Giant headquarters are obsolete

We should distribute theater-level headquarters into many small nodes, because they’re less vulnerable and make better decisions.

I used to think giant headquarters were a new phenomenon. They’re not. Westmoreland had a staff of 3,000 in Vietnam. As you’ll see in the diagram below, we’ve changed very little since then.

I’ve served in a few JTFs myself, and I’ve interviewed others about their time in such places. I’ve talked with 264 officers so far, from recent SAMS grads to Combatant Commanders.

Below are four proposed reforms, each moving us closer to a remote, distributed headquarters. Most are already occurring at a small scale in a few units. It’s time to adopt them more broadly.

Typical headquarters of a Combined Joint Task Force. This article is also intended for component commands and Subunified Commands, which are similarly structured.

1. Reachback and halfback

Large staffs aren’t always bad. But it’s dangerous to put everyone in a single location, where they’re vulnerable to a missile strike, communications loss, or act of sabotage.

The intel community and special operations forces already make heavy use of reachback, a concept in which part of the staff works from homestation. This distributes risk geographically. A second option, which I call “halfback,” puts much of your staff in a nearby country. Eisenhower put most of his staff in London, and the counter-ISIS coalition headquartered in Kuwait. This makes it easy for staff members to travel forward when needed and keeps everyone in roughly the same time zone.

Let’s take this a step farther. Distribute most of your staff across three or more bases in different countries and the US, leaving just a small group forward with the commander. This forces your day-to-day work into remote collaboration tools, which in turn makes it far easier to involve collaborators from outside the organic staff.

2. Allow select staff to work fully remote

Imagine you’re the J3 at a corps headquarters. You’re about to deploy as a Joint Task Force, but your best iron Major is leaving military service. He has three kids and no interest in another deployment. So you say,

Joe, I want you to deploy from home. Stay in one more year, and be part of something that really matters. You can work from the office, from your own home (with SIPR), or from wherever you choose.

In return, I expect you to work lawyer hours. Do everything you can to make sure we create the best plan to achieve our desired end states

Not everyone does well with so little supervision. But I bet Joe will.

3. Make it easy to add outside players

Today, when you need an operational planning team, you do the best you can with the people on hand. Have you ever been on a planning team that faced decisions for which it was absurdly unprepared, but which someone else in government could have solved in an instant? Like allocating equipment to the Peshmerga with just a Wikipedia-deep understanding of Kurdish politics. Or deciding where to put Ebola Treatment Units, with no epidemiologist on your team.

It’s hard to assemble the right planning team quickly. This is harder than it was 50 years ago for Westmoreland’s staff, as more specialties become relevant to your Joint Task Force — from IO planning to cyber defense.

For serious planning at the JTF level, ask, “who is the best in the world to solve this?” Since you’re already working remote, you should be able to engage them in a matter of minutes to hours. No matter which organization they work for.

Imagine that the person you most want for a particular ad hoc team works for the Defense Threat Reduction Agency (DTRA). You need her for a total of about six hours. Will her leadership have a problem with this? I doubt it. In exchange for six hours of work, DTRA proves its relevance to your major operation.

With a good rolodex and the right infrastructure for remote collaboration, your possible teammates now include the entire government.

4. Make it easy for subordinate units to contribute to plans

It’s already common to bring subordinate units into your OPTs. Now that remote collaboration technology has improved, you can take it a step further.

A Special Forces team leader is 5 echelons below the Joint Task Force or Subunified Command. I imagine this captain — and others similarly distant from the headquarters — contributing directly to a theater-level plan.

It’s hard to do this when your plan is made of Word docs and PowerPoint slides. It’s hard to distinguish high-quality contributions from the noise. Even harder for disparate contributions to add up to a coherent whole. This is one of the technology problems I’m working on right now.

Conclusion

Individually, these aren’t big changes. Each is already happening on a small scale in some units. Together, they add up to a Ghost Headquarters. It’s a less attractive target, because no single node is critical. Loss of comms at a single node is a nuisance, rather than a disaster. And ad hoc teams are constantly forming and disbanding, with precisely the right people for the task. It means well-informed command decisions, at the speed we need for great power competition.

I work on these problems every day at Onebrief — software for agile military planning. Here’s how you can get involved:

1. We run monthly exercises to test Onebrief against a red team. If you’re a military officer serious about planning, you’ll make $500 for participating in an exercise as a tester. Details here
2. We’re hiring! Our most critical open role is a senior product designer. The ideal candidate is experienced (UI, UX and data visualization) but hands-on.
   Location: Remote (US only) | Salary: $95k-$125k | Equity: 0.72%-1.44% | Job description