Under former Secretary of Defense Ashton Carter and the third offset strategy, the Department of Defense launched a love affair with Silicon Valley. The relationship was a logical fit for Carter, a former physicist and professor at Stanford University. From the original Silicon-Valley based Defense Innovation Unit Experimental (DIUx) to the Defense Digital Service and the tech-heavy Defense Innovation Advisory Board, Carter’s innovation initiatives under the third offset drew unabashedly from Silicon Valley. These organizations have adopted organizational, technological, and cultural practices from the Northern California tech hub. Employees are “bureaucracy hackers,” “technology exorcists,” and “rogue leaders,” while outreach speaks of “ecosystems” and “partners.” Offices are decked out with contemporary furniture, white boards, and Apple products. The staff, while often active and reserve military, typically wear civilian clothes — and not the stuffy suits of the D.C. beltway. A public affairs report by the Department of Defense touted the director of Defense Digital Service by proudly announcing that he was “the only director of a Defense Department organization who shows up to work wearing a hoodie.”
The belief that undergirds these initiatives, and these very un-Department of Defense cultural practices, is that appropriating lessons from Silicon Valley leads to greater defense innovation. This seems like a compelling argument. Silicon Valley is the world’s leader in innovation, especially in technological innovation. In 2015, Silicon Valley based businesses generated $24.5 billion dollars in venture capital investment. That means that Silicon Valley’s venture capital investment is greater than the GDPs of approximately half of the world. Further, with over 19,000 patent filings the region is one of the top three internationally. It does all this by attracting extraordinary talent, boasting one of the most educated workforces on the globe.
However, as with all analogies, there are important differences between defense innovation and Silicon Valley that make the use of the model inappropriate — or at least incomplete — for defense innovation. As Secretary of Defense James Mattis evaluates the follow-on to third offset, it is a useful time to look closer at whether there may be better commercial analogies for defense innovation. Perhaps, after a two-year love affair, it is time to swipe left on Silicon Valley.
Swiping Left: Problems with the Silicon Valley Analogy
Despite the overwhelming success of Silicon Valley, there are significant downsides to its innovation model. Failure is rife — an estimated 90 percent of start-ups fail and only 1 percent become the billion-dollar companies we typically think of as examples of Silicon Valley success. Further, critics worry the innovation model is difficult to replicate, especially by public enterprises and in organizations and cultures that are generally risk averse. Finally, the splintering of production chains makes Silicon Valley’s lean staffs and global outsourcing incompatible with organizations that focus on post-start-up production innovation.
These critiques highlight large potential incongruities between defense innovation and Silicon Valley. Defense as an organization tends to be risk averse and not primed for failure. Further, what Silicon Valley is best at — incubation innovation — is not optimized for defense models. At least for American defense organizations, the greatest challenge in technological development has not been in generating good ideas (think for example about the amazing success of organizations like DARPA), but in taking these good ideas from incubation to operational integration. Defense acquisition wonks call this transition between technology prototyping and full operating capability the valley of death. Unfortunately, Silicon Valley tells us little about how innovation can solve this valley of death problem in large, cumbersome organizations like the Department of Defense that are bridled with Congressionally-mandated inefficiencies and deeply-entrenched bureaucracies.
But perhaps the largest difference between Silicon Valley and the Department of Defense is that Silicon Valley’s most successful businesses are technology producers. This is probably a fundamental divergence from defense, for which the primary objective is not the production of technology but rather the achievement of national security objectives, (defending the homeland, protecting allies, projecting power, defeating adversaries in conflict). Defense contracts for technology and often incubates technologies to attain these objectives, but ultimately defense is a technology-enabled organization and not a primary producer of technology. Therefore, models of innovation that focus solely on how to innovate new technologies fail to help solve innovation problems caused by challenges acquiring or implementing technology within missions.
Defense — particularly in the United States — often focuses on technology as a stand-in for military success. However, a military with all the best technology still may not be able to translate that technology to victory. Think, for example, about the vast array of technology created for Iraq and Afghanistan that was unsuccessful without innovations in training, doctrine, operations, and strategy. The literature on military innovation is clear on this point: technology only goes so far towards military effectiveness. Instead, it is the militaries that combine the development of new technologies with organizational or doctrinal change that ultimately succeed in battle and war. Like the ill-fated Silicon-Valley based Juicero, too much focus on technology without non-material innovation efforts or without thinking about larger strategic objectives leads to the misallocation of resources and a nation that is ill-equipped for future competition.
Swiping Right: Alternative Commercial Analogies for Defense Innovation
Being a tech-enabled organization instead of a primary tech producer has significant implications for models for innovation. Tech-enabled models need to focus less on how to invent new technologies and instead need to account for relationships with technology vendors, for balancing cyber-vulnerabilities with technological development, and for thinking about what to produce in-house versus what to outsource to primary tech-producing companies. Two tech-enabled industries, banking and healthcare, provide particularly interesting analogies for defense innovation. Each requires greater understanding and each have their own limitations, but both present new insights for the Department of Defense
Banking is a particularly interesting innovation model for defense. Like defense, banking relies on technology to achieve a non-technological objective — managing and making money. When we think about companies like J.P. Morgan Chase, Citi Group, or Wells Fargo, we don’t think of them as tech companies. However, even though banking doesn’t primarily produce technology, these companies do rely on technology for financial success. Modern-day banking is increasingly a technological problem in which the safeguarding of virtual information, the transmission of currencies through cyberspace, and day to day financial transactions are dependent on technology. Like today’s defense organizations, banking is both incredibly effective because of its digital capabilities and also incredibly vulnerable to attacks in the cyberspace domain. Technology, therefore, is a double-edged sword for the financial sector.
