2.06: ‘The Ben and Erik Roadshow’

Night mortar fire. Author’s photo. Afghanistan, 2010.

There are a lot of expressions which you hear parroted in almost every unit in the Army, but no one really lives by. Take the adage, ‘Train to standard, not to time’. It’s very easy to understand and it briefs well but try taking less than an hour to complete your unit physical training and see what happens. We all know ‘The first report is always wrong’, but most commanders lose their temper when they get them. We also tell people ‘Never read PowerPoint slides to your audience’, and then sit quietly while someone does exactly that, never cutting them off.

The thing about each one of these is, just like other proverbs, they’re generally true just sadly ignored. One of the most important ones that suffers this plight is ‘Intel drives operations’. At almost every headquarters in the Army the S3 shop (Operations) is the biggest by far, and it tends to drive operations, intel be damned.

I moved up from my battalion XO job to take on the Group S3 role.¹ I couldn’t have timed it better, because I got to work alongside Ben, the Group S2 (Intelligence). Ben was a fellow advocate for adopting digital processes and leveraging data. We hit it off immediately.

Our first endeavor was to kill the command’s VIP PowerPoint brief. Most units have a standardized deck of slides (usually with too much text and often not up to date) which they use to brief visiting leaders, be they generals or congressional staffers. Ben hated PowerPoint as much as I did, and he also hated keeping the slides up to date. So instead, we got the team to help us make a KMZ of all the various topics, and we’d use Google Earth to brief off of.

The formula was pretty simple. We uploaded everything we could into it. Maps of key locations, flows of trade, key partners, and of course where our units were. By having it all in layers, we could turn things on and off as needed. In addition to helping avoid the clutter all too common in PowerPoint, this also meant we could bring multiple data layers up to highlight how they influenced one another.

Ben and I developed a rough outline of topics, but the key difference in the brief was giving control over to the guest. We’d pull up the map on the screen and ask, ‘Where do you want to go?’ As soon as they decided, or if they deferred to us, it was Ben’s turn to open. He’d introduce the topic and provide his intel perspective. He’d finish off with an open ended closer, inviting either a question or for me to tag in. I’d provide the ops” perspective, and then do my own en passant back, opening either for the guest or Ben to keep the conversation going.

The style was terrifying to most of the battalion commanders. They kept pinging us for the slides to the brief, and refused to accept there were none until we were at the table and no print outs emerged. ‘What are we going to talk about?’ they would frantically whisper to Ben and I. ‘Whatever the General wants to talk about, sir,’ I’d coyly reply, usually followed by Ben adding ‘It’ll be fun, sir, relax’.

And it was. Military and civilian both, our guests routinely cited the brief as one of the highlights of their trip. This was principally because it was a conversation, not a slow death by bad PowerPoint. The conversation could flow wherever their interests did, with new layers and data being brought up and hidden as needed to facilitate the conversation. The map at the front of the room stopped being a focus and became just a background to help frame the back and forth at the table. When the discussion inevitably ran long, time and again we found our guests asking for just a little more, ‘I want to keep this conversation going’. We even had generals who came by just to get the brief after one of their fellow flag officers had told them they needed to check it out.

The only challenge was you had to know both everything that was in the KMZ, and where it was. I devised a better way to do this years later. But at the time it meant the Assistant S3, my friend Nick had to doing some serious mental judo. He had to manage the challenges of keeping tabs on where the conversation was at the table while also recalling where in the data structure the right data was to get it onto the screen. It’s not unlike having a London cabbie learn ‘The Knowledge’, and I don’t doubt it altered Nick’s brain in much the same way. He never slipped though. He was so good at it in fact we had more than one General caught bewildered and thinking we had our own voice activated AI before we let them know it was just the captain in the corner moving the globe.²

The conversations were driven at their outset by Ben and his intel perspective. He got top billing in the ‘Ben and Erik Roadshow’ as it became called, because he drove the conversation. The briefing was so popular we were even called upon to give it from Fort Bliss, Texas to a congressional delegation via video teleconference (VTC).

Ben and I, along with a small contingent of the Group staff, were down in Texas for an exercise dubbed the Army Warfighter Assessment 17 in October of 2016.³ The exercise’s focus was on LSCO as opposed to COIN.⁴

Both styles of conflict have data requirements, but the volume of targets in LSCO is orders of magnitude higher. In COIN you can have days’ worth of footage on a single target to sort through as you struggle to find the needle in the haystack. In LSCO the targets are almost immediately recognizable. Tanks don’t blend it like insurgents do. The challenge instead is how many different targets you need to track.

