A paratrooper assigned to U.S. Special Operations Command Europe descends to a drop zone near Alzey, Germany during a static-line training jump, April 14, 2023.

A paratrooper assigned to U.S. Special Operations Command Europe descends to a drop zone near Alzey, Germany during a static-line training jump, April 14, 2023. U.S. Army / Sgt. Laura Bauer

DOD Managers Need a New Approach. SOCOM Can Lead the Way

The Pentagon's McNamara-era philosophy is woefully insufficient to a world that changes daily. U.S. Special Operations Command is working to craft a better one.

Many proposals to speed up the Defense Department’s adoption of artificial-intelligence tools cite flaws in the acquisition process or the need for technological innovation or experimentation. But at its heart, AI adoption is not a tech problem or an evangelism problem: it is a management problem. DoD is hamstrung by a management system born in the 1950s and 1960s. The Department needs a new management philosophy, enabled by data, analytics, and AI—and it needs a place to incubate this new way of doing things. Special Operations Forces, a microcosm of DoD and a pioneering organization, is poised to be a pathfinder.

For all the important changes of the past half-century, DoD’s management system remains anchored to the federated approach of industrial titans that won World War II and the systems analysis/cost control methods introduced by Robert McNamara. To be sure, one of McNamara’s management precepts was “get the data.” But approaches that may have worked when the Library of Congress held an estimated 124 gigabytes of information are hopelessly overmatched when the world generates millions of gigabytes per second. The decision-making processes and cycles that were appropriate for the pre-internet era are now inundated by the force created by the mass and acceleration of data.  The central question is no longer an optimization problem of defense resources to strategy, as it was during McNamara’s era—it is how to accelerate performance to keep pace with the rate of change in the modern world.

Certainly, data, analytics, and AI have begun to infiltrate the Department’s business and operational matters. Leaders today have unprecedented visibility to “see” the Department in ways that McNamara could not have envisioned. Two of the most prominent examples are the Advancing Analytics, or ADVANA, program, which has made substantial progress in tying together the Department’s business data; and Project Maven, which has made similar gains on management of the operational and tactical matters. Data interoperability provides an opening not just for basic business intelligence, but for driver-based planning and true enterprise business performance management. Promising efforts have emerged to help illuminate whether investments generate desired outcomes. Some recent examples include the Department’s “Pulse” initiative to help measure business health and the Navy’s “Performance to Plan” effort, which aims to improve tactical aircraft readiness. 

But this piecemeal approach is insufficient to the urgent need to harness AI amid near-peer competition. Without a new management philosophy, enabled by data, analytics, and AI, DoD will be hard-pressed to deliver agility and improved performance in the face of rapid change.

SOF can set the standard

The past two decades of counterterrorism operations saw Special Operations Forces burnish their reputation as a pathfinder for the Joint Force—including by producing modern data management and AI tools that are now widely used to great effect. But a new era brings new demands. The latest National Defense Strategy emphasizes campaigning—across geography, domain, and time—as a key “way” to achieve the “ends” of maintaining a rules-based international order. The nation needs SOF to campaign globally in support of integrated deterrence while continuing counterterrorism operations and standing ready for crisis response. And with stable, timely budgets less frequent and Overseas Contingency Operations funding a thing of the past, SOF must make better use of its taxpayer dollars. 

Machine augmentation of decision-making is not just a tactical problem – it is a strategic problem.  It is no longer enough for Special Operations Forces to ensure proficiency on the battlefield. Better business operations and cross-regional campaigning are needed for long-term success against strategic competitors and financial topline pressures. The central problem is to increase performance—both to go faster and to lift all boats in our own enterprise and with our joint, interagency, and international partners. The solution must include adopting data and ML/AI for enterprise management, business operations, and campaigning. 

U.S. Special Operations Command is the perfect testbed for new management theories based on data, analytics, and AI. A microcosm of the Joint Force, it has operational responsibilities executed in coordination with and through the geographic combatant commands. It also has service-like authorities to organize, train, and equip special operators. SOCOM has an annual budget of around $14 billion; a diverse and global workforce of some 70,000; and air, land, and sea assets. Its data has the good and bad pathologies of the all services and combatant commands. If SOCOM can learn to use data, analytics, and AI to improve business operations and campaign globally, it can provide a gold standard for others in DoD to emulate. 

The good news is that the command need not start from scratch. Just as the Pentagon adopted the management playbook from General Motors in the 1950s, SOCOM is drawing inspiration from large successful companies that have digitally transformed management and operations. It is blending these digital-era approaches with lessons from its recent decades of counterterrorism and crisis-response operations: a team-of-teams mindset, relentless mission focus, data-driven tactical operations. 

The result is a blueprint for the Joint Force. Data-driven management operations create a virtuous cycle that drives greater enterprise performance, just as the “find, fix, finish, exploit, analyze, and disseminate” construct drove better performance for special operators and their joint, interagency, and international teammates. Data and machine learning improve performance at all echelons. 

Three key areas

In practical terms, modernizing management means transforming and aligning three key areas: planning, data, and governance.

Professional planners in DoD are schooled on Measures of Performance and Measures of Effectiveness, but dynamically assessing campaign performance has always been a challenge. Planners might draw inspiration from an approach pioneered by Intel and Google: the Objectives and Key Results construct. Clear OKRs with time-bound, measurable objectives and an associated data structure that provides real-time feedback on performance are within our grasp—paving the way for truly data-driven trans-regional campaigning in support of integrated deterrence. The same applies to other managerial priorities that are associated more with the business operations of SOCOM: personnel, materiel, and resourcing portfolios. Data analytics and machine learning can augment enterprise-wide planning to bring implementation and outcomes into greater focus.  Data analytics and eventually ML can also reveal opportunities to drive greater efficiency and effectiveness into these bureaucratic ecosystems, illuminating trade-offs needed to accelerate modernization efforts. 

Making the most of data requires focus on the critical, though less flashy, aspects of digital transformation—mainly data stewardship, data interoperability, data quality, and digital literacy. It means accountability in executing the “data decrees” and creating a data-centric architecture that enables data interoperability. SOCOM is investing in such an architecture and the foundational elements necessary to become a data-drive organization. The goal is to instrument the Command for greater performance in execution.

Governance means ensuring broad enterprise understanding of and alignment on goals and objectives. It requires assessing performance as revealed by the data. The SOF emphasis on “truth over harmony” is tailor-made to benefit from data-driven, evidence-based insights. Data and machine learning have the potential to transform traditional governance processes by establishing a single source of truth for enterprise operations, freeing up time to instead focus on resolving barriers to execution. 

SOF needs transformation in all three areas to move from a data-informed organization to a data-driven one that can harness machine augmentation for its business operations and campaigning. And if this pathfinding is successful—and it must be—the rest of the Defense Department has a blueprint to follow.

Dan Folliard is the Chief Digital and AI Officer at U.S. Special Operations Command. He is a career civil servant and was formerly a senior executive in the Office of the Secretary of Defense, serving in both OUSD(Policy) and in the immediate office of multiple Secretaries and Deputy Secretaries of Defense.

The opinions expressed here are the author’s and do not represent an official position of U.S. Special Operations Command or the Department of Defense.