Tuesday, September 29th, 2026

Thursday is October 1st.

That means Q4 starts in two days, and most of the operators I talk to are walking into it with a plan. Revenue targets. A launch. A hiring decision they've been sitting on since July. A list of projects with little colored dots next to them.

Good. You need a plan.

But I want to talk about the thing nobody puts in the plan, which is the person who has to execute it. You. The one making every pricing call, every hiring call, every "should we take this client" call, every "is this good enough to ship" call. The single asset every other asset in your business depends on.

And you're running it on default settings.

Not because you're lazy. Because nobody ever handed you the manual. You got the business, you got the clients, you got the inbox, and somewhere in there the assumption was that you'd just show up at roughly the same level every day and make roughly the same quality of decision at 7am and at 4:45pm on a Thursday after nine calls.

You don't. Nobody does. And the gap between your best hour and your worst hour is where a lot of Q4 is going to get quietly decided.

So today I want to give you the protocol I actually run. Not a morning routine. Not a biohacking stack. An operating manual for the operator, built for the next ninety three days, which is exactly how many are left in the year as of this morning.

Let's go.

Here's the thing I got wrong for about fifteen years.

I managed my calendar like every hour was the same hour. An hour is an hour. Block it, fill it, move on. Sales call at 3pm, proposal writing at 10am, pricing decision whenever it came up, inbox all day in the gaps.

What I didn't account for is that the version of me who shows up at 10am and the version who shows up at 3:30pm are not the same employee. One of them is sharp, patient, and willing to say no. The other one agrees to things to end the conversation.

I'd been letting the second guy make some of my most expensive decisions.

When I finally tracked it, and I'll show you how in a minute, I found that my genuinely good thinking lived in about a three hour window. Three hours. Everything outside that window was fine for execution, fine for calls where I already knew the answer, fine for admin. It was not fine for anything where the outcome depended on judgment.

And I'd been scheduling judgment work wherever it fit.

That's the first shift. Stop asking how many hours you have and start asking which hours you have.

This is the part everybody wants to skip, and it's the part everything else depends on.

For five working days, set a reminder every hour on the hour. When it goes off, write down two things. What you were doing, and a number from one to five for how sharp you feel. Not how busy. How sharp. Could you make a hard call right now and trust it?

That's it. Takes about twenty seconds an hour.

If you want the version where you don't have to rely on your own memory, Rize.io tracks where your time actually goes in the background, and it's brutally honest in a way that self-reporting never is. I run it because I lie to myself about my hours and it doesn't.

At the end of five days you'll have about forty data points. Look for three things.

Your peak window. The two or three consecutive hours where you're consistently a four or five. For most people it's somewhere in the first half of the day, but I've met plenty of operators whose peak is 8pm, and they've spent a decade forcing strategy into a morning slot that was never theirs.

Your trough. The stretch where you're consistently a one or two. For a lot of people it's mid afternoon. This is where you are most likely to say yes to something you'll regret.

Your drains. The specific activities that drop your number for the next hour or two. Not the hard ones, necessarily. For me it's a certain kind of meeting where nothing gets decided. I walk out a two every time.

Write all three down. That's your baseline, and it's the first page of your operating manual.

Now you rebuild the week around what you found, and the rule is simple.

Judgment goes in the peak. Pricing, hiring, strategy, writing anything that has to persuade someone, any decision that's expensive to reverse.

Execution goes in the middle. Calls where you already know the answer, delivery work you've done a hundred times, reviews.

Admin goes in the trough. Inbox, scheduling, invoicing, the stuff that needs doing but doesn't need you at your best.

That sounds obvious written out. Almost nobody does it. Look at your calendar for this week and I'd bet at least one decision that matters is sitting in your trough, because that's when the other person was free.

The move is to protect the peak like it's a client meeting. Because it's the most valuable meeting you have. Put a block on it every day for the next thirteen weeks, name it something boring so nobody asks, and do not let anyone book it. Including you.

I resisted this one for years because it sounded like wellness advice, and I was running a business, not a spa.

Then I read the actual research and changed my mind.

The short version is that the regularity of your sleep matters more than you think, and in some analyses more than the total hours. SUCCESS ran a solid piece on it in June that walks through the numbers, including a study that analyzed more than 61,000 adults and found that people with the most regular sleep schedules had somewhere between 20 and 48 percent lower mortality risk than the least regular. The researchers' conclusion was that sleep regularity was a stronger predictor of all cause mortality than sleep duration.

That's longevity data, not productivity data. But you don't need a study to know what an irregular week does to your Thursday. You've felt it.

So here's the protocol and it's the simplest one on this list. Pick a wake time. Hold it within about a forty five minute window every day, including weekends, for the next twenty one days.

Not a bedtime. A wake time. The bedtime tends to sort itself out once the wake time stops moving.

I held mine through a Saturday I really didn't want to. The next Wednesday was the first Wednesday in months where I didn't hit a wall at 2pm. One data point. I'll take it.

This is the one that's new, and it's the one I think most operators are getting wrong right now without knowing it.

Earlier this year, a team at BCG published research in Harvard Business Review on something they called AI brain fry. They surveyed 1,488 US workers and defined it as mental fatigue from excessive use or oversight of AI tools beyond your cognitive capacity. The people experiencing it reported a buzzing feeling, mental fog, trouble focusing, slower decisions, and headaches. And the organizational costs showed up where you'd expect. More errors, more decision fatigue, and higher intention to quit.