LSCO is a different fight than COIN in a lot of ways, but one of the ways the Army is still struggling to adapt is to the fires focus. LSCO still has the need for maneuver forces to mass and take objectives, but with today’s sensor and communication capabilities, mass also gets you killed. It is a fires heavy fight, one which often subordinates maneuver to fires. The Army, which has fetishized maneuver forces for almost all of its 248 years, has struggled to make the turn.

Ben and I made it back in the fall of 2016, when we needed a way to track all the targets we were getting. Seeing them wasn’t enough, we needed to track them and ensure nothing was double counted, and also update them as they were actioned by the fires cell. And we had to do it at a scale we’d never encountered before. Both Ben’s intel analysts and my fires cell started making their own separate trackers on white boards, but it quickly became clear that wasn’t going to keep up with the volume.

So, we made a SharePoint List, which is really just a database. But instead of two trackers, we made it a single one. The intel side of the equation led, creating new entries for any targets that had not been previously identified, and updated their details as reports came in. Fires would take each target, figure out the right weapon, and pass the request up to the army division running the wargame, logging the target number, and what munitions the division used when. Intel could then update the effect on target as a new report came in, and either close the target or poke the fires side for a reattack.

No one emailed anything back and forth, everyone could see the status of all the targets from their own desk, and they could each make updates as necessary. But where people sit also matters. In our operations cell, the fires team sat directly across the table from the intel team. They were forced to look at each other every time they looked up from their computers, and so crosstalk and coordination was fast and flat. The whole system was improvised on the spot, made and modified by the team themselves. And it worked well, not just tracking the targets, but also giving us data on which sensors were producing better than others, and which munitions were having the right or wrong effect.

The lack of effect was what prompted me to walk over to the division headquarters on the third day of the wargame. The division had a HPTL and right at the top of that list was a Gravestone Radar, which was serving as the backbone of our enemy’s air defense system.⁵ For three days everything that flew over the line of troops between our forces and theirs was shot out of the sky almost immediately. The number of downed aircraft due to this radar system possibly exceeded the total of airframes that actually existed in the US inventory. My issue was three days ago one of our forward teams had found it, but it wasn’t dead yet.

In contrast to our operations center, the division’s was much larger, situated inside massive tents with air conditioners running full blast in the Texas sun to keep it to at least an uncomfortably warm temperature inside. I found their head of operations in the center, and asked him why the Gravestone wasn’t dead yet. He was confused, responding with a, ‘First heard,’ before beckoning the battle captain over to ask him for details.

We started by going to the fires cell, which was through a narrow passageway into another tent. They confirmed they’d received the request for fires from our fires cell, but they had been unable to confirm the target with their intel cell. So we walked back, through that narrow passage, across the OPCEN, down another narrow passage, and over to the intel cell where it appeared the request from the fires cell had never been picked up. If the OPCEN was the heart, the fires cell was in the left hand, and intel was in the right hand. It was literally a case of the left hand not knowing what the right hand was doing. Unsurprisingly, the principal means of communicating across the staff was in PowerPoint slides briefed at the nightly Commander’s Update. Our intel / fires database was live, always updating and always sharing everything with everyone who looked at it. The Divisions information only refreshed once a day at the CUB.

Our Gravestone had fallen off the radar⁶ because an intel rep had decided our report wasn’t credible, so it didn’t make the slide. In contrast, our targeting database tracked even false reports, updating the reliability of sensors, while avoiding the catastrophic error of just one person deciding a report wasn’t credible. Everyone could see everything, in real time. Ours was a system built on sharing, tracking, and executing data. It wasn’t constrained to nightly update meetings, and it wasn’t wasting its time making PowerPoint slides. Intel drove ops in both our organizations, but in the division, they drove them to a standstill. Data was the difference.

1. In Special Forces we have groups in lieu of brigades, but they’re roughly the same size.

2. Nick is currently working to bring machine learning LLMs to our DoD networks to automate and accelerate our systems. I have a sneaking suspicion that once being relegated to a human Siri may be some part of his motivation.

3. As mentioned previously, the Army’s fiscal year starts in October, not January, hence the name.

4. Large Scale Combat Operations, or what most lay people think of when they hear the word ‘war’. COunter-INsurgency, the fight the Army had been prioritizing for most of the previous decade and a half.

5. High Payoff Target List. All the things the enemy will have a very hard time living without. Preferably losing it will cause them to stop living entirely.

6