Not too many years ago, if the Army had wanted you to have an opinion about your next assignment, it would have issued you one with your duffel bag. The same had been true for units, as well. Leaders at every level largely left it to the Army to fill their formations with individuals based not on talent, but on having the right rank, branch, and availability date. In 2019, this changed when the Army introduced a market into its assignment system. Gone were the constraints that offered familiar comfort to both units and officers but that also prevented maximizing the talents and productivity of our officer corps. Within the assignment marketplace, some found the competition and ability to tell their own story liberating. Others cursed the comparative chaos and extra work it created. However, in the end, the Army’s first large-scale implementation of assignment markets not only demonstrated they can improve how we assign our people, but they can also uncover for units and officers the keys to succeeding in future markets.
This fall, nearly fifteen thousand officers and five hundred units are participating in the Army’s second major iteration of the Army Talent Alignment Process (ATAP) through the Assignment Interactive Module Version 2.0 (AIM 2.0). Analysis and feedback from last year’s ATAP showed many areas of success. Officer preferences, often shaped by their talents, were the most important factor in determining future assignments; units could directly influence who entered their formations by ranking officers they preferred; and the Army discovered additional information about the many talents resident in our officer corps and about the unique talent demands of each job. More importantly, the Army learned several lessons from its first iteration of ATAP that can help officers, units, and the Army improve.
What Officers Can Do
Officers can greatly influence their chances of receiving the jobs they prefer in the market. Most importantly, moving officers should build their resumes in AIM 2.0. If officers are the engine of the marketplace, then the information they share on their AIM 2.0 resumes is the fuel that keeps it humming along smoothly. Complete and informative resumes allow units to better identify officers with backgrounds or experiences that are a particularly good fit for a specific job or type of work. Surprisingly, though, only 60 percent of officers in last fall’s marketplace took the time to describe their talents within the AIM 2.0 resumes. Choosing not to do so is a missed opportunity as officers with resumes benefited significantly, receiving 40 percent more #1 votes from units than officers without detailed resumes. There is no magic formula for what to include on a resume, but discussions with units suggest that most were looking for officers who could write reasonably well and who put forth the effort to prepare a resume.
In addition to building resumes, moving officers should interview with units they are interested in joining. In a survey given to all units at the end of last year’s marketplace, 35 percent said that interviews were the most important factor in determining how to rank officers in the market. Even though AIM 2.0 is online, the marketplace is designed to facilitate person-to-person interaction, albeit predominately via phone or video conference. Moving officers who reach out to unit POCs and schedule interviews early in the marketplace have better chances of finding the jobs they want than officers who wait for units to contact them.
A third way an officer can maximize the benefit of ATAP is to submit truthful preferences to the assignment market. While this is not immediately obvious, it becomes clearer once an officer understands how they are paired to jobs after the marketplace closes. The Army uses a deferred acceptance algorithm to match officers to jobs. While that might sound complicated, it is a relatively straightforward process—but one that works best when officers and units submit truthful preferences. The advantage of a deferred acceptance algorithm is that it produces the best match even when a #1 to #1 match—meaning an officer and a unit both rank each other as number one—is not possible. For example, if a moving officer’s most preferred job is in the 82nd Airborne Division, and the 82nd ranks him #2, then it is still in the officer’s best interest to rank this job #1. This is because there is no guarantee the 82nd will receive its most preferred officer—who may be looking at the 101st. In other words, officers cannot achieve a better outcome by misreporting their actual preferences, or “gaming the system.” In fact, officers often hurt themselves when they do this. The design is intentional and allows officers to list “reach assignments” without penalty. So, go ahead and reach for that dream job knowing your chances of success in the market will not be diminished. (You can learn more about deferred acceptance algorithms at here or here.)
There is substantial evidence that officers did not understand this point in the 2019 market. Post-marketplace surveys revealed that 31 percent of officers admitted that they did not place their true first preference #1 in the marketplace. Furthermore, roughly 75 percent admitted that at least some of their preferences were not truthful. Unfortunately, these officers potentially hurt their chances of receiving their most desired assignments. After all, no algorithm can know what you truly want if you do not share that information. If you are a moving officer, your best strategy is to truthfully rank as many jobs as possible according to your true preference for each job. The deeper you rank jobs, the better chance you have of getting a job that aligns with your preferences and takes advantage of your unique talents.