This double-edged sword leads banking to invest heavily in information technologies and talents. They employ thousands of computer scientists and software developers to produce technology solutions in-house while also partnering with start-ups and established tech companies for Fintech, data analytics, and security technologies they don’t have the comparative advantage to produce. J.P. Morgan Chase employs 40,000 technologists and spends $9 billion on technology and $2 billion on security. (To put this in perspective with the U.S. Department of Defense, the Department of Defense’s primary IT agencies, Defense Information Systems Agency and Cyber Command, employ only 21,000 personnel.) According to J.P. Morgan CFO Marianne Lake, this is why Chase should be characterized not as just a bank but also “a technology company.” Banking, therefore, is an excellent case to understand how large organizations outsource technology and interface with vendors, how they decide what technology to produce and maintain in-house, how to manage and recruit technological talent, and how to invest in cybersecurity.
Like banking, healthcare is also a technology-enabled industry. From the increased use of robotics in advanced surgery to the digitization of patient records, and the use of databases to manage large healthcare organizations, the healthcare industry increasingly relies on technology to keep people healthy. This makes healthcare — similarly to banking and defense — particularly vulnerable to cyber-attacks on the technology upon which it depends. Cyber-attacks on healthcare are some of the most prolific within critical infrastructure and these attacks have surged as medical companies increasingly digitize records. How the industry responds to these vulnerabilities can help inform the Department of Defense, which is also vulnerable to attacks on increasingly digitized and autonomous technologies and operations.
Perhaps where healthcare looks more like defense and less like banking is in its relationship with risk and regulation. Like defense, healthcare also has high stakes for failure. Whereas Silicon Valley values and even rewards failure as part of its innovation model, healthcare has a much more risk-averse and heavily regulated approach towards risk and failure. Oversight is as much a part of healthcare innovation as acquisition is within the Department of Defense. Indeed, the high ethical stakes of innovation within healthcare create a lot of the same normative barriers to innovation as those that impact the proliferation of emerging technologies in defense, (for example, artificial intelligence and autonomy). This has created significant challenges for innovation within the healthcare industry; defense can learn a lot about how the healthcare industry responds (both successfully and poorly) to these cultural and organizational barriers to innovation.
Where Does This Leave the Third Offset?
Swiping left on Silicon Valley innovation models doesn’t mean there isn’t a place for Silicon Valley in defense innovation discussions. The strength of organizations like DIUx is not necessarily in their appropriation of Valley culture, but instead in their ability to bring non-traditional firms with emerging technology into the Department of Defense. DIUx’s power to do this is because of how they have innovated acquisition models, not because the organization itself can better innovate new technologies for the future battlefield. Tech-enabled sectors like banking and healthcare lead concerted efforts to court and invest in Silicon Valley vendors. The Department of Defense needs better ways to do this and DIUx provides a first gateway to streamlining acquisition processes and using more commercial models of emerging technology. (This is particularly important for quickly changing technology like software.)
Additionally, concerns that the military “culture” is impeding innovation are likely valid. The Department of Defense acquisition process — partly because of its congressional oversight and partly because of its own perverse incentives — has become extraordinarily risk averse. And to make matters worse, service cultures tend to bias toward capabilities and operational constructs that benefit their elite. Therefore, for services and operational specialties that base their identity on particular weapons platforms or mission sets, innovation can be especially painful. The third offset strategy ignored many of these cultural disincentives for innovation and instead focused on technological solutions; the next strategy should take a harder look at how cultural barriers to innovation can be overcome.
Finally, this short analysis of innovation reveals the very large role software acquisition and development plays for any major tech-enabled organization in the information age. The Department of Defense has a horrible track record for developing, maintaining, acquiring, or securing software. Its contracting process was not designed for the fast pace of software development, its traditional vendors are not on the cutting edge of software development, and its organizations are not optimized for maintaining and keeping up with software changes. Therefore, the software products it develops or invests in are rife with user-interoperability issues, technology lags, and security vulnerabilities. Much of the technology the Department of Defense needs to invest in is in Silicon Valley, but the innovation processes and organizations required to capitalize on that technology increasingly reside instead in tech-enabled industries.
As we re-evaluate the third offset, looking at other industries for innovation analogies can help us design a strategy that is about more than developing technology and instead seeks to innovate the way we acquire and institutionalize technology. We need to think more deeply about if and when we need to create technologies in-house (floppy disks for the nuclear command and control?), how we attract and staff a workforce of technologists, how we interface with cutting edge technology developers, and how we institutionalize and yet de-centralize the safeguarding of our digital capabilities.
Dr. Jacquelyn Schneider is an Assistant Professor at the U.S. Naval War College. Her work on national security, technology, and political psychology has appeared in Journal of Conflict Resolution, Strategic Studies Quarterly, Washington Post, Bulletin of Atomic Scientists, and War on the Rocks. The views presented here are hers alone and do not represent those of the Naval War College, the U.S. Navy, or the Department of Defense.
Image: U.S. Army photo by Michael Beaton