4. The Long Tail of Expertise

“We sold more books today that didn’t sell at all yesterday than we sold today of all the books that did sell yesterday.”

—Josh Petersen, Amazon employee¹

Overview

Whoever thought the power function would become an icon of popular culture? Chris Anderson’s and Clay Shirky’s terrific writings have achieved just that. The power function, once relegated to workhorse status in statistics, mathematics, and modeling, now graces the cover of popular books and appears in hip, trendy magazines. Its newfound status results in the ease in which it explains the market phenomenon of businesses such as Amazon.com.

The Long Tail concept has found applications in marketing, inventory strategy, Internet statistics, research, and media sales. It is invoked here, qualitatively, as a means to illustrate some observations made on a daily basis at InnoCentive and which are more broadly applicable to innovation in general and the entire notion of expertise. Step one is for you to have a clear picture of the image in mind as the arguments are discussed. A power function plot, with its long tail is shown in Figure 4.1.

Figure 4.1. A plot of a power function showing the very long tail that approaches, but never quite reaches, the baseline as it extends forever to the right.

For the moment, concentrate only on the curved gray line that swoops down from the upper left and disappears off the chart on the lower right. The right portion of this curve is what has been referred to as the long tail. For this discussion, the left side will be called the head, in contrast to the tail. Using the bookselling example, you can label the y-axis sales volume (or popularity, if you prefer). And the x-axis would be simply each book in the world arrayed from most popular at present on the left to least popular on the right. The head of this curve would thus contain those books on The New York Times bestseller list, and anything mentioned on the Oprah Winfrey show. If you are a small bookseller and have room for only 100 titles, naturally you would want the 100 books represented on the leftmost part of this curve—for the simple reason that they would generate more revenue per inch of shelf space than any other 100 books. The point made by Anderson and the long tail argument is this: On any given day, customers are more likely to want to purchase more books from the hundred-and first-popularity on down than they are books among the 100 most popular. But because you don’t know precisely which of those less popular books are likely to be purchased that day, you can never make a rational stocking decision other than the one already made, which is to stock the most popular books.

These arguments are, at the same time, obvious and subtle. Businesses make such rational choices all the time, and because of their obviousness, their subtlety is never examined.

Defining and Hiring Experts

Expanding on the ideas briefly introduced in Chapter 3, “A New Innovation Framework,” the way in which this long tail argument creeps into hiring practices—and even the notions by which expertise and qualifications are defined—will be discussed. This will be followed by an examination of the subtle consequences of continuing to practice in that way.

Most people feel that they are quite clear on their definition of what qualifies someone to be called an expert. It is easy for them to say that Stephen Hawking is an expert in cosmology and I am not, that Jared is an expert woodworker and I am not, that I am an expert chemist and Dwayne is not, and that Dwayne is an expert mathematician and I am not. Such statements feel well-defined and easily justifiable on the basis of degrees, training, experience, publications, and contributions. But, even as I exclude myself from the category of expert plumbers, I would argue that, although I know less than those experts, I do know more than many other people. And there are those who know more than me but may not quite reach the level of “expert.”

Recognizing the continuous nature of these definitions, you can realize that no boundary line exists and that experts merge seamlessly with novices and even those who can be described only as totally ignorant on the subject of plumbing.

The first reaction to an argument like this is understandably, “So what?” So what if it’s a continuum; so what if you have to draw some arbitrary line? So what if you make the measured decision to use your neighbor’s Saturday plumbing skills to put in a new sink instead of hiring an enormously qualified, and probably overpaid, expert? Baseball pitching skills probably follow the same continuous pattern, and the Yankees are content to look for the pitcher residing on the leftmost of the curve and offer him a sufficient salary to entice him to join the team.

All hiring decisions are generally presented as if it were no more complex than the recruiting of a new pitcher or making that Saturday afternoon plumbing choice. But in the case of full-time permanent employment offers, it is often quite a bit murkier. One typically frames the recruiting process in those same simple terms: “Find me a top-grade CEO,” “Find me an expert electrical engineer,” or “Find me an outstanding market analyst.” But the commitment of hiring a permanent employee is actually quite different. You might even have an immediate situation driving your choice. Perhaps a new CEO is needed to help enter a new market sector or to manage through an explosive period of growth. Perhaps an electrical engineer is needed to debug stray radio frequency signals that are compromising a new product release. And perhaps a market analyst is needed to immediately prepare reports on newly acquired telecom capabilities. But it is almost invariably true that if these individuals are hired, they will remain with the organization after, sometimes long after, those immediate needs are satisfied.