What Units Can Do
Just like officers benefit if they truthfully preference as many jobs as possible, units benefit by ranking as many officers as possible. Units that rank all of the officers they are interested in will do better than units that only consider officers who rank their jobs #1. If a unit does not receive its most preferred officer, the deferred acceptance algorithm will attempt to match that unit with its second-most preferred officer, and so on. Therefore, units should not exclusively chase #1 to #1 matches. Instead, units should rank all officers that interest them, which will increase their chances of getting an officer who is a good talent fit.
A second suggestion for units is to put your best foot forward and professionalize your hiring processes. Many officers in last year’s market found little more than boilerplate job descriptions and had difficultly contacting units and scheduling interviews. Lacking detailed information about the job and unit’s hiring process, officers would call the incumbent listed in a job’s AIM 2.0 profile. This often resulted in a negative first impression of the unit and a missed opportunity for the unit to connect with an officer interested in its job. The best units made it easy for officers to understand the unique aspects of their jobs, were enthusiastic about selecting their teams, and executed well-coordinated and structured interviews.
We also recommend units engage with officers in their formations who are in the marketplace and who expect to move this summer. For units, these officers can provide a great resource to identify other units’ best practices, share what attracts officers to jobs in the marketplace, and help scout for their talented replacements in the market. Commanders, likewise, can serve as references on the AIM 2.0 resume, supervisors can call units on behalf of their moving officers, and units can lead professional development workshops to train their officers on how the marketplace works. Because both the unit and officers within the unit are in the marketplace, both can benefit from what the other is seeing and doing.
What the Army Can Do
Finally, there are several changes the Army should implement to make the marketplace more effective for both officers and units. First, the Army should better educate the force on how officers pair to units. As the first ATAP iteration showed, officers and units that work with ATAP incentives will have better outcomes than those that do not take advantage of ATAP opportunities. As more units and officers participate in the marketplace, they should naturally grow more acquainted with how to make the process work best for them. But the Army should also actively seek to inform units and officers on how best to leverage the marketplace to their advantage. A marketplace with well-understood rules and processes builds trust, encourages more information sharing, and is more effective in aligning talent.
Second, the Army should limit the number of “signals” officers can send to units to indicate their interest. During last year’s marketplace, officers adjusted their preferences multiple times to signal interest to different units at different times. This resulted in an unlimited number of signals for officers to use and generated substantial problems. It also frustrated many units who could no longer be certain which officers were genuinely interested in them and which officers were likely to change their preferences immediately after a phone call or interview.
The Army should give every officer a specific number of “market coins” to be sent to units of their choosing independent of their preference list. This would help units learn which officers are interested in their jobs without requiring officers to repeatedly change their preferences. The method of using signals to quickly identify interested participants has a proven record of making other matching markets more efficient. Admittedly, officers should think carefully about how they use these signals, but with this mechanism in place they would no longer need to strategize or manipulate their preferences, which are kept secret. Signals also put officers on a level playing field: every officer can receive the same suggestions regarding how to use their signals, and no officer can gain an advantage by repeatedly adjusting preferences. Furthermore, when officers have a limited number of signals that cannot be reused, units that receive the signal can be confident in the officer’s intentions. Likewise, units that receive very few signals will know that they need to work harder to garner more interest in their positions.
A final recommendation is to continue building tools within AIM 2.0 that help participants find the right talent in the marketplace. Units and officers alike found AIM 2.0 difficult to navigate from a talent perspective. The Army should start by including dropdown boxes categorized by relevant talent dimensions such as previous military experience, education (level and discipline), civilian credentials or qualifications, language experience, and hobbies or interests. KSBs—Knowledge, Skills, and Behaviors—which units can use to find officers with particular talents, should be incorporated into these talent dimensions. Such dropdown boxes are most effective when units have the ability to customize the categories they search within, and when the Army can add to or remove elements from dropdown fields as more information becomes available. For example, if several officers reveal that they have a particular computer programming skill on their resumes, then this specific skill could be added to dropdown boxes in future markets.
