Unlocking the Future of Army HR with Predictive Analytics

By Col. Gregory S. Johnson, Director, Officer Personnel Management Directorate, U.S. Army Human Resources Command and Lt. Col. Kristin C. Saling, Deputy Director of People Analytics, U.S. Army Deputy Chief of Staff, G-1

A data-informed, predictive, and forward-looking human resources?enterprise is a critical ingredient to winning the race for talent.?In a talent environment that grows more competitive by the day, the Army needs to put its most proactive foot forward to acquire, develop, employ, and retain the most capable individuals, and upgrade its HR systems and procedures from passive engagement to predictive, positive engagement.?

With the growing complexity of talent and personnel data, answering traditional HR questions about rates must give way to predictive and targeted retention modeling and engagement. Models that extract requests for skills in talent marketplaces and forecast the need for those skills in recruiting and training platforms, and comprehensive data collection efforts to capture important quality-of-life and quality-of-work metrics that contribute to the overall employee experience would allow us to shift strategically.??This is important to operating in a modern environment where by data is readily available. Our internal capabilities need to shift to match this emerging requirement.

This shift also takes a proactive, engaged, and skilled HR workforce, an Army familiar with data and analytics, and leaders who challenge their teams to analyze current policies, procedures, and regulations to ensure we update business processes to align with the information our data and analytics are providing us.?

The Army’s HR enterprise is already a data enterprise. It collects and manages?a massive amount of personnel data, tracking and forecasting strength, predicting losses, and making recommendations on the quantity of Soldiers to access, promote,?and distribute to schools, training, and operational locations.?However, there is a significant difference between managing through descriptive analytics and predictive?analytics, and?forecasting by quality and talent attributes rather than quantity.?

Analytics professionals group analytics into four basic categories – descriptive, diagnostic, predictive, and prescriptive.?These analytics progress from least to most complex, and least to most data intensive.?

·???????Descriptive analytics summarizes attributes of your data and tells you what has happened (e.g. the number of accessions and promotions in the month of March).?

·???????Diagnostic analytics?measures attributes of data and past data to answer why something has happened (e.g. looking at historical trends for accessions and determining why the Army will or will not make the recruiting goals for the year).

·???????Predictive analytics uses data to predict future trends and forecasts (e.g. how many people are likely to leave the Army in a particular grade or military occupational specialty).

·???????Prescriptive analytics prescribe an action to take in the future (e.g. recommended adjustment to recruiting targets based on current accessions and predicted attrition).?

As you move from descriptive toward prescriptive, the models get successively more complex and more powerful, but require exponentially larger amounts of high-quality data?and increasingly skilled analysts.?

The Army has attacked its data problem through the consolidation and curation of large data sets through analytic platforms like Army Vantage and the Person-event Data Environment (PDE), and establishment of enterprise programs like the Integrated Personnel and Pay System – Army (IPPS-A).??

But the Army must proactively plan how to use these data lakes and analytic programs in the HR space to answer the right personnel questions to create the necessary analytic tools in time to support important decisions.?These models can’t be built without proper planning, data collection, data architecture, and skilled data practitioners in the HR enterprise.??

Building predictive data models throughout the HR enterprise will enable the Army to proactively use data to do more than?engage Soldiers and unit leaders on a transactional basis. The Army can take on a number of challenging?problem?sets with these models.??

1.?????Predictive and Targeted Retention. The Army is developing predictive retention models that forecast attrition to the individual level and provide insight about where the greatest risk exists. These models, paired with an understanding of how and when to apply retention interventions, can inform a targeted retention strategy to maximize the use of limited resources?with targeted engagement.?

2.?????Performance Prediction. Persistence models are already helping special operations commands determine candidates most likely to succeed in training. When applied to other elite, high-cost training programs, these models can assist the Army with targeting recruiting efforts at the optimal candidates. ???

3.?????Emerging Skills Analysis. Predictive models applied to performance, assessments, and market selection can help the Army identify emerging knowledge, skills, and behavior (KSB) requirements. Recommendation models can help units find the best candidates to fill these emerging?requirements. The Army people enterprise can use these predictive models and identified skills to also inform requirements across the enterprise, and feed information to training organizations that would allow them to adapt their curriculum to develop more of these skills according to demand.

4.?????The Army people enterprise can use the models, add these skills to similar requirements across the force and the training enterprise can use them to create the right volume of the skillset to meet future demand. ?

5.?????Employee Experience. The Army captures a large amount of service member experience information through surveys. Using data from these surveys and predictive analytics, the Army can anticipate when Soldiers who have selected certain opportunities and had certain experiences might have difficulties that can be anticipated and mitigated. One example of this is capturing quality of life issues through the Department of the Army Career Engagement Survey (DACES).?

6.?????Optimal Teaming.?As the Army continues to study and conduct research on how to create and lead the best teams, predictive analytics can play a role in matching the types of teammates that are most likely to work well together. When recruiting and selecting people for positions, leaders need to have a good understanding of what KSBs they need not just in an individual, but to round out the team, and predictive analytics can help leaders identify the KSBs needed to fill in their blanks.?

Determining the problems that the Army needs to solve and questions we need data to answer is an essential part of having a dynamic and proactive HR?modernization program. Having large amounts of data is great for a predictive analytics program, but creating analytics to ask the right questions, and having the vision for proactive personnel management, starts now, with an engaged, upskilled, and analytically-minded workforce.?How do we get there??

Our first and most important step to prepare for a modern, dynamic and data-driven environment is to look at our talent. The right application of talent, and for that, a talented?upskilled?HR workforce is essential to pulling this off.?HR leaders need to instantiate the data collection and data architecture needed to support predictive modeling,?hire?or upskill data engineers and analysts, and upskill the HR workforce to meet the Army’s talent management needs, now and in the future.?We can and must do this.

Colonel Johnson is a career Army Adjutant General Corps Officer. He was a Distinguished Military Graduate of the University of San Francisco, earning a Bachelor of Arts in United States History. He also holds a master’s degree from the United States Army War College, a master’s degree of Policy Management from Georgetown’s Public Policy Institute, and a master’s degree in Education from the University of Oklahoma.

Lt. Col. Kristin Saling is the Deputy Director of Army People Analytics in the Office of the Assistant Secretary of the Army (Manpower & Reserve Affairs). In this office, she is responsible for developing, managing, and synchronizing the Army's strategy, policy, research, and analytics development across the Army's people enterprise. Lt. Col. Saling holds a Bachelor of Science in Operations Research from the United States Military Academy and a Master of Science in Engineering Management from the Missouri Institute of Science and Technology and a Master of Science in Systems Engineering from the University of Virginia.