Be Quick, Be Useable, Be On Time: Lessons in Agile Delivery of Defense Analytic Tools

Executive Summary

When faced with urgent national security needs, the Department of Defense (DoD) has historically bypassed conventional processes and allowed a greater degree of agility to flourish for critical programs. Over the past decade of combat operations, this again has held true. For all the criticism of defense acquisition, there have been examples of extremely successful efforts that quickly delivered new capabilities responsive to urgent warfighter needs.

These efforts were characterized by an intense focus on the user, a commitment to rapid delivery, and an acceptance of an incremental approach to improving capabilities. Instead of seeking the ideal solution, programs adopted a continuous cycle of deliver, learn, adapt, and improve.

As our active war fighting commitments draw down, and budgets tighten, we risk losing important lessons. The urgency behind these agile efforts will decline at the very time when we may need them most. The importance of agile practices for the Department is growing. These practices need to expand beyond urgent combat needs to be the norm for more mission capabilities, including analytic tools.

Four key factors are driving this need for increased agility. They include: the rate and unpredictability of advances in technology; the difficulty specifying requirements early in a program; the wide range of future potential conflicts and adversaries; and the rapid expansion of data sources and data volumes. Taken in concert, these factors call for alternative approaches for how systems, specifically those heavily dependent on information technology, are developed and acquired.

There are significant differences between agile and conventional processes that pose challenges to more widespread adoption of agile methods. Historical cases, contemporary experience, and commercial lessons can inform DoD efforts to scale up existing agile pockets to institutional processes. To be successful, efforts need to be coordinated across functional and process boundaries and cannot be left to the acquisition community alone, or instituted in isolation.

This analysis identified three critical pre-conditions that need greater attention in the ongoing discussions on transitioning to more agile methods in DoD. First, recognize the central role of user collaboration throughout development. Second, create mechanisms to easily bring together multifunctional teams to fuel the cycle of deliver, learn, adapt, and improve. Finally, foster a culture characterized by agility as a routine, vice requiring senior leader intervention to break down barriers, through consistency in what is stated as important and valued, such as agility and adaptability, with what is measured and incentivized.

Fortunately, there is contemporary experience within defense organizations with agile methods. There are successful practitioners with years of experience that can seed expansion of these practices. Areas such as intelligence analysis and cyber require this agility now. Continuing to view these practices as exceptions limited to urgent national security needs is ill advised.