Executives, human resource departments, and recruiters are smart enough to recognize that they are seeking a broader sort of individual, one who can solve not only the issues of immediate concern but also the issues, appropriate to their expertise, that will surface in the future. Unfortunately, because these are issues and challenges that haven’t surfaced yet, they cannot be named, and they cannot be specifically recruited for. And thus, the practical consequence of “expertise” within the business world becomes awkwardly defined as “the ability to solve a domain-specific problem posed in the future.” It’s not a real crisp definition; it’s not a definition you would look forward to building metrics around. It’s just the practical reality of what you’re actually doing when you hire someone.

You hire an editor because you believe she will do a good job of editing manuscripts that haven’t been written yet; you hire a chemist because you believe she will do a good job of synthesizing compounds, that not only do not exist yet, but also which, at present, you do not even know you want; and you hire a bookkeeper to compute unmade sales, unmade costs, and use reporting rules yet unwritten by accounting standards bodies.

So how, exactly, can you measure a job candidate’s fitness for a future need? The answer to this question is why this entire book began with a quote about gambling and horse racing: “The race is not always to the swift, nor the battle to the strong, but that’s the way to bet.” Employers are placing bets on employees, and their recruitment filters are designed to identify “the swift and the strong.” Is that not what a race handicapping system is designed to do? It is designed to use measurements that can be made today to enable you to predict the horse and jockey that will win tomorrow. The values that can be measured today, such as recent track times, health, weight, experience, and so on, are not the values crucial to your success; that is, all you really want to know at the moment it counts is, “Did the horse win?” Those other properties you measure aren’t what you actually want to know; they are simply what you can know, and their use is predicated on a belief that they are related, in some predictable fashion, to what it is you actually want to know. This is how races are handicapped; this is how employment is handicapped, this is how expertise is handicapped. The measures may be different, but the principles are the same.

It’s time to take another look at the long tail and its head of expertise. Figure 4.2 fills in this newly articulated definition of the y-axis, relevant to the employment process. There are also zone labels and shading on this graph to enable you to more conveniently track the arguments that follow. You recognize the ultimate unmeasurability of this Y-value, but engage in a little thought experiment and pretend that such a measurement is possible for the sake of the following.

Figure 4.2. Three zones of the long tail curve. Zones A and B are our defined “experts.” It is them we seek to employ to bring us innovations. But we always employ a minority as in Zone A, while the majority of experts work elsewhere, Zone B. Generally ignored is the zone of greatest overall area (cumulative probability in this graph), Zone C. See text for tapping this enormous resource for innovation.

Via the collation of degrees, prior jobs, recommendations, grades, and so on, you can assume that you have hired well. You have recruited employees that lie to the leftmost of the chart. You have stocked your limited bookshelves with The New York Times bestsellers and the book recommendations of Oprah Winfrey. But now say it again. You have stocked your limited bookshelves with only The New York Times bestsellers and the book recommendations of Oprah Winfrey. Of course you have only stocked some of each of these key assets. As already pointed out—most of the smart people don’t work for you. You have hired those shown in Zone A in Figure 4.2. The skilled employees your competitors have collectively hired are those shown in the larger Zone B.

The Untapped Potential

Executives listen to Bill Joy and worry about the relative size of Zone B; he’s right—most of the smart people work for somebody else. And you can worry that your employees are being wooed by those competitors, even as you woo the employees of your competitors. All the while, Zone C, the long tail, remains out of your thoughts. And why shouldn’t it remain unthought-of? There’s only so much shelf space. Why waste it? This is what Chris Anderson describes as the logic of scarcity. A logic that seems less and less applicable to the Internet age.

