Hey, Will here,
I grabbed lunch with this guy I know who got out. Not "sold the business and retired to Bali," got out. He still owns it. The whole thing. But he's out. Like, doesn't take calls during dinner.
Took 30 days completely off the grid last September. And the business? Revenue actually went up while he was gone.
So obviously, I'm sitting there thinking he probably hired some expensive GM, built out this whole leadership team, spent months documenting every process. Nope.
Here's what he told me: "I sat down on a Tuesday in February with the eight decisions I made every week. Didn't get up until I'd turned each one into a prompt. Now my team just runs the prompts. Took me six weeks. Didn't cost me anything."
That's it. Oh, and his phone stayed face down on the table the entire lunch. Didn't even glance at it once.
He didn't hire his way out of the day-to-day. He changed his thinking and found a better way.
And look, I'm not saying this to make you feel bad. But here's what I noticed: You're probably working harder than you were a year ago. The business? It's working about the same. He flipped that a year ago.
While you're grinding through another week answering the same questions, making the same decisions, putting out the same fires... he's at lunch with his phone face down.
I'm just saying, maybe there's another way to do this thing.
📋 The 60-Second Read
Three things before your coffee goes cold:
The pattern. Most founder-dependence advice dates back to before AI. Hire executives. Build management depth. Document SOPs. Six years and a payroll you can't carry.
The principle. The 12% of operators making AI work aren't using better tools. They've encoded their judgment. The other 88% are asking employees to use AI when it helps.
The prompt. The Encoding Test. A meta-prompt that takes one decision you make every week and turns it into a template your team runs without finding you.
Even if you close this right now, that's what matters this week.
🔍 The Dig
What my friend figured out has a name. The Encoding Gap.
Most operators see the problem and reach for a 1990s answer. Hire executives. Build management depth. Document SOPs. The advice is correct. It's also from a world without AI.
That world had one path: put another human in your seat. Five to ten years. Big payroll. Most operators never finished. 2026 has another path. Encode your judgment into a prompt before you encode it into a person.
Two studies, late 2025. PwC: 12% of CEOs got revenue gains and cost savings from AI. 56% got nothing. MIT: 95% of AI pilots delivered no measurable ROI. My read: the 12% don't have better tools. They've turned recurring decisions into systems their team can run.
That's the gap. Not between operators and AI tools. Between operators who encoded and operators who didn't.
6,240 hours × $200 = $1.25M. Twelve hours a week that shouldn't route through you, fifty-two weeks, ten years. Paid in time before a buyer ever shows up.
The exit discount on a $1M EBITDA shop runs $1M to $1.5M. That's the smaller figure. The bigger one is paid weekly. You don't have to wait for an exit to start saving it.
Three types of operators. You're one of them.
🪨 THE LINCHPIN. Phone face-up. Zero prompts. Treats every decision as proof of leadership.
🛠 THE FIELD GENERAL. Delegated the calls. Team calls anyway. Same hour count. Different excuse.
⚡ THE DIALED-IN OPERATOR. Eight decisions, eight prompts. Phone goes silent at 6 p.m. Friday and stays silent. Took 30 days unplugged last year. Revenue went up.
✅ Linchpins build paychecks. Field Generals build revenue. Dialed-In Operators build hours that compound back into a life.
💬 The Quote
Your judgment isn't sacred. It's a template not yet built.
🎬 The Move
This week, one thing.
Pick the decision that keeps landing on your desk. The one that interrupts your Tuesday morning. Pricing call. Customer escalation. Proposal review. Hiring call. Whichever one your team always pulls you into because "only you can make this call."
Don't hand it off. Encode it.
Write the prompt that captures how you actually make the call. The inputs you need. The criteria you weigh. The edge cases that matter. The output format you want. Five lines if it's simple. Ten if it's complicated.
One decision. One prompt. Before Friday.
The first one takes 20 minutes. The second one takes 10. By the fifth one, you're down to 4 minutes. And your team no longer needs you to make decisions you've already made a hundred times.
🎯 The Prompt
This week's Move handles one decision. The Encoding Test handles the rest.
⏱ 90 seconds + a 5-minute interview · 📋 You need: a decision you make weekly · ✅ You get: a copy-paste template your team can run.
Tomorrow I'm going to make a decision I make at least once a week.
The decision is: [describe the kind of call that keeps landing
on your desk — pricing, customer escalation, proposal sign-off,
vendor approval, hiring veto, etc.]
You're a fractional COO who's helped operators write prompts that
capture their own decision-making. Walk me through 5–7 questions
that extract:
- The inputs (what info is on my desk before I decide)
- The criteria (what makes a yes vs. a no)
- The edge cases (when I override the rules)
- The output (what my team actually needs from the answer)
After I answer, write the prompt as a copy-paste template. Make
it specific enough that two different team members would produce
the same answer if they ran it.
Don't be generic. Don't be helpful. Be precise. Ask clarifying follow up questions,Ninety seconds in. A working template out.
The first one feels uncomfortable. That's the point — the reason your judgment feels uncodable is the same reason it's still on your desk every week.
Run it before your next pricing call.
🎤 Open Mic
Three questions, pick one. Reply.
What's the first decision you'd systematize? One line.
Have you tried writing a prompt for one of your judgment calls? What broke?
Which decision do you genuinely think can't be encoded? Anonymous if you want it.
I read every reply. The ones that land in next week's Open Mic help other operators see themselves.
— Will
P.S. The Encoding Test is one of forty prompts in The Operator's Prompt Pack. You've seen four. The other 36 are what the 12% are using, while the 88% are still stuck in meetings. Five more in the Strategy section alone for decisions you're making right now that shouldn't route through you anymore. The gap widens every week you wait. Get the pack here!
What do you think of this newsletter?
Love reading newsletters? Here’s the best app for newsletter reading.
This newsletter may include paid promotions or affiliate links. We may earn compensation if you sign up through them. See our Privacy Policy here.
