(As featured in https://warontherocks.com/2020/08/dont-just-copy-and-paste-a-better-model-for-managing-military-technologists/ with Jim Perkins and Richrd Kuzma)
As software has eaten the world, there are no more “tech” and “non-tech” companies. Instead, there are organizations that leverage data and automation, and organizations that fade into irrelevance. The Army risks irrelevance if it does not address its digital talent crisis: It does not have enough digital talent and it does not manage its digital workforce well. The U.S. Army recognizes that it needs modern tools — software, data analytics, cloud computing, etc. — to succeed at multidomain operations, its strategy to fight and win across land, air, sea, space, and cyberspace on a 21st-century battlefield. Creating these capabilities requires a digital workforce: software developers, systems engineers, product managers, data scientists, cloud architects, user experience engineers, technical program managers, and other roles. The recent creation of the Army Futures Command Software Factory is a positive step to address the digital talent gap, but the Army should be bold and create a “Digital Corps” to manage its growing digital workforce.
Despite years of recommendations — and explicit direction in the 2020 National Defense Authorization Act — to build digital expertise within the workforce, the Defense Innovation Board wrote last year that the Pentagon has “taken only modest steps.” The Army has made some small, disparate efforts to harness uniformed digital talent through the Army Digital Service, AI Task Force, Enterprise Cloud Management Office, and the new Software Factory. But enterprise-wide change cannot be sustained through piecemeal efforts. Furthermore, “upskilling” or providing mid-career training to these servicemembers without changing their personnel management systems will create digital orphans — skilled servicemembers with no home (that is, no career track) and no institutional protection within the Army. It is not enough for the Army to focus on putting the right people in the right places at the right time — it should also develop strategies for keeping them there.
The Army should create a Digital Corps as a 21st-century talent management framework for 21st-century skills. It should be modeled after another field for highly skilled specialists: the Medical Corps. Doctors, nurses, and healthcare administrators are critical to the functioning of military healthcare, but don’t fit within the traditional talent management system. The Medical Corps manages personnel policy, promotions, training, and certification of medical personnel in a way that the regular Army could not hope to.
To understand where and why the Digital Corps fits within the talent management system, we should understand the other talent models: traditional, specialist, and professional.
The traditional model manages personnel in the core fields of the Army — infantry, armor, engineers, and artillery. These fields are unique to the Army and require training only within the Army. Cadets are not required to have particular skills to enter and pick branches mostly on preference. Since the skill requirements for these fields are well-known and for the most part static, well-established career managers treat personnel and roles as interchangeable parts. A job in these fields merely requires a person of a certain career field and pay grade, such as an infantry captain — any will do.
Personnel — both officer and enlisted — in traditional fields move between leadership, staff, and developmental assignments as their careers progress, grooming them for the broad needs of “generalship.” Indeed, most of the generals in the Army come from, value, and select these traditional fields and skill sets for promotion, according to a recent RAND study. Military planners have had decades to learn how to properly assign people and resources to traditional units. The resulting system is an assembly line that moves traditional personnel from one established billet to another, treating them as items to assign to a unit, where the goal is filling vacancies, rather than a personalized approach to talent management where the goal is to align skills to teams and missions. The Army built the cyber branch in this model and has struggled since inception to retain talent.
Specialist fields like aviation and special operations forces are somewhat similar in that they are also unique to the Army — though there are civilian helicopter pilots, few shoot Hellfire missiles or get shot at. Specialist fields also have an established Army training pipeline. These branches differ from infantry, armor, engineers, and artillery in two key attributes: They require a certain level of assessed proficiency — either the Army Flight Aptitude Screening Test or special operations forces assessment and selection — for entrance, and they have well-known but highly technical skill sets that atrophy quickly.
The military recognizes that pilots lose proficiency in flying quickly, so time in the cockpit is prioritized. Unlike traditional branches, aviation is heavily staffed with warrant officers due to technical expertise requirements. Putting aviation officers into desk jobs actively deteriorates their skills — they get better at flying by flying a lot. Aviation officers also have many specialization options based on aircraft variants and special training courses, so the branch has many occupational specialties and additional skill identifiers to track and manage its talent pool within the same industrial-era framework that traditional branches use.
Finally, the military professional corps — medical, legal, and chaplain — functions differently from traditional branches. These roles are not unique to the Army and have civilian training and accreditation pipelines, generally require expertise prior to entering the Army, and require specialized skills not needed within the traditional Army. In medicine, the services delineate between technical experts (e.g., Medical Corps) and administrators (Medical Service Corps). A hospital could not run without both sides of the house. This model allows officers who are highly skilled experts to practice their professional skills rather than serve in traditional leadership roles as platoon, company, and battalion commanders or as staff officers (although they can). This model — allowing physicians to serve a career from captain to colonel while remaining a clinician — is similar to how technology firms delineate between technical experts (software engineers, data scientists, technical program managers) and operations roles (marketing, logistics, account managers) while allowing for technical experts to be promoted in non-managerial positions to focus more deeply on their area of expertise.