The logic of plenty says, “Look at Zone C!” It’s not just bigger than Zone A; it’s bigger than Zones A and B combined. The, at first dubious, and certainly nonobvious, conclusion is that there is more problem-solving capability among the “less qualified” than there is in the “more qualified” populations. How can that be? Simple. There are more people less qualified—many more people less qualified—many, many, many more people less qualified. The long tail of skills has always been—and is today—largely untapped and institutionally discounted. The tapping of this long tail, Zone C, in spite of the technical correctness of the arguments just made, still feels wrong; it still feels like a bad idea.

Why would you hire a computer hobbyist or a construction engineer, when you need the best electrical engineer possible? Of course you would not. It is the history of “innovating from within,” of cloaking your innovation processes in secrecy, that compels you to too readily equate commercial innovation with hiring innovators. Tapping into the enormous capacity of the long tail cannot be done via the hiring process—and therefore cannot be part of closed innovation. It’s not that history isn’t full of stories of problems solved by innovations coming from the tail, of breakthroughs put forward by the tail. Some have been mentioned and will be mentioned again: patent clerks rewriting cosmology, telecom engineers predicting solar particle storms, weekend handymen contributing to environmental cleanup, and patent attorneys designing new routes to poly-carboxylic acids. But wait! Some of these examples are probably not a familiar part of your college coursework. That’s okay, they’re everyday experiences for those who help the world tap into this long tail of knowledge and expertise—and replace the logic of innovation scarcity with the logic of innovation plenty. And you’ll find each of those stories in this book.

Perhaps one of the earliest, and most systematic, attempts to document the contributions of a portion of Zone C to product innovation is the effort of Eric von Hippel at MIT’s Sloan School of Business. Von Hippel is the author of Democratizing Innovation (referenced earlier). Using a long tail argument of his own, von Hippel defines an easily identifiable cohort of “nonexperts,” which in our imagery would be the occupants of the left side of Zone C—the head of the tail as it were. In other words, if you were to undertake the tapping of Zone C for ideas, von Hippel defines a protocol for rapidly identifying the potentially most productive members of that zone. von Hippel calls these key contributors “lead users.” They may lack engineering and design qualifications that product companies seek in their R&D staff, but they are the customers who buy the company’s products. Actually they are a special subset of customers; they are the ones who fearlessly dismantle them and tweak them for purposes of their own. They are the gearheads of the 1950s unbolting the stock carburetor from their newly purchased Mercury and ultimately extracting another 10hp from the engine. Because they are the buyers, and they can often be identified, they represent a readily tapped portion of Zone C that it is inexcusable for business to leave unexploited.

Tackling the Long Tail

Power functions have an interesting property known as self similarity, which is to say that if you take any part of the function, it still looks like the whole function. So as you peer into Zone C—and exclude for a moment Zones A and B—the lead users become the “experts” in the “head” of Zone C, and the remainder of Zone C becomes—yet again—the zone of greater cumulative opportunity—another long tail!

Although historical attempts have been made to tap the tail, and are recorded elsewhere in this text—for example, the Orteig Prize (for crossing the Atlantic), the Longitude Prize, and the Millennium Prizes in mathematics—the Internet has opened Zone C possibilities hitherto unimagined. Platforms have emerged that enable challenges and needs to be broadcast and potential solutions to be acquired from all three zones, with indifference to source.

In early academic analyses of such systems, this approach was dubbed “broadcast search” by Harvard Business School professor Karim Lahkani and his co-authors.² They have identified several Internet platforms, such as Topcoder and InnoCentive, and at least semi-quantitatively confirmed the Zone shadings in Figure 4.2, which is to say that a great many of the crowd-sourced successful solutions are originating from sources that would not have been placed, a priori, in the leftmost portion of the curve. As stated in one paper by Lakhani, along with Lars Jeppeson, Peter Lohse, and Jill Panetta, “Problem-solving success was found to be associated with the ability to attract specialized solvers with a range of diverse scientific interests. Furthermore, successful solvers solved problems at the boundary or outside of their fields of expertise, indicating a transfer of knowledge from one field to others.”³

It would be nice to have a direct “head to tail” comparison of the problem-solving effectiveness of the different zones. And several exist in today’s business literature. One such example appeared in the NASA case at the end of the previous chapter. A solution that had been sought for about 30 years within Zone A was solved in months when Zone C was tapped.