One user of an agent orchestration platform put it perfectly: there's really too much going on for you to reasonably comprehend.

I recognized myself in that immediately.

Here's what happened to me, and I suspect it's happening to you. AI made a lot of individual tasks faster. So I started running more of them at once. Three chats open, an automation building in another tab, a draft being critiqued in a fourth. It felt like leverage. By 3pm I was more tired than I used to be after a full day of writing everything myself, and my decisions in the afternoon were noticeably worse.

The work didn't disappear. It changed shape. I stopped doing the task and started supervising five of them, and supervision turns out to be expensive in a way nobody warned me about.

So I built rules. Here are mine, steal whatever fits.

One thread at a time on anything that matters. If I'm using AI to think through a pricing decision, that's the only thing open.

Batch the reviews. I don't check AI output as it arrives. I let it stack and review it in one sitting, in my middle hours, not my peak.

No new AI work after my trough starts. Anything I kick off late in the day, I review badly.

And I run everything through one place. I use Galaxy.ai because it puts the models I actually use under one roof, which means fewer tabs, fewer logins, and less of that scattered feeling of running a small air traffic control tower out of my browser.

The point isn't to use AI less. It's to stop treating your attention like it's free just because the tokens are cheap.

This one's short and it's saved me more money than anything else on the list.

Pick a time of day after which you do not make irreversible decisions. No signing, no hiring, no firing, no pricing changes, no "sure, let's do it" on a big commitment.

Mine is 4pm. After four, the answer to anything that can't be undone is "let me look at this tomorrow morning and get back to you." Nobody has ever been upset by that sentence. Not once.

Add a second trigger. If you slept badly, the cutoff moves to noon. Same rule, earlier.

This is the cheapest insurance in your business. The decision still gets made. It just gets made by the version of you that's actually qualified to make it.

Your peak window is worth nothing if you spend the first twenty minutes of it in your inbox.

I learned this the dumb way. I'd guard three hours on the calendar and then open email "just to check" at the start, and by the time I got to the real work the best part of the window was gone and my head was full of other people's priorities.

So the first ten minutes of the peak have one rule. No inputs. No inbox, no Slack, no phone, no news. You open the one thing you're working on and you start.

If you've got meetings eating your peak because notes and follow ups have to happen somewhere, get the note taking off your plate entirely. I run Fathom on every call so I'm not spending my good hours reconstructing what somebody said on a call I was half present for.

Every Friday, twenty minutes. Three questions.

Did I protect the peak? Count the days out of five.

Did I hold the wake time? Count the days out of seven.

Did I make any irreversible decision outside my cutoff? Yes or no, and if yes, what was it.

That's the scorecard. Notice what's not on it. Revenue, pipeline, output. Those matter and you're tracking them elsewhere. This scorecard tracks the inputs that produce the quality of the person making those decisions, and it's the only thing on your whole dashboard you fully control.

Thirteen Fridays between now and the end of the year. Thirteen scores. You'll know by the fourth one whether this is working.

Copying somebody else's protocol wholesale. My peak is not your peak. Run the audit first or you're optimizing a guess.

Adding eleven protocols at once. You'll hold all of them for four days and none of them by the second week. Start with two. The wake time and the peak block. Add the rest when those two are boring.

Treating a bad week as proof it doesn't work. You will miss days. Travel happens. Kids get sick. Clients have emergencies. The protocol is a default you return to, not a streak you have to protect.

Letting the protocol become the work. If you're spending more time tracking your energy than using it, you've built a hobby. The audit is five days. The review is twenty minutes. That's the whole overhead.

I built the template for all of this because I got tired of rebuilding it on calls.

It's called The Operator Protocol. Inside you'll find the five day energy audit grid with the hourly scoring already laid out, the peak, trough and drain map, fill in versions of all seven protocols with the blanks for your own numbers, my AI ceiling rules written out so you can edit them, a decision cutoff card you can screenshot and keep on your phone, and a thirteen week Friday scorecard that runs from October 2nd through Christmas.

No cost, no sequence waiting behind it, nothing to sit through. Duplicate it into your own workspace and use it.

Get The Operator Protocol here

If you run the audit and your peak turns out to be somewhere strange, hit reply and tell me. Those are my favorite emails. And if you go through it and realize the real problem is that the business can't survive you protecting three hours a day, that's not a calendar problem. That's a structure problem, and it's the exact conversation Pinnacle Masters exists for.

Don't build the whole manual. You won't, and you shouldn't.

Set the hourly reminder. Tomorrow morning. Five days, one to five, twenty seconds an hour.

That's the whole assignment.

By next Tuesday you'll know something about yourself that you've been guessing at for your entire career, which is when you're actually good. Most of the people I walk through this are surprised. A few are annoyed, because their peak has been going to their inbox for a decade.

Q4 is going to ask a lot of you. The least you can do is know which version of you is answering.

Thirteen weeks. Go find your hours.

One step, one day. Grace over guilt. — Dan Kaufman

Tools I use and trust. If you sign up through these links I may earn a small commission at no extra cost to you. I only recommend things I actually use.

Make.com, automate your workflows without writing code

Fathom.video, AI meeting notes that actually capture what matters

Galaxy.ai, all your AI tools in one place

Rize.io, time tracking that makes you honest about your hours

Beehiiv, where serious newsletter operators publish