You think you know your process. Then someone hands you a pen and a napkin and asks you to draw it for a stranger who won’t nod just to be polite. That’s the only test that counts.
I was at lunch with a technical client. Burger joint. Sports on every screen, a TV in every corner doing its own thing. The kind of place where you’re half-shouting and nobody minds.
This client has an engineering team, and that team tests things most people would consider miles outside the scope of a marketing project. They don’t want the deliverable. They want the process. How does this actually work. What’s the flow. What are the questions you ask, and when, and why those.
Most of those questions get answered midstream, after the contract, in the real discovery, when you’re already in it. But we were having the conversation early, at the table, with a western burger in front of me.
And one of the engineers leans in and charges me with one specific piece. He wants to know how we take technical language and figure out what the target market is actually looking for, and then arrange that information so it doesn’t feel schmarmy, and then broadcast it back out into the world. Partly standard marketing operating procedure. Partly target-market development. And partly: how the hell do you actually do this, man.
He called it a three-body problem. I agreed. It is.
So I did what my dad did a hundred times at a hundred tables. I reached for the napkin. Grabbed the pen.
And I started to draw. Here’s the process. Here’s the flow. Here are the questions we ask right here, and here, and the thing we’re listening for when we ask them. Out of my head and onto the napkin: the problem, the solution, the technical pieces that have to be in place so we’re actually meeting the client where they are, understanding what they need, and representing it back out to the world. The same machine that learns from the client while it sells for the client. A flywheel, drawn in pen, on a napkin, in a sports bar.
Ten minutes of back-and-forth. Mostly me, talking and drawing and unfolding the thing as it got bigger than the napkin wanted to be. A tear in the corner. A barbecue stain from the western burger that I’m pretty sure is now part of the diagram.
By the end, we had the whole thing.
Here’s what it actually taught me
It wasn’t about the marketing process. It was this: I didn’t fully know the process until I had to draw it for someone who would catch me if I faked it.
That’s the part nobody tells you. You can sit at your desk and write your process down and feel great about it, and be completely fooling yourself. Because when you’re alone, you autocomplete your own gaps. You write “and then we qualify the lead,” and your brain silently fills in everything that phrase means, because you already know it. The sentence looks finished. It isn’t. You just can’t see the hole, because you’re standing in it.
A napkin with a skeptical engineer across the table does not let you do that. He sees the gap before you do. He doesn’t nod when you reach the fuzzy part. He just waits. And that silence is the test. You either have the next line or you don’t.
Yeah. That’s the whole instrument.
Why the napkin beats the whiteboard
We don’t even use whiteboards much anymore. We’ve got docs, decks, a tool for everything. And every one of those tools gives you somewhere to hide.
The whiteboard in your own conference room is the worst offender, because the people watching are your people. They already know the folklore. When you hand-wave the messy step, they fill it in for you without even noticing, and you both walk out thinking the process is solid. It isn’t. It just got autocompleted by a room full of people who’ve absorbed the gaps by osmosis.
A doc lets you leave things fuzzy and call it done. Nobody’s standing there. You can write “handle the exception appropriately” and move on with your life.
The napkin is small. It’s public. It’s watched. There is nowhere to hide on a napkin in front of a stranger who won’t pretend to follow you. That constraint isn’t a bug. That constraint is the entire point. The napkin doesn’t test whether your process is good. It tests whether your process is real.
And here’s the thing the kitchen taught me. The thing the AI agent has been missing all year, the part most deployments leave completely blank, is Ingredient #2 of the Professional Recipe. The Context. Every rule that lives in your head and not on the page. Every Tuesday-restock convention, every never-route-to-Mrs.-Alvarez exception, every we just handle that the new hire can’t possibly know. That’s the ingredient. And the napkin is the cheapest tool ever invented for finding out how much of it is actually written down anywhere.
Spoiler: less than you think.
Try it this week
Here’s the move, and it costs you a burger.
Take one of your people to lunch. Your ops lead, your top tech, whoever owns a process that matters. How you qualify a lead. How you scope a job. How you decide what to build next. How you handle the call that goes sideways.
Ask the hard question. If you can’t think of one sharp enough, go ask AI to write you a brutal one. Give me the question a skeptical engineer would ask to find the hole in this process. Then hand your person the pen. Hand them the napkin. And say: run the process for me. Draw it.
Then watch.
If they draw it clean, if a skeptic could follow it without re-explanation, they know it. That process is real, and you can build on it. You can write it down, hand it to a new hire, hand it to an agent, and trust it to hold.
But if it comes out as “well, normally, but sometimes…” If they explain the same step three times. If they draw a box and then talk right past it. You didn’t find a communication problem. You found the exact place the process was never actually defined. The same place your new hire keeps getting stuck. The same place the AI agent keeps confidently doing the wrong thing. You found it in ten minutes, over lunch, for the price of a western burger.
And if they won’t even pick up the pen? That tells you something too. The people who actually know their work are usually a little proud to draw it. The ones who won’t try are protecting the fact that there’s nothing there to draw. You may be working with the wrong people.
The napkin doesn’t care how smart anyone is. It only asks one thing: has this work ever been defined well enough for someone to draw it for a stranger? If yes, you’re ready to build. If no, the AI isn’t your problem and the new hire isn’t your problem. The napkin is just the cheapest place to find that out.
The napkin doesn’t test whether your process is good. It tests whether your process is real.
Original framework. Distilled from client work.
~ source material · Original framework. Distilled from client work.