The Army also recognizes the importance of specialist labels in lieu of an over-generalized “medical officer” label. Immunologists and orthopedic surgeons are both physicians, but you would prefer the latter for your knee surgery. Instead of the “any infantryman will do” mentality, these professional corps have a variety of technical specialty and skill codes to more precisely define areas and levels of expertise. Assignments managers work closely with each person to account for unique roles and skills not tracked in the system (for example, a physician who also does research or medical education).
None of these models is perfect for managing the unique challenges of the digital workforce, but all have something to offer in shaping the new Digital Corps. The traditional model works well for a force that needs to scale across hundreds of thousands of people but fails for people with specialized skillsets. The specialist model prevents skill atrophy and prizes specialization but struggles to account for the expertise required in accessions and dynamic changes in required skills. The professional model comes closest in shepherding highly skilled individuals within the Army talent system, but still requires tailoring to fit the needs of the digital force.
Unfortunately, there will always be some uncertainty or gap around military force planning for emerging technologies. No model can project what talent is required to man future technology force structures with the precision needed to fit within existing Human Resources Command manning structures. The need for a holistic understanding of both talent development and assignment requires a model that communicates the full personnel lifecycle. Said differently, the Army needs an even more flexible model.
Rather than adopt an existing “professional” model lock, stock, and barrel, we will evaluate digital talent management through the Army’s People Strategy. The strategy is built on four pillars that comprise a soldier’s entire service experience: acquire (recruit, screen, hire), develop (train), employ (assign), and retain (pay, evaluate, promote).
Step one in building a digital force is acquiring digital talent — identifying, recruiting, screening, and hiring candidates. Despite arguments over the merits of technologists in uniform, the Army’s decision to invest in AI experts and software developers signals that they clearly recognize that making digital skills the exclusive domain of civilians and contractors creates a dangerous condition where the Army doesn’t understand how its core information technology systems function. This lack of awareness invites future strategic risks as key leaders are unable to make rapid and effective decisions about employing or developing nuanced capabilities. Since digital skills and best practices are more prevalent in industry and academia, the Army should leverage a blend of all relevant ways to acquire talent. These include Cyber Excepted Service authorities (a Department of Defense effort to hire skilled civilian cyber professionals), Cyber Direct Commissioning authorities (which enable skilled civilian technologists to serve in uniform at ranks ranging from second lieutenants to colonels), and future investment in cadets.
First, the Army should look to traditional commissioning sources like the Reserve Officers’ Training Corps and service academies. Once cyber security was recognized as a defense priority, West Point and Annapolis established world-class cyber centers and cybersecurity majors. Digital skills like software development and artificial intelligence have been identified in the 2019 and 2020 National Defense Authorization Acts — therefore the Army should respond by increasing the supply of digital officers. The new applied statistics and data science program at West Point is a start, but the efforts should not stop there given the unfilled need for software developers, cloud architects, and other skills.
Investing in cadets should be complemented by leveraging direct commissioning authorities, but the military should move carefully. There are risks associated with relying heavily on direct commissioning for the Digital Corps — especially now as the culture is forming. The shortened training program that direct commissioning entails is a clear trade-off in the service contract: The service accepts risk of less military training, decreased indoctrination, and higher projected turnover rates in exchange for commissioning civilians who already have special expertise. Filling the nascent Digital Corps with “outsiders” who don’t understand military culture and norms will erode trust and potentially create a culture of “technologists who happen to wear a uniform” instead of “soldiers who happen to code.”
In addition to creating an increased supply of digital officers, the Army should create the relevant assignments for these leaders to fill. The Naval Academy built a $100 million cyber center but only allows one midshipman per year to join the Navy’s Cyber Warfare Engineer field. In a similar vein, while the Army has placed some junior leaders into digital transformation roles, many other emerging technologists have been blocked due to the lack of direct sponsorship from a general officer or bureaucratic constraints like their professional timelines or professional education requirements.
Until the Army has a robust digital personnel acquisition pipeline it should focus on developing digital talent from the personnel it has. It is important to distinguish familiarity from expertise. Although creating a “digitally literate” force is necessary in the digital world, it is no replacement for digital experts. Though soldiers across all branches are taught battlefield first-aid, the Army would never rely on artillery officers training in combat lifesaving in lieu of the Medical Corps. Coding boot camps and entry-level certifications create familiarity, but will not yield the full transformative value of software. The military will need to rely on experts who can see through hype and evaluate expensive education programs honestly.
Developing talent can again rely on existing programs developed by civilians in academia and industry. A great example of this is the Army’s AI Task Force educating a cohort of “AI Scholars” at Carnegie Mellon University. For members of the Digital Corps, encouraging extensive use of the Pentagon’s Training With Industry program would be an invaluable broadening assignment rather than merely a “nice-to-have” and could be considered a requirement for promotion to either major or lieutenant colonel. Soldiers could make themselves competitive for these opportunities by seeking out some of the free and low-cost online courses taught by world-class experts.
Lastly, unlike the medical and legal professions, there is no centralized accreditation board for technology expertise. There are many certifications, but not a true licensing credential. Degrees in computer science, electrical engineering, and information technology are strong proxies, but some candidates — including the authors — are self-taught and learned by doing. The tech industry values skills over credentials and tests this through technical interviews. Military bands conduct auditions. The Digital Corps should, too.
