that this project could be constructed with Disney as the employer in alliance with Siemens. Siemens got the benefits of Disney's world-class employees, reward structure, and culture with decades of experience marketing to children, without having to create a similar structure internally. The move was enabled by an existing alliance between Disney and Siemens for building theme park rides; the two had learned the trick of working together, and that set the stage for an unforeseen collaboration on hearing aids.
Does the work of your organization require the talent to reside inside your organization, or do you, like Siemens, simply need a way to access the right talent in another company? If the work you need to do is outside your core value proposition (like marketing to children was to Siemens), might there not be better talent you could borrow from outside? Do you structure your alliances based on optimizing the work, or based simply on financial or technical elements?
How does one develop a comprehensive portfolio of noninsulin diabetes drugs? You might think that giant pharmaceutical firms could take that on, but even for them it is a daunting challenge to perform at world-class levels on all the many elements of drug development. AstraZeneca and Bristol-Myers Squibb are robust competitors, but their leaders led through the work, realizing that the best way to fight diabetes was to do it together. In 2007, they formed a global diabetes alliance to discover, develop, and commercialize new drugs for type 2 diabetes. Add in Bristol-Myers acquisition of Amylin Pharmaceuticals in 2012 and the alliance had the capability to offer a full spectrum of treatment options.6
This is a good example of borrowing and buying capability rather than building it internally. In their book Build, Borrow, or Buy: Solving the Growth Dilemma, Laurence Capron and Will Mitchell argue that knowing when to build, when to borrow, and when to buy capability is critical to success.7 The trouble is most leaders lean too heavily on one tactic instead of applying the appropriate solution to the situation.
AstraZeneca ended up buying the alliance in 2014, essentially incorporating employees who were formerly outside its boundary and bringing them inside. Does that mean the alliance was a mistake? No. It is instead an example of another way to lead through the work: Envision your organization as flexible, constantly changing its shape, rather than as a rigid structure. In 2007 AstraZeneca extended its organizational boundary to overlap with Bristol-Myers Squibb in diabetes research; in 2012 Bristol-Myers engulfed Amylin. In 2014 as Bristol-Myers Squibb began moving in a different direction, it made sense for AstraZeneca to fully absorb the alliance into the main corporate body.
If AstraZeneca thought in terms of fixed structures and rigid organizational boundaries, they would never have achieved their current strength in diabetes treatments. They saw the work of winning the diabetes game as being about moving pieces available somewhere in the world, not just moving the pieces available within the organization.
We are all familiar with outsourcing and the economic value it provides through specialization and its ability to mitigate the impact of product demand fluctuations. Alliances have similar advantages, but they introduce a much fuzzier set of relationships. The alliance between AstraZeneca and Bristol-Myers Squibb on diabetes treatment didn't just share employees, it also shared intellectual property. That fuzziness is important. It is both a challenge and an opportunity. When it comes to leading through the work, the traditional boxes we use to define what is inside and outside an organization are breaking down.
Earlier we showed how an organization called Topcoder was a source of freelance computer coders to solve Ion Torrent's compression problem. It illustrates how work is escaping the confines of regular full-time employment. Yet, Topcoder is much more than a source of free agents. It is an example of something called a talent platform that not only provides an alternative source of workers but offers insights about what it fundamentally means to lead through the work. We will deal with talent platforms in depth in Chapter 4. Here, we offer some highlights to show just how fundamentally they change how you think about leading through the work.
Upwork, the leading site for freelance work, was designed to be a marketplace that matches work to free agents. Need a logo? You can find a designer on Upwork. Need a part-time administrative assistant? Upwork can help you find one. Need a brand strategist? The talent you need, for as long as you need it, is a few clicks away. In many ways, Upwork is an Internet-based replacement for a temp agency – at least that is what it was when it started.
Think of it like the consumer buy-and-sell sites Craigslist and Kijiji, but instead of buyers and sellers of used household goods finding each other, work and talent find one another. A leader lists a task that needs to be done and free agents offer their services. Alternatively the leader can search the listings of free agents to see who is available. It is similar to job boards like Monster or CareerBuilder, except regular full-time employment isn't being offered or sought; and it offers services to help overcome barriers that get in the way of working with off-site free agents.
