The Work You Stop Doing

This is part 2 of 4 of The Growth Team of One. Part 1, The Job Nobody Wrote Down, wrote the job down and tagged every line. This part decides what to stop.

The first thing I cut as a growth team of one was the wrong thing. I had a list, finally, and the list was too long, so I did what the list seemed to be asking for. I dropped the note I sent when a trial started. It was small, it was mine to send or not, and nobody had ever asked me to send it. I kept the weekly report, because the report had a meeting on Monday and the note did not.

Nobody mentioned the report for the next month. Somebody mentioned the note in about a week: a trial that had started on a Tuesday and heard from us the following Wednesday, eight days later, by which point they had picked something else. I had cut the one line on the list a customer could feel and kept the one line nobody was reading.

I had sorted the list by how it looked. That turned out to be the wrong column.

How work looks, and what it does

Last week I asked you to write the job down by recurrence and tag every line: only you, could run on a rule, or nobody owns it. If you did, you are looking at a long list and feeling the pull to shorten it. Good. That pull is where the mistake happens.

Work has a look, and the look arrives before you know what the line does.

Optional-looking work is small. It is triggered by something rather than scheduled. Nobody asked for it, so nobody will ask where it went. The follow-up when a trial starts. The CRM update when a deal closes. The note when a known account comes back to pricing. Each takes a few minutes, each happens when something else happens, and each is yours to skip without a conversation.

Mandatory-looking work is scheduled. It has a template, so it has a shape before you start. It has a meeting, so somebody appears to be waiting for it. The weekly report. The dashboard check. The board slide, formatted three ways because three people like it three ways. Each is on the calendar, and the calendar does not take requests.

Those are the parts. Now set the look aside and ask what each line does. The optional-looking line is triggered by a buyer, and the value of responding falls by the hour. The mandatory-looking line is triggered by the calendar, and the calendar does not care whether anyone read it.

So the look is a disguise, and it disguises in both directions. The line that looks skippable is the line a customer feels. The line that looks load-bearing is often the line nobody would miss. Flip each one and see which it is.

The notice test

Here is the question that replaces the look. Who notices if this stops, and when.

The names: nobody, you, the founder, a customer. The clocks: by lunch, this week, this month, never. Every line on your list lands in one cell of that grid, and most of them you can place in a few seconds, because you already know. You have just never written it next to the line.

Two corners matter.

The first is nobody, this month or never. That corner is the stop list. If nobody would notice for a month, the work is not holding anything up. It is holding a slot on your calendar.

The second is a customer, by lunch. That corner is the work you never stop, however small it looks or however little anyone asked for it. If a customer feels the gap before the end of the day, the line is doing real work at exactly the moment the work is worth the most.

The weekly report is the test case. Not who would say they read it. Who would notice, and when. Skip a week and count who asks. The honest count is small, and it often includes nobody, including the person it was for, who gets the number in standup and never opens the note. That is not a criticism of the person. It is a description of the report.

The rest of the grid is useful too. A line the founder notices this week is a conversation to have with the founder. A line you notice this month can run on a slower clock. But the decisions live in the corners, so place every line and read the corners first.

The stop list

What ends up in the stop corner is mostly mandatory-looking work with no reader. The check nobody acts on. The note nobody opens. The review of the test that has been "next week" since spring. The reconciliation of two numbers nobody disputed.

Stop it this week. Not "phase it out." Stop. A month from now you will have talked yourself back into half of it, and the only evidence you had, that nobody noticed, will have been replaced by the memory of how important it felt.

The way to stop it without a fight is to say two things out loud, to one person: what you stopped, and who would have noticed. "I stopped sending the Tuesday numbers note. The number is in standup. Nobody would have noticed for a month, so I am taking the hour back." That sentence makes the decision visible, which is the on-the-record version of quitting something, and it invites the one correction that matters. If somebody says "I read that," you have found a reader, and the line goes back on the list with a name next to it.

Chris Hart, Civic's CEO, asked in his piece on planning 2027 with the team you already have which repetitive work the team should stop doing by hand. The stop list is the first honest answer, and it is shorter than you expect, because most of the repetitive work is not in this corner. It is in the other one.

Build the list below. As you tick lines, the chart from the top of this piece loses a box. The chart with everything lit was part 1. The chart with boxes struck is this part.

What you do not stop

Now the other corner. A customer, by lunch.

Look at what is in it. The follow-up when a trial starts. The reply to the inbound form. The note when a known account comes back to pricing. Finding out that an integration broke before the customer does. Every one is small, triggered, and unasked for. Every one is a line a customer would feel before the end of the day.

Every one is also a signal with nobody on it. A signal is something a buyer just did, like coming back to your pricing page, and the work in this corner is the response to one. That is why it is time-sensitive, and why it looks optional: the trigger came from outside your calendar, so your calendar has no slot for it. I wrote about one of these lines in Follow-Up Has No Owner. The signup lands, goes on a mental list titled "maybe today," and is on no list at all three days later. That was the line I cut first, years before I had a name for it.

You do not cut these. You also do not keep doing them by hand, because by hand is how they came to look optional. You hand them to a rule: when this happens, do this, from this owner, and write down that it ran. A rule has no calendar, so it has no slot to protect, and it does not get busy. That is the whole of part 3, next week, and this corner is its input. If a line here was also tagged "nobody owns it" last week, it goes to the top.

Chris made a neighboring point yesterday about letting an agent run the setup of your tools and keeping the go-live decision for yourself. Who notices if it stops is a cousin of that question. Monday's piece put it another way: the tool that holds all the context is still not the one that decides, the context layer is not the decision. The stop list is a decision. So is the never-stop list. Nobody else can make either one for you.

What the record already shows

Bryn is the product we make for this, and we ran it on ourselves before anyone else, and we still do. If any of this already runs for you as a Play, a rule you wrote once that says when this happens, do this, then the notice test is already answered, and the answer is in the log. The record Bryn keeps tells you who noticed: every run, what triggered it, where it went, and whether a person approved it first or the rule carried it on its own. Approve mode, where Bryn drafts the step and waits for your yes, is where the never-stop lines usually start. Bryn is not another dashboard to watch. It is the governed execution layer that runs Plays through your stack. In plain terms: you write the rule once, Bryn carries it out inside the tools you already use, and nothing runs without