The potential payoff of the Army’s transition from a centrally managed officer assignment system to an assignment marketplace is high. However, it will take several more years of hard work from officers, units, Human Resources Command, and other stakeholders to continuously improve it. As officers gain more control over their careers and units gain more control over who joins their teams, better talent alignment will increase productivity, improved officer satisfaction will boost retention, and the sharing of talent supply and demand data will give the Army valuable information about its most important resource—its people.
Maj. Kyle Greenberg, Lt. Col. Mark Crow, and Col. Carl Wojtaszek are labor economists teaching in the Social Sciences Department at the United State Military Academy. Their research focuses on talent management, labor market outcomes, and human capital development within the Army.
The views expressed are those of the authors and do not reflect the official position of the United States Military Academy, Department of the Army, or Department of Defense.
Image credit: Cpl. Rachel Diehm, US Army
Thank you for this well written article. I am going through this process for the first time this cycle. One issue that you did not introduce is HRC intervention. In effect, a particular branch can choose to “manage” key positions. An officer will receive an email stating that they were selected for a position, with instructions to rank that position #1. The unit is also instructed to rank the selected officer #1 artificially creating a 1 to 1 match. The result is a partially fixed market.
The second element that this article does not address is the use of labels. Officers are assigned labels (e.g. Nominative, Post CSL, JDAL) which can be used as screening criteria for what positions they can match with. It is unclear if this would actually prevent a mutual match, but officers are being instructed that they will only be assigned to positions that match their assigned label. For instance an officer not identified with a nominative label may not match with a nominative assignment, even if the unit and officer mutually select each other. Conversely, a an officer with a Post CSL label cannot match with a position not labeled as Post CSL.
This is my second time using the marketplace and I like it quite a bit, in theory. As Perry mentioned, a lot of jobs in our category have been pre-slated and they don't reflect as such. It's very frustrating trying to guess which positions are already filled and which ones are open for contact. You can only signal interest to so many units and trying to not waste those signals turns into attempting to game the system. I'm not sure which ones are still open, so I'll switch up my preferences every so often to try and maximize my signal to the ones that are actively looking. It would be a little easier if the unit contact information was updated and accurate. I would also recommend allowing for more signals. I have known officers that ended up with their 45th choice or more. With a large pool of jobs, you never know which one will be a great fit for sure. A little more wiggle room would be awesome.
The ATAA is actually not a faithful representation of the Gale-Shipley algorithm (GS).
1. The information about preferencing shared between unit and officer is an addition to the algorithm.
-Units cannot be expected to rate all officers in the marketplace. They therefore reasonably rely on the signal of interest in order to narrow their search space.
-Right now, this signal of interest is the ‘green check’ that signals to a unit that an officer has ranked the unit in the top 10% of their preferences.
-This cap naturally drives officers to change their preferences many times, in order to maximize the number of units that will see them as interested.
-In GS, rankings are not revealed before it comes time to match. The National Resident Matching Program solves this interest problem by requiring Medical Students to apply to the residencies they are interested in. This imposes a cost on the students (so that they cannot freely indicate interest to all programs) but allows the student to apply to as many programs as they wish (not arbitrarily limiting them at 10%). They also keep rankings separate from applications so that students can be honest in their preferences.
-AIM2 should incorporate some mechanism to show interest that does not rely on an officer ranking a unit in their top 10%. This could be a simple form for officers to fill out when applying to different units, and allow each unit to specify a short prompt in order to prevent officers from copying and pasting the same generic note of interest multiple times.
-The ‘top 10%’ rule also discourages officers from listing ‘reach’ units as the authors advocate. If only the top 10% of units ranked are alerted to an officer's interest, then wasting these precious slots on reach assignments is not in an officer’s interest. Instead it encourages officers to ‘play it safe’ and only rank assignments in their ‘target range’, in order to maximize the chance of being matched with something ‘good enough’ and avoid the possibility of missing each of the reach assignments and being relegated to the officer’s last choice.
