It's Sunday night and you're already doing math on the week ahead. Not the fun math. The other kind, where you count the hours against the days and they don't fit, so you start deciding which evenings you'll quietly hand back to the business. You've done this so many times it barely registers as a decision anymore. And underneath the fatigue is a second feeling you don't say out loud, a low hum of resentment that the whole thing needs you this much, that it can't run a single day without you touching it.
That resentment comes from how much of the business still runs through you by default. Nobody designed it that way. It just happened, one unassigned decision at a time.
Here's how I've come to read a founder's calendar. Every recurring block of your time exists for one of two reasons. Either it's work only you can genuinely do, which is rare, or nobody has built the system that would let someone else do it, which is almost always the real answer. The hours are a readout. Block by block, they tell you exactly which parts of the business still have nothing underneath them.
You could work harder this week. Because, of course, you already know how, and so does every founder reading this, which is exactly why it won't fix anything. The useful question is much narrower and eveb more uncomfortable: which of your hours truly matter only because you happen to be the one who penciled them in?
Sort by dependency, not by category
Here's the framework, and you can run it today without buying anything or hiring anyone.
Pull up last week. Not next week. Last week, because last week actually happened and next week is still fiction. Go block by block. But don't sort the blocks the way you normally do, by category. Meetings, email, sales calls, that grouping tells you what kind of work it was and nothing at all about whether the work needed you.
Quick aside: if your calendar does not accurately reflect how you spent your time, then the first thing you need is a week of good data. As I think about my personal experience, there was a time where my calendar had little to no resemblance to where my time was actually spent. So much so that I had to start keeping a second "Time Spent" log which tracked where I actually spent my time.
Back to your calendar, or your Time Spent log. Sort it by dependency, not grouped by activity type. For each recurring block, ask one question: if I vanished for two weeks, no phone, no laptop, genuinely gone, does this stop? Does it continue, but lose most of its value?
If it stops , that block is a system you haven't built. Name it. The approval only you can give, the client who only trusts you, the number nobody else knows how to pull, the decision that routes to your desk because it always has. Each one is a specific, buildable thing that's currently disguised as "just part of the job."
Trust me, find the lowest leverage hour on that list and document it, then delegate it, automate it, or delete it. Recapture a little time each week until the list is empty. This is investing in yourself, your sanity, and uncoinceidentally, it's the best thing for your business!
If it stops running without you, that block is genuinely yours to keep: the handful of relationships that are actually yours, the few decisions only you can make, the direction of the whole thing. That work is real, but it's usually a much shorter list than most founders expect.
Run this honestly and the split tends to land somewhere ugly. For a lot of founders, 60 or 70 percent of the week turns out to be the first kind: work that only comes through them because the machine to route it elsewhere was never built. Nobody ever sat down and decided to make themselves indispensable there. They just never got around to building the lane that would let someone else carry it.
Two founders you already recognize
Picture a 30-person company where every refund over some dollar amount waits for the founder's yes. It started as prudence, back when a single bad refund could sting. Now it's just a plain old bottleneck. The founder approves those refunds because the policy was never written down. An unbuilt system sitting in plain sight, costing time week.
Or the founder who's the only one who can produce the monthly numbers, because the report lives half in a spreadsheet and half in their head. Every month, the same evening disappears into the same export. The report is simple. The problem is that it's undocumented, and undocumented work has exactly one possible operator.
Both are just systems nobody built, later relabeled in the founder's mind as something they're "good at" or "particular about."
Yes, it's slower. That's the whole reason it never happens.
You already have the objection ready. Building the system is slower than just doing the task. The refund takes you ninety seconds, and writing the policy that lets someone else own refunds takes an afternoon. That's true, and it's exactly why the task never gets systematized. The ninety seconds wins against the afternoon every single time, right up until you add up a year of ninety-second decisions. Even more, what work is not happening? What is the true opportunity cost of those 90 second decisions? I promise you, it is way more expensive than you realize.
And yes, someone will do it worse than you. At least at first. That's the price of the system existing at all. So you're choosing between a rough couple of weeks while someone learns and owning that task for as long as the company exists. Ninety seconds a day, forever, is the expensive option. Funny enough, after a little time, the system executor ususally does it better than you did, because it is core to their role, not something you're just doing because you need to.
How to run it this week
Block ninety minutes, print out last week, and mark every recurring block with a D or an M. D for "the business depends on me for this," M for "this is genuinely mine." Then take the D list and sort it by one thing only: which missing system, once built, hands back the most hours for the least effort. Build that one first. Not the scariest one, not the most interesting one, the one with the best ratio. Then the next one down. As I said above, if your calendar doesn't reflect reality, take the week to log your time.
If a full week feels like too much, don't do the full week. Mark up one day. Take Tuesday. You'll find two or three D-blocks that have no business being yours, and two or three is enough to start.
A calendar full of your own indispensability is a map of every system you haven't built yet. Your time is your most valuable resource it. Protect it with your life.
Keep building,
– JW
