A flat and accessible space to collaborate is a very cool thing to have, but it’s not much more than a dry erase board without data. Tragically, the army is in a perpetual state of data loss.
In just one special forces group there are 54 separate ODAs each executing over sixty to eighty JCETs a year, most of which have their own unique partner force. There’s more training than just JCETs. Teams execute off-post training events, subject matter expert exchanges, joint and partner force exercises, and combat training center rotations. Each team does multiple combinations of these in a single year. The tacit knowledge on a team is voluminous.1 But ODAs are always busy, so the next mission often means things don’t get written down from the last one. Anything not written down is on the verge of being lost or forgotten.
Layer in the PCS cycle and most organizations are bleeding irreplicable knowledge at a catastrophic pace.2 Some army units see between 1/3 and 2/3 of their personnel physically move to a new post every single year. Almost half the army seems to get a new job every year.
There are some software tools already in place to capture these soldier lessons as data, such as CIDNE, but they are too often clunky and too regimented.3 They also aren’t focused on getting information valuable to the teams, only answering higher’s RFIs. You can capture the obvious details of an exercise, but there’s no place to tell another team how to avoid getting a Roppongi shake down. There’s no block to explain how to navigate your weapons and ammo through three different countries’ customs. There is nowhere to capture the information a team really wants or needs to know to succeed when they RIP with you into the mission. Systems like CIDNE only capture the basics and that’s if the teams fill them in. Where there’s no return of value to the team, there’s less incentive to spend hours typing the data up. At best you get compliance.
There’s a sweet spot of Venn overlap between ease of input and ease of extract that defines how effective your data collection tool will be. In addition to making it as low cost as possible for your users to enter data, if you really want them to value input, you’ve got to make it easier for them to get data out. The chicken-egg dilemma requires most tools to do both well if they are going to see real adoption. However, it also means if you can find a tool that does both, it will see compounding returns. Good search doesn’t just make getting information out of a database easier. It improves the information in the dataset.
OneNote gave us an easy-to-use tool to easily capture soldiers’ experience and share it flat across the formation, but for me, its killer feature is unmatched search. This may come as a surprise given almost every other Microsoft Office search tool is horrible. SharePoint is famously bad at finding what you want, but OneNote is brilliant at it. Just hit Ctrl-E, and type in anything you’re looking for. OneNote will find it in the title, in the text, and — thanks to OCR — even in the pictures and print outs you embed in the notes.4 And it’ll find them fast. A good search function dramatically reduces the friction of building a new dataset, because you don’t need to worry as much about file structure.5
But, back to the chicken-egg problem, a good search tool is nothing without a large source of data to dive through. OneNote really shines in this space. Every note in the app has effectively infinite space, which allows users to store anything and everything they want.6 By using OneNote our staff officers were able to put any key updates at the very top of their staff estimates and then include any and all references or details just a short scroll beneath it. With easy to generate hyperlinks to different sections, notes, even individual paragraphs, moving through the entire notebook became easy.
Once we had the SOP up for the staff, the next project we turned to was capturing all the information we could from those 54 ODAs onto the ‘Country Pages’. We built separate OneNotes for all the countries we engaged with made draft templates to help the teams fill them out. The staff started by adding all the digital copies of training and JCET packets we already had on share drives, but included data about ranges, about partner forces, about everything we could find. We revived the area studies that had vanished somewhere in the GWOT. Then, to show off a little, we made an interactive map Ux on SharePoint, so all you needed to do was click on the country where you were looking to go.7
We also built out a one stop for all the OPT packets.8 These included everything from programs of instruction and risk assessments, to points of contact with local law enforcement and even past budgets. As with the country pages the goal was to get everything we knew into the OneNotes. We made a map interface for the whole project, so while you had the option to search by ‘mountain level 3’ or ‘scuba’, you could also just click on Oregon and get every single training packet we had for every spot in the state anyone had used.