Upwork successfully competes against temp agencies partly because of the efficiencies of being automated, partly because it is useful even if you just have a small task rather than a whole job, and partly because it can tap affordable talent in the developing world. Upwork is important if you are a temp agency competing for market share or a leader looking for some extra help. If a talent platform was just the equivalent of a big room filled with tasks and free agents wandering around to find each other, then it would not be particularly exciting. And if Upwork was the only talent platform out there, it would be interesting, but hardly world-changing. However there is much more to talent platforms than this simple view.
Consider the talent platform Ion Torrent used: Topcoder. Whereas Upwork is usually seen as a way of getting work done more cheaply than using employees, Topcoder intends to tackle programming tasks so difficult that your employees cannot do them.
Topcoder challenges employment on two fronts. As a leader, when does it make sense to get work done with a fixed group of employees (assuming you have an employment brand to attract this highly desirable pool of talent, and they would pick you over Google) versus giving the work to more talented programmers on an as-needed basis? As a talented programmer, when does it make sense to tether yourself to a corporation when you could fly free as a Topcoder? The bigger question has to do with the scale of the change. Are we headed toward a world where most programming work is done via talent platforms?
What Topcoder is to programming, Tongal is to advertising. Tongal strives to be a better way for firms to get advertising videos made. It's a talent platform that enables crowdsourcing of ideas and the production of commercials. It attracts work from top brands like Lego, Anheuser-Busch, and Procter and Gamble. In the old, big-budget world of mass-market TV advertising, traditional advertising agencies may have an advantage, but among the fragmented audiences of the Internet and cable TV, those big budgets are unsustainable. For commercials, talent platforms like Tongal are a big part of the future.
A quite different kind of talent platform is Amazon's Mechanical Turk. Amazon's platform is named for the Mechanical Turk, one of the most notorious machines in the history of artificial intelligence. The Turk was an eighteenth-century chess-playing robot that astounded the intelligentsia of the time. No, your sense of the history of technology is not awry; the Mechanical Turk was a clever fraud. A man was hidden inside the robot and it was he who provided it with the intelligence to play chess.
Even in the modern world of computing there are some things humans do better than machines. Amazon's Mechanical Turk (MTurk) feels like a machine, but it cleverly takes little tasks and farms them out to anonymous human workers hidden behind the interface. Consider image recognition, such as being asked “Is this a picture of a kitchen or a bathroom?” This sort of task is easy for a human but hard for a machine. When leaders at Amazon confronted the problem of handling large numbers of microtasks a computer could not do,
Jeanne Whelan, Jessica Hodgson, “AstraZeneca, Bristol-Myers Deepen Diabetes Alliance,” Wall Street Journal, published January 31, 2013, www.wsj.com/articles/SB10001424127887323701904578275290772944154 (accessed April 6, 2015).
“AstraZeneca and Bristol-Myers Squibb Diabetes Alliance Provides $5 Million Grant for American Diabetes Association's Pathway to Stop Diabetes Research Initiative,” Bristol-Myers Squibb, published January 16, 2014, http://news.bms.com/press-release/astrazeneca-and-bristol-myers-squibb-diabetes-alliance-provides-5-million-grant-americ (accessed April 7, 2015).
Jennifer Fron Mauer, Laura Hortas, Timothy Power, Sarah Lindgreen, James Ward-Lilley, and Karl Hård, FierceBiotech blog, posted January 16, 2014, http://www.fiercebiotech.com/press-releases/bristol-myers-squibb-and-astrazeneca-complete-expansion-diabetes-alliance-t (accessed April 7, 2015).
AstraZeneca press release, posted February 3, 2014, www.astrazeneca.com/Media/Press-releases/Article/20140203-astrazeneca-acquires-bms-share-of-diabetes-alliance (accessed April 9, 2015).
Laurence Capron and Will Mitchell,