Perhaps the best, prospectively designed study was carried out under the direction of Tod Bedilion, director of technology management of Roche Diagnostics. Roche took a problem that had been worked on over many years by both Roche and its partners. The challenge was to quantify a flowing clinical sample. The challenge was then broadcast to researchers globally throughout Roche and after collecting proposed solutions, the challenge was then posted to the large InnoCentive network of diverse solvers. At the time of posting, InnoCentive was unaware that this challenge had the history just described. Over the posting period, the challenge was viewed by almost 1000 solvers and 113 proposals were submitted from all around the world. In an article by London Business School professor Julian Birkinshaw and the editor of Business Strategy Review, Stuart Crainer, Bedilion is quoted: “The proposals were incredible... In contrast to the internal network, rather than being one or two lines, many were multiple pages. Some people had done experiments. There were diagrams. There were drawings that filled an entire notebook. We would have been delighted if we could have got much of the work out of our own research organization... I couldn’t put ten people in a room and have a brainstorming session or a seminar for two days for the same cost with all the travel involved. And I would have gotten a few hundred sticky notes rather than an entire notebook with 113 separate detailed proposals.” The authors go on to say, “And, most important of all, there was a result. Basically, in 60 days, Roche was able to solve a problem that it and its partner have been tinkering with and optimizing for the last 15 years.”⁴

Reinforcing the non-cash utility arguments put forth in Chapter 2, “The Future of Value Creation,” Bedilion observes, “Clearly, the financial incentive played a part here, but we think there is more going on—people also seem to get intrinsic value out of sharing their expertise through this community.” Bedilion’s observation mirrors that made years earlier by Lakhani as he examined the motivations of InnoCentive solvers. In the same Lakhani paper cited earlier, it is noted that the likelihood of making a winning submission correlates more strongly with intrinsic motivations, such as the satisfaction of problem-solving or intellectual curiosity, than it does with the cash award (though not statistically different).⁵ This is, of course, not an argument to abandon cash bounties, but a recognition that multiple utilities are in play and important to the process. Many authors have commented on the “wisdom of crowds,” putatively starting with James Surowiecki’s book by the same name.⁶ Enormous efforts are made to define and measure collective intelligence, including multidisciplinary undertakings such as the MIT Center for Collective Intelligence, under the direction of Professor Thomas Malone. Following are a few of the authors’ own observations about where the problem-solving value of Zone C, taken as a whole, originates. Out of respect for the scholarly work in the categorizing and measuring of crowd intelligence, you are referred to elsewhere for in-depth scientific arguments. But we do not shy away from addressing the question of just where the solving power comes from in a crowd. It would be too simple to just shrug and imagine that broadcasting challenges to the crowd precipitates out the occasional expert or near-expert: the needle-in-a-haystack argument. And that happens; but that is about finding needles, a search problem, and not the power of the crowd as a whole, as a body. Legitimately, it would seem that the only explanation needed is the “long-ness” of the tail and the summing of many small probabilities versus the summing of a few larger ones. But our own observations of problems, solved in a broadcast search or crowd-sourced manner, prompt us to propose three characteristics of the crowd—three origins of its problem-solving capability: diversity, marginality, and serendipity.

Diversity, Marginality, and Serendipity

The first of these is diversity. Having a larger number of people tackle a problem would obviously be of limited use if they all used the same approach. This is more than just a numbers game as in “many hands make light work.” A mental picture of the diversity argument is to imagine “space.” Although this actual space is “problem-solving space,” with the axes defined by the parameters of the approach, imagining that space familiar to all of us, as a great celestial vacuum, sparsely populated by stars, will work perfectly well for this metaphor. One characteristic of rich, complex problems is that there are so many ways NOT to solve them.

Think of the few scattered stars in a volume of space as being solutions to your problem and the cold dark areas representing failure to solve the problem. Problem solving routinely begins with a hypothesis or a belief of what a solution might be. If that first attempt fails, the results of the effort are gathered and analyzed, and a new attempt is made. Think of that problem-solving process as an explorer. The testing of an hypothesis is the equivalent of that explorer stepping into space where he thinks there is a star, a solution to his problem. His analyses, calculations, and logic have led him to guess where a star might be, and he steps into space to find his surroundings cold and inhospitable with only a glimmer of light in the distance. He takes stock of the situation and jumps again, pausing to assess his new location. He does this repeatedly, each time his evaluations telling him if he is “getting warmer.” And thus he navigates his way from his original guess to the solution-star.

