Z.24: Confessions of a Firefighting Arsonist
My name is Erik and I’m a firefighting arsonist.
I presume most Downrange Data readers come to this Substack looking for answers, or at a minimum, novel ideas. I hope across the more than 98 posts many of you have found the insight you were looking for, or at a minimum a rhyming voice of frustration.1
Today’s not one of those posts where I know better. Today’s a post where I come clean about my own failings and ask for your help.
This all started last fall when my good friend and co-author Nicholas Frazier told me to check out his MIT professor’s book.
Nelson P. Repenning, Donald C. Kieffer There's Got to Be a Better Way
Right from the introduction, the book was calling me out, in detail.
The authors described how organizations call in their supposed best project managers to right a sinking ship, handing over the reins to people who have a knack for getting things done under pressure. Those who can circumvent processes and make magic happen. Except these firefighting arsonists are actually inhibiting the firm’s ability to grow and compete, because they aren’t fixing the process; they’re going around it. Too many of these sorts of pyros can cripple firms as they hire and promote the best firefighting arsonists over real change agents.
Anyone who’s served with me for a month has heard me decry the army for doing the exact same thing. I’ve described it as our tactical corgi problem: the military obsessively cultivates us to be tactical thinkers over our careers, meticulously turning out shortsighted tacticians who can solve immediate problems. Then suddenly at the Senior Service College they ask these stubby-legged fluffs to be strategic thinkers and find most of us can’t make the turn. After twenty years they want the long legs and foresight of Great Danes, but that’s not the traits they selected for.
Reppening and Kiefer are correct, of course. My firefighting skills are certainly a large part of how I made Colonel. Their book details how the real need is to attack the overall process that’s not performing. Anything less is at best a temporary fix, and at worst so damaging to the system itself that it has to be cut out like quick-clot. Sure, it stops the bleeding, but that’s only the solution in a crisis.
I can’t dispute Reppening’s and Kiefer’s critique. Any regular reader will know the myriad ways I bypassed the systems of record to get the job done. They worked in that moment, but I’m uncertain what happened after I left. This is a critical problem.
‘I think maintenance is the Achilles heel of soldier-developed projects,’ my fellow commander Zach Griffiths once reminded me. Zach’s not much of a firefighting arsonist, largely because he’s better at process building. He’s the reason the army has The Harding Project. Zach didn’t bypass the system like me. He went straight to the Chief of Staff of the Army and told him this was worth investing in.
Stormking Analytics has a recent piece where they give me a more flattering name than firefighting arsonist. To them I’m an operational leader.
Their piece details why the army has a dozen systems of record, all of which are dwarfed by literally thousands of Excel spreadsheets. I’ve had to listen to more than one KM lecture me that Excel is not a database. Well, in a world where you don’t have real-time access to a database, or where you need to connect multiple databases together to get an answer, Excel is the default tool.
Let’s start with asynchronous connections. My co-author had a great example of why the rest of the world forgets how inimical this is to military ops. He was meeting with Palantir to try and fix their system. He tried to explain the critical need to access at least some of the data while disconnected. When the Palantir rep asked, ‘Why would you disconnect from the internet?’, Nick realized he needed to start their entire conversation over again.
Excel isn’t just the real-time problem mitigation. It’s the go to solution for a lot of my problems. Stormking Analytics provides what might as well be athe most flattering definition of firefighting arsonist:
‘Excel is not chosen because it is elegant or ideal. It is chosen because it is available, flexible, and locally controllable. A spreadsheet can be built in minutes, adjusted in real time, and tailored to a specific problem without waiting for a development cycle or governance review.’
Excel, or more specifically, CSV is the lingua franca of the army’s true data corps. It’s the only way to get data to move and dance the way we need it to. Want proof? Download your data from any database, even Substack. Guess what? It’s coming in CSV.
I’m not a firefighting arsonist because I want to be, but because fixing the military systems isn’t within my control. Soldiers don’t have the ability to edit and update, so we become firefighting arsonists in the gap, and our CSVs / Excel spreadsheets proliferate. I’m not alone. Jennifer Pahlka’s organization Recoding America is trying hard to fix our bloated systems, but she time and again finds the same problems.2
As a garrison commander, I’ve been able to gain access to a lot of the databases that the army already has. But, none of them touch. I had hopes for the new Vantage AI and database system; but thus far I haven’t been able to self-learn my way through it. When our recent Senior Leader Forum introduced an FA59 who knew the program, I quickly asked him who our rep for Vantage was.3 Crickets.
Stormking Analytics recently posted an article listing organizations’ use of AI across a hierarchy.
Stormking described how most organizations are not using true AI at a tier three level.
My org is not currently using AI at scale at all. We are not even in first tier. We are at 0.
Which brings me back to Reppening’s and Kiefer’s book. A lot of different author’s fixes are focused on tech solutions. Mine certainly have been. But the book’s authors highlight the real problem is tracking the flows of information and putting the right people in the right place to make decisions. Reppening and Kieffer term this ‘making the invisible visible’.
David Epstein found somthing simialr when writing his most recent book.
Epstein cites a cutting-edge genomics lab who solved their organizational problem by actually drawing a funnel and mapping how everything fit into it. Once they could actually physically see the inputs, the answer to their challenges was obvious.
Which leaves me where I currently am. I know the army has databases of information which, if I link them up, can answer easy math problems. I need to get those systems to talk to each other without needing me to re-download the CSVs every time. But I also need to see the workloads people around me are actually carrying, so we can look at how to actually fix the broken systems around us instead of just sprinting around them.
If you have any Vantage experience or expertise in the latter hit me up in a DM. Because right now the only tools I have are CSVs and a lighter.
This post is #99.
Recoding America should be on every high school reading list. Recoding America > Once an Eagle. Come at me bro.
FA59s are Army Strategists.
We’ve been pulling Vantage into Power BI for our tools. Building Vantage dashboards is a completely other universe.
Thx for the Shout Out! FYI: We’re struggling with Vantage as well, and have also been using the same approach as @bryan.