Digital problems — like medical maladies — require certain expertise to fix, so digital experts cannot be treated as interchangeable parts. A machine learning expert may know little about website design, so the Digital Corps will need a detailed and evolving taxonomy of digital skills. Thankfully, the Army Research Institute and Talent Management Task Force have been working on a taxonomy of knowledge, skills, and behaviors, and the Army is working to build a software-based talent marketplace that helps candidates choose jobs and jobs choose candidates based on skill supply and demand. Instead of keywords on a resume, military technologists can corroborate their technical skills via GitHub or a Code.mil repository.
Once these roles are created, military technologists should be flexibly managed to ensure that they are assigned to teams that can best use their capabilities. If commanders had to compete for talent by pitching their projects and problems to technically talented servicemembers capable of building digital solutions, commanders might be less likely to relegate those hard-won technologists to PowerPoint slide-making. The need for a centrally managed talent pool was a core part of the Defense Innovation Board’s Workforce Now recommendations, and should be included in the Digital Corps.
Once employed, retaining talent is a two-part challenge: The Army can choose to show people the door (involuntary) or people can choose to leave (voluntary). Bleeding talent in either way is costly.
Involuntary attrition occurs most frequently when a service member is passed over for promotion, reaches a mandatory retirement date, or fails to meet fitness standards. During promotion boards, officers are grouped by specialty into competitive categories. Competitive categories collect specialties that can compete against each other for promotion. Creating a competitive category for digital technologists — as there is for medical — will likely improve officer promotion rates and may strengthen bonds between existing functional areas in technology areas by providing a collective sense of purpose and identity. Furthermore, having technologists evaluate other technologists may help to focus the leaders on technologists’ contributions, whereas it’s plausible that non-technical administrators rating technologists may emphasize administrative minutiae or physical fitness over digital skills.
Frustration is likely to be a larger factor and is harder to fix. What happens when a highly skilled developer is not given the right tools to create software, needs to jump bureaucratic hurdles to do her basic job, or should leave her field of expertise for three years to check a box for a broadening assignment? These are harder problems. The last, however, is a talent management issue that can and should be solved with the Digital Corps model.
Let’s assume that the services commit to creating a Digital Corps. What happens next? What outliers will they need to confront? Organizational change often fails when leaders fail to consider second-order effects or edge cases.
Immediately, the Army should prioritize Reserve Officers’ Training Corps and West Point computer science and information technology programs. These will take time to produce results, but they are strategic investments. Next, they should allocate training funds for online courses and recruit volunteers for the training. The Air Force has already made initial progress here with its investments and experiments in DigitalU and the Project Nexus coding bootcamp. Those who take the courses and pass are the top candidates for the initial Digital Corps billets, in addition to anyone who applies directly based on existing skills. This will give Army leadership time and space to maneuver and address the larger challenges.
Then the Defense Department needs to address how to get tools and authorities into the hands of these teams. The Army and Air Force are each working to create secure software development platforms, but each factory should be able to purchase licenses and hardware or provision cloud resources rapidly, or these experts will sit idle, grow frustrated, and leave.
Next, the services should understand their mid-term roadmap for civilians and contractors. Hiring civilians requires much shorter onboarding and they can make valuable contributions in pay grades of GS-7 to GS-15 or even as a “highly qualified expert,” but big tech companies aren’t the only talent poachers. The services should work to avoid losing uniformed talent to become contractors or civilians in other federal agencies due to pay or removing them from roles where they get to use their digital skills. Perhaps just as importantly, these personnel should be provided with unique leadership and cultural structures to flourish, which are distinct from the command and control concepts found in traditional military hierarchies.
Lastly, there are already existing specialties with overlapping skills, and service leadership should sort out equities. Cyber should be brought under this new corps and Army cyber should make clear the difference between their new tool developer specialty and a software developer. Operations research specialists are similar to data scientists but may need further training, while network operations should get pulled into the Digital Corps directly. The services should also create a path to upskill selected volunteers from existing branches — the leader of the Army’s new software factory is a helicopter pilot after all.
The questions presented here are not esoteric — managing technologists is a “right now” problem for Army Futures Command, Special Operations Command, and multiple Air Force, Marine Corps, and Navy software organizations. They are writing code and building tools that are in the hands of users in contact with the enemy every day.
It’s easy to shoehorn technologists into industrial-era models, but following the traditional branches will inevitably lead to the same mismanagement and attrition that is fueling current calls for reform. Talent management is an investment, not a cost, and the effort to execute a bold vision for uniformed technologists will pay off over the long term.
The military of the future will not merely shoot slightly further and move marginally faster. By creating software factories, and the talent factories that sustain them, the Army and Air Force clearly recognize that the character of conflict is changing — just as it did during previous industrial revolutions. If the U.S. military wants to remain a dominant force in the world, it should keep adversaries at bay. How better to demonstrate the characteristics of offense — surprise, concentration, audacity, and tempo — in the current security environment than by seizing the opportunity to create a force built for the information age?