Now, imagine instead that you have not one explorer but a thousand explorers. Each of these explorers begins with a different perspective on where a star might be found. They each begin with different instruments to do the calculations. They each begin with different experiences and intuitions about finding stars. And so they each jump into a different region of space. These are not random jumps. If they were, then, in all likelihood, every one of our thousand explorers would find themselves cold and dark. They are informed jumps (though to varying degrees) and as such it is not unimaginable that some would find themselves in regions relatively warm and bright with only a short distance remaining to a solution.

Marginality is the second source of problem-solving capability in the crowd. Although you have become acclimated to the notion that deference to experts is a wise choice of action, you need to recognize that the acquisition of expertise is inevitably accompanied by the acquisition of constraints. Experts are taught not only to solve problems but also how to solve them. They are taught that there are right approaches and wrong approaches. They are taught what has been tried and failed, and should not be tried again. These constraints may hinder their ability to see problems in a fresh light and approach them without bias. We are not suggesting that experts are miseducated or maltrained; recall from the arguments about the long tail that the experts resided at the highest probability, the “head” of the graph, and that by virtue of their education and training. And if, contrary to the counsel of Runyon, the race was always to the swift and the strong, you would need to proceed no further. But because you know this not to be the case, you need to accept that constraints go hand-in-hand with expertise. This has led to attempts to measure those biases and deal with them explicitly. In a publication draft that Karim Lakhani shared with us, he wrote: “Because individuals become socialized to the norms and beliefs of their fields and organizations, remaining at the margins while keeping up to date and actively pursuing access to resources offers those marginalized a different set of perspectives and heuristics than those at the core of the professional establishment.”⁷ Such observations have been published, in connection with broadcast search methodology, in a Lakhani and Jeppeson article in Organizational Science,⁸ and in the arena of psychology and sociology, in Neil McLaughlin’s article in Sociological Quarterly, appropriately titled “Optimal Marginality.”⁹

Our third and final argument for the source of problem-solving capability within a crowd is serendipity. The role of serendipity is not an unfamiliar one, and it is discussed at the conclusion this chapter with a retelling of the story of Archimedes, in an attempt to extract new learning’s from an old tale. Serendipity seems for the most part to be one of those factors completely outside the control of either the institution or the individual attempting to invent or innovate. And yet, its importance to the advancement of knowledge and technology is evident in the lore—whether you speak of Newton getting beaned by an Apple, Kekule dreaming of snakes, Maxwell’s envisioning of gear trains, or Archimede’s “Eureka!” bath.

These historical observations, along with your own personal experiences in solving complex problems should lead you to conclude that novel solutions often arise due to the confluence of three distinct factors: skill or training, a clearly articulated challenge or need, and a personal experience (not domain experience), either past or present (usually viewed as the source of the serendipity).

The Tear Gas Connection

Our experience at InnoCentive, where we have watched the crowd solve problems hundreds of times over, leads us to repeated observations about how, and how frequently, these three factors of diversity, marginality, and serendipity, play a role in bringing forward novel problem-solving approaches. Although we could illustrate this with a great number of examples, let’s choose just one to show how all three forces were at work.

An InnoCentive client had an application for a tetracarboxylic acid, a novel molecule, that was not readily produced at a competitive cost. The challenge to design an efficient means of preparing this material was posted on the InnoCentive website and attracted potential solvers from around the world. During the 4 months that the challenge was active, 247 people from 35 countries, representing 247 jumps into problem-solving space (diversity)—opened project rooms to see the details of the challenge and sign intellectual property transfer agreements. At the conclusion of the 125-day posting period, 17 individuals had submitted novel proposed solutions. In a blinded review of these submissions, by the seeker, the most promising approach was submitted by a patent attorney. To be fair, the attorney, David Bradin, was no stranger to chemistry and had worked as a chemist before earning his law degree: marginality.