2. A true stable solution to GS requires the ability to choose to go unmatched. That means, if an officer is not given any of the assignments they ranked, then they should not be mandated to fill the least desirable slots (those still left open). This would require the Army to incentivize undesirable positions with things like increased pay, responsibility, or other perks in order to attract talent to undesirable positions.
All this AIM2 business is allowing me to see what other interesting jobs are out there, sure I rank them, but I still reach out to the ones I care about. If some respond I moved them up in the rankings.
I am currently competing in the 21-02 marketplace and find the whole process immature, frustrating and distasteful.
1) Units are extremely uneducated about the AIM marketplace, its process and what the officer is dealing with on their end. Units may be reviewing officer resumes but aren't aware that they can contact officers and sell their unit to those officers. Those that I have dealt with that ranked me in their top 10% told me that they didn't know how to contact the officers. It appears that the information provided in the AIM system isn't available to the units beyond what is on the resume (which lacks any fields for POC info). Many units consider a phone call from an officer asking for information about the unit/position as a sign of extreme interest making for frantic phone calls from the unit and for uncomfortable conversations when they don't see the officer mark the unit with the "green checkmark". Units also are unaware of the business rules for the marketplace demanding years worth of OERs, PT cards, additional resumes and agreements to 2-3 years worth of service to the unit. I found joint units and agencies to be the most uneducated and to some extent hostile to individuals contacting them asking for more information or consideration as they believe they have already selected 1-3 candidates in the first phase and their task is complete.
2) Information about units and positions was substantially lacking 70% of the time. Many units are uploading power point presentations about the unit while failing to provide any useful details about the position, the sub organization/staff section the candidate would work in, and POC info. I can't count the number of times I had to contact the incumbent and ask them to speak with the hiring POC to contact me back because the information provided was wrong or the POC was "busy" with other matters. This lack of upfront information eats up a crazy amount of time on behalf of the officer while they are also doing their current job. Units are also failing to realize that officers have families and in addition to information about the unit and position, need to understand the family side of life. Schools, housing, extra curricular activities, traffic, spousal employment, etc. All of this information must be researched and discussed with the family for 50+ assignments; its daunting and emotionally draining. Providing as much of this information upfront as possible (or at the very least providing solid links to the information) would save multitudes of time for all candidates. Many units during interviews couldn't answer any family life questions and were not prepared to provide contact information for the installation.
3) The online AIM system is clunky for both officers and units. Officers have no place to record notes about assignments. Units have no place to record notes about officers. The AIM assignments and officer lists are not exportable to any format so everything must be done on paper. Additionally, the website itself suffers terribly from programing bugs flakes out scrambling the information provided on the page (requiring constant refreshes) or it will display information that was filtered out, causing confusion. There is no in system messaging service requiring units and officers to contact each other using traditional methods of email and telephone. Sadly not everyone is in the same time zone so most messaging falls to email and I don't know about most of you but I get way too much email already; as a result I have missed several emails from units. As a leader and mentor, I don't have access to the AIM system for my subordinates and thus no way to advise and counsel them on their assignment choices unless they screenshot and print me a list. Considering HRC has basically stepped out of the line of fire for assignments and is no longer willing to act in an advisory role for junior officers, seniors really need some way to know what is being offered so they can better counsel their troops.
4) There is a lack of closure/disclosure on behalf of the officer and the unit. Some units graciously contacted me and let me know where I was located on their list and I greatly appreciated that because it allowed me to rank other units higher on my list. Others played their cards close to their chest hoping to have the officer rank them higher on the list as a fall back if their #1 pick fell through. I'm sure this is also happening with officers who are being non-committal to their 2-n unit choices in hopes that they will be rank higher as a fall back to their number 1 pick. This cat and mouse game is very unprofessional and distasteful in my professional opinion but based on the system the Army set up, this seems to be how they want the assignment game played. The units I was upfront with and notified where they were on my Order of Merit List, dropped me from their top 10% the next day.