenge/high complementor co-innovation challenge scenario of quadrant 3. Here, the product-based logic breaks down in the face of ecosystem dynamics. Here, the first firm to overcome its execution challenge needs to wait. Here, the value proposition remains incomplete until all the complementors solve their problems as well. Here, the pioneer has no advantage. In fact, the pioneer is at a slight (–0.3 percent) market share disadvantage relative to laggards.
The implications of these findings are powerful. What they reveal is that it is fundamentally wrong to think about first-mover advantage as a fixed characteristic of an industry—any industry, including your own. Instead, we see that the extent of first-mover advantage depends on the nature of challenges in the ecosystem.
When development efforts are accelerated, costs rise exponentially. Whether or not they win a prize, a certainty for firms that race to be first is that they will pay a higher price for the attempt. Wise innovators will locate the nature of their endeavor on the first-mover matrix and adjust their speed accordingly. Aggressively pursuing first-mover advantage makes most sense when the innovator’s execution challenges are high. This is the occasion for greater efforts because this is the time of greater rewards. It also shows that when complementors’ co-innovation challenges are high, the rewards for pioneering are low. This is the occasion for watchful patience; the time for active preparation, but not for aggressive launch.
Smart-Mover Advantage
If you allow yourself to look beyond your own innovation challenges at the entirety of the value proposition, timing clues abound. It shouldn’t take a visionary to see that demand for MP3 players won’t take off until users have easy access to the actual product. It doesn’t take a crystal ball to foretell that sales of a new lithography tool generation will flounder until the key complements are ready. And it doesn’t take a lot of effort to figure out whether co-innovation challenges for a new innovation are high or low: in the semiconductor lithography industry we found articles discussing details of co-innovation challenges that were published as long as six years before the technology was commercialized.
Nevertheless, it does take willpower to resist the natural urge to rush forward. Self-imposed delays go against the grain of most leaders and organizations. But in an ecosystem world, the costly risk of “hurry up and wait” is a call for discipline and perspective. The ecosystem is a puzzle that needs to be assembled. The prize does not go to the first player to put down the first piece—because nothing happens until the puzzle is complete. The prize is only awarded after someone puts down the last piece.
Steve Jobs’s iPod journey is an exemplary illustration. Jobs had the perspective to realize that digital music players were in a different quadrant than portable cassette players. The Walkman started life in quadrant 1. The MPMan started life in quadrant 3. Jobs understood that the natural trajectory of challenges is toward the upper left of the first-mover matrix. The “system” works to resolve co-innovation challenges, while industry rivals figure out execution. His insight was to “surf” the co-innovation wave, knowing that its challenges would be resolved over time. His brilliance was to wait to expend his energy on the execution challenge.
Jobs waited, but while he waited, he also differentiated. Apple didn’t launch the iPod as a product. In combination with its iTunes music management software, the iPod was a solution. As the iPod’s co-innovation risks faded away in 2001, Jobs launched. With his proprietary hardware-software combination, he didn’t just put down the last piece, he put down the last two pieces. And he made sure they interlocked. By shifting to offering solutions, Apple increased the execution challenge for itself as well as for everyone else, effectively lowering the value of competitors’ previous efforts and increasing the barrier for rivals to achieve future success. In doing so, he thrust the iPod into quadrant 2, to capture smart-mover advantage. (In chapter 8 we will see Jobs deploy a variant of this strategy with the iPhone and the iPad tablet.)
Smart timing—for a novel product or an entirely new technology—requires a sober look at who and what else is out there to help or hinder the effort. Understanding these dynamics informs decision makers, helping you dial up or dial down your level of aggression going forward. As ecosystems become ever more pervasive, it is critical to ask the question: “Are we in a Walkman world or in an iPod world?” As you choose your moment to strike, looking around at both the structure of the ecosystem and your role in it, you are now in a better position to win. There’s still uncertainty, of course. No triumph is preordained. But understanding your effort from an ecosystem perspective stacks the odds in your favor. The blinders are off. Welcome to the bigger picture.