You might ask, “Where’s the serendipity?” After being announced as the winning solver, Bradin explained that when he first saw the problem his thoughts turned immediately to tear gas and some chemical strategies used in its preparation. In an obtuse series of mental links, Bradin applied his tear gas experiences to seeing the problem in a new way.

Of course, that’s not to say that “tear gas” was the only potential provocation or experience that would enable a fresh perspective. But in as much as it is virtually impossible to apply serendipitous connections or events to rational problem solving, it seems that the only logical resort is to a crowd—opening up a greater likelihood of such connections occurring. Bradin says that he continues to scan the challenges posted as part of the InnoCentive website, “for those where I might have a flash of insight.” This is clearly a matching mechanism that is not operating when challenging research problems are assigned within an institution.

Eureka! The Right Question at the Right Time

As briefly mentioned, in connection with serendipity, the role of the challenge is not to be ignored in capitalizing on the intelligent behavior of the system. This can be illustrated by recalling the oft untold portions of the Archimedes story and the discovery of buoyancy or water displacement. Hiero, the tyrant of Syracuse, had ordered an artisan to construct a crown for him. The tyrant had paid for the crown and supplied all the gold that was to be used. Naturally Hiero was smart enough to have the crown weighed upon its receipt to ensure that all the gold was accounted for. But amalgams and alloys were known to the people of Hiero’s time. Gold was strengthened by the addition of base metals without significantly altering its appearance. How was Hiero to know that the gold he delivered had not had a portion removed and the remainder mixed with lead to produce the expected weight. Well, one feature of the alloy, gold plus lead, was that it would have a different density; that is, the weight of a given volume of the alleged gold mixture would not be the same as that of pure gold. Hiero could verify the honesty of the artisan by measuring the volume of the crown.

For this, he took it a local citizen skilled in the art of mathematics. Archimedes examined the crown, acknowledged the beautiful workmanship, and then delivered the bad news: The crown was an irregularly shaped object, and although Archimedes could easily have delivered to Hiero the accurate volume of a cube or a sphere, or even a cylinder, he could not do so for an irregularly shaped crown. And, of course, melting the crown back into a cube would’ve merely started the process all over again (although possibly with a new, live artisan).

That night, as the tale is usually told, Archimedes retired to his home and took a bath. As he lowered himself into the tub, he noticed the water rising along the sides of his “regularly shaped” bath receptacle. As Archimedes watched, he realized this was a method to effectively measure the volume of an irregularly shaped object, whether that irregularly shaped object was Archimedes’ middle-aged body or Hiero’s ornate crown. As the story goes, Archimedes was so excited that he leapt from the tub, never bothering to dress, and ran naked through the streets of Syracuse shouting, “Eureka! Eureka. I have found it. I have found it.”

This story is related to high school science classes as an illustration of creativity, the “a-ha! moment.” Students are encouraged to remain inquisitive about their world, to realize that clever solutions to problems may very well lie outside the present day strictures of what is known, and that even complex mathematical problems may have a simple nonmathematical solution. Less attention is paid to Hiero’s role in this process other than as a setup for the important event that was to follow. (Chapter 6, “The Challenge Driven Enterprise,” discusses the topic of the challenge’s role as well.)

But let’s here illustrate the criticality of Hiero’s question, and the manner in which it was posed, by speculating (probably rightly) on something not recorded in either history nor this oft-told anecdote. This wasn’t Archimedes first bath!

How was it that this math- and science-oriented, intelligent person had failed to discover a breakthrough—that sent him screaming naked through the streets—on any of his many prior baths? We are prone to think of questions and challenges as necessary precedents to discovery, without delving much more deeply into how they are constructed, what properties make them powerful, or even where they came from in the first place. Everyone is familiar with clever courtroom dramas in which the nuanced wording of the question either elicits an incriminating response, or in which the question itself contains a conclusion that the attorney wants the jury to hear. Advances in science and technology are rarely considered in light of that courtroom drama. They are not considered exercises in wordplay. You can imagine that they live in a realm somewhere above language—independent, a consequence of laws that exist free of what they are named or even how they are described.