Then, just like the SOP, we set it up so everyone in the entire SF group could edit it, all from their desk, with no need to ask permission or submit for review. I’m sure to some people that sounds scary, all those users with direct access to all that data. The concerns about data corruption weren’t unfounded. But we believed that the information we were losing was potentially more valuable than what we had, so we focused on removing every barrier to inputting it into the database.
Our commander was certainly concerned. When we briefed him on the country pages, we tried to explain to him that it worked just like Wikipedia. He looked at us skeptically and asked, ‘What happens when you get two 18Fs who are going back and forth and arguing, changing each other’s posts?’9 Before I could reply, the Group Command Warrant chimed in, ‘Then we’ve won, sir’.
Chief got it. He explained to the commander that if we had two soldiers arguing passionately about the finer details of an area study, then we had a level of buy-in that was self-sustaining. That’s part of what the super-powered search function was for OneNote. If you could effortlessly find what you were looking for, soldiers would use the database. And they would add to it. While I was never naive enough to believe we’d get every single soldier inputting everything they knew into the OneNotes, I knew we’d start to capture the gigabytes of tacit knowledge we had been losing every year. Knowledge that could never be replaced.
And the OneNotes were gigabytes. They continued to grow so large we ended up having to make separate files for each country. As each team uploaded their documents for mission approval, we added those, and when they returned, we encouraged them to update the OneNotes with the latest they’d learned. The OneNotes grew, tentatively at first, then faster. They kept growing as I PCSed on to my next assignment.
Until they didn’t. A new group XO came on board, one who didn’t know OneNote but was comfortable with PowerPoint. When it came time to migrate the unit’s SharePoint to the latest version, the decision was made to ditch the Country Pages and the OPT library. The files were ‘too large’, and KM was never a fan of them anyways. (Footnote: Knowledge Management. In DoD these are civilian contractors hired to improve our data systems. I have yet to meet a KM who actively did this though). The OneNotes were not in accordance with the SOP from higher after all.
It was my failure in the end. I was so focused on fixing the everyday data loss of soldier’s PCSing, I didn’t pay attention to the catastrophic loss of a single disinterested leader. I also should have done more to bring KM onboard. To show them when it comes to data, large file sizes can be a feature not a bug. I’ll write more on KM in 4.2, but after years of constantly having fights with KM, I did what most SF guys do. I went around them. I’d given up on KM instead of working with them to allay their worries, to show them the value of what we were building. As a result, we lost a library of knowledge, not to fire, but to ignorance.
Tacit knowledge can be hard to define, but it’s lessons really learned, usually because you learned them the hard way.
Permanent Change of Station, or when you move to a new base in the military. The army says the average servicemember moves every three years, but I’m averaging one every year and a half.
Combined Information Data Network Exchange. One of the DoD’s data repositories of tactical information.
Optical Character Recognition leverages ML to translate visual text into data.
I’ve been using OneNote since 2016 to organize my own personal records and often forget entirely where some of my commonly used notes are actually stored, since it was easier just to hit Ctrl-e and type in a keyword that it was to parse through folders and tabs.
OneNotes can scroll in any direction forever. In theory there is some sort of file constraint, but I’ve never found the bottom of one). Contrast this with PowerPoint where everything has to fit in that little box on the screen, which is why there are so many unbearable slides with tiny ass text no one can read.
There were columns on the left and right with country names and flags for the geographically challenged.
Off-Post Training. Much of the training SOF does is best executed off traditional military bases, requiring lots more paperwork to get approved.
Intel Sergeants, and historically the keepers of the area study.
I've had similar experiences with full time KM guys. The only successful attempts I've seen to manage knowledge were all grass roots efforts.
I have a hard time grasping why the KM org or G3 orgs don’t see the inherent value in sharing data rather than hoarding it. Imagine how many jr officers would appreciate shaving hours of workload off their plates they could use to spend time w their families. Imagine further, those same officers actually encouraging young men and women to join rather than feeling jaundiced.