But scientific and technological advance do not occur at a constant rate and on all fronts simultaneously. They are sparked by something. The frontiers are moving in fits and starts according to human conditions, human desires, and human language. The salient difference between Archimedes’ 10,001st bath and his 10,000th was Hiero’s question, and, no doubt, Hiero’s enormous control over the consequences of the answer. It is a given that, almost certainly, this principle would have been discovered either at another time or in another place.

Hiero’s query acted as what Professor Paul Carlile, at Boston University’s School School of Management, calls a boundary object, which is a device, either conceptual or physical, that enables a boundary of some type to be crossed.¹⁰ These could be boundaries of academic disciplines, boundaries of terrain—that is, water and land—or boundaries of time. The circumstances of Hiero’s question produced a shift in time and space so that an event occurred at this particular point in time and space rather than at some other. Challenges, or well-formulated questions-asked, are the boundary objects that enable human progress to occur as it does, in fits and starts, and the effectiveness of that progress is often a function of the effectiveness of the challenge itself. As frequently remarked, “A problem well-stated is a problem half solved.”

The tale of Archimedes is told here not only for its obvious messages regarding creativity and innovation, but also to suggest that it could be practiced by innovators in a much more plentiful, more accessible, manner. It is common experience that serendipity is shrugged off as one of those things that just sometimes happens when your head is down and you’re working hard. It’s good, but it’s not schedulable; it can’t be invoked at will—but is also not an entirely rare occurrence. To access it, and to do so consciously, means a crowd is indispensable. When looking for a unique solution, the key, with foresight, would be not just to ask an Archimedes, but to ask him as he lowers himself into his bath. Unfortunately, we don’t have foresight. To mimic that result, you need to access many potential sources of knowledge, each engaging in a unique train of thought. You need to find Bradin as he thinks of tear gas, and all you know for sure is that he occupies a crowd, a crowd that must become part of your overall problem-solving strategy—a crowd accessible only if your organization taps Zone C, the long tail. Goodbye to the logic of scarcity; welcome to the logic of plenty.

As Anderson himself explains in his October 2004 Wired article, “What’s really amazing about the Long Tail is the sheer size of it. Combine enough nonhits on the Long Tail and you’ve got a market bigger than the hits. Take books: The average Barnes & Noble carries 130,000 titles. Yet more than half of Amazon’s book sales come from outside its top 130,000 titles. Consider the implication: If the Amazon statistics are any guide, the market for books that are not even sold in the average bookstore is larger than the market for those that are.”¹¹

The history of innovation is marked by many events in which the uniquely prepared mind capitalized on a singular set of circumstances. In the last century, centers have been built for innovation and technical excellence by application of the rules of rational compromise and the logic of scarcity. Admittedly, you still can’t program serendipity into the process, per se, but you can employ the crowd and the logic of abundance to create a greater likelihood of serendipity and other problem-solving opportunities to be manifest. Rephrasing Anderson’s comment on “non-hits,” and speaking of any specific well-stated challenge, you might say, “The number of novel ideas available from those who aren’t ‘experts’ is greater than the number of novel ideas available from those who are.” This open approach to problem solving can open the door to accelerated innovation and to solutions to globally critical problems of nourishment, environment, health, and energy. Can you really afford to leave any zone in your map of human capability untapped?

Case Study: How the Oil Spill Recovery Institute Tapped the Crowd to Be Better Prepared for Arctic Spills¹²

On March 24, 1989, the massive oil tanker, Exxon Valdez, hit Bligh Reef and ruptured, spilling 11 million gallons of crude oil into Alaska’s Prince William Sound. The Exxon Valdez disaster affected 1,300 miles of coastline and threatened many species of wildlife from fish to mammals to birds and even insects and bacteria.

The Exxon Valdez case has been studied by numerous researchers and is taught in colleges and universities around the world. The facts of this event live on in secondary school textbooks on science and the environment. While Prince William Sound recovers, and even as oil remains on some of the beaches, 20 years later, it is vitally important that plans are made NOW to obviate, or at least mediate, any future spills and the environmental consequences.

The Oil Pollution Act, passed by Congress in 1990 in response to the Valdez disaster, created the Oil Spill Recovery Institute, headquartered in the Prince William Sound Science Center in Cordova, Alaska. Subsequent legislation has insured that the Institute and its mandate will persist as long as drilling efforts continue in Alaska.

As they have for two decades, OSRI continues to play out scenarios and evaluate potential responses. At a conference several years ago, it brainstormed some of the unique hurdles it may face in a future arctic spill and agreed to collect, study, and document the potential responses. One clear complication of an arctic spill is the frigid temperature. Among the tools available during cleanup is the skimmer vessel known as a minibarge. These hold about 250 barrels (a little over 10,000 gallons) of crude oil, mixed with seawater, as the surface water is skimmed and deposited into the tanks. When the tank on a minibarge is full, the contents must be moved to a larger vessel or to holding tanks along the shoreline.

If it’s –10 degrees Fahrenheit (and it often is), the oil is going to thicken beyond the point of transferability. The pumps used to transfer the oil are powerful submersion pumps that are designed to be dropped directly into the minibarge holding tanks. Although capable of transferring something as viscous as peanut butter, it is imagined that in a frozen minibarge tank, the running pump would simply “chew” its way to the bottom, leaving a slowly collapsing hole and emptying the tank only over a prohibitively long time. If a method could be found for expeditiously emptying the barges in the coldest weather, it would speed cleanup and provide reserve barge capacity year-round as a safeguard.

OSRI is a grant-making body, funding research related to oil spill recovery at high latitudes. Lead by research program director Scott Pegau, it identifies and funds research to address gaps in understanding and capabilities. As Pegau says, “This specific problem (of cold, viscous oil transfer) has been around for years—as have many others.” Over time, Pegau and his team have consulted with petroleum engineers from around the world. Wanting to explore beyond the real-world constraints he had to work with, Pegau turned to a problem-solving network, InnoCentive. The challenge for “reducing viscous shear” was posted on InnoCentive and broadcast to solvers in 120 nations. At the end of the posting period, there were 27 submitted responses to the challenge. Of these, four different ideas were deemed worthy of further consideration, and one was awarded the bounty OSRI had posted. Pegau recognized that many promising leads were coming from those who had never worked with oil but who recognized the similarities in behavior to other materials with which they did have experience.

The winning approach came from an industrial hygienist who spent his typical day measuring workplace exposure to hazardous materials. John Davis, the awarded solver, has a master’s degree in chemistry from Illinois State University. Under the direction of Professor Cheryl Stevenson, his thesis project was “Anion Radicals of Cyclooctatetraene.” There was virtually no connection between his professional specialties and the OSRI needs. But Davis sees himself, first and foremost, as a problem solver, and he could readily grasp the nature of OSRI’s problem. In past efforts, Davis was helping a friend and found himself pouring concrete. It was while on the construction site that Davis learned how to manage the movement of highly viscous fluids, such as setting concrete. As a scientist, it was readily apparent to him that the techniques he’d employed pouring cement could be an answer to moving the oil off those frozen barges.

Davis’ solution involved the use of sonic vibrations to “mobilize” the fluid and cause it to behave as if it were a much less-viscous substance. Using methods proven elsewhere, he postulated that the sonically agitated oil would “vibrate” its way into the pumps under the available pressures and allow itself to be pumped away from the barges so that they could return to their duty of skimming the spill. So, what Davis proposed, in a nutshell, was the use of sounds to save our sounds. Luckily, there have been no new spills in Alaska, but if and when there are, this new technique is now “in the quiver” to respond rapidly and minimize their impact.

Even before Scott Pegau’s arrival at OSRI, the Institute had considered using public challenges and awards to advance its research agenda and solve some of the unique problems it faced. But no clear mechanism had been identified until the InnoCentive platform was launched and provided a low-cost access to promotion, challenge structuring, and a widely accessed web presence. Working with InnoCentive, OSRI had placed a “bounty” on solving this problem of $20,000.

Since their success with this first challenge, OSRI has posted five additional challenges on InnoCentive. Collectively, the challenges produced 215 submitted ideas, and OSRI has made awards on three of the five. It is a source of hope that OSRI and clever solvers are openly collaborating to ensure that future adverse environmental impacts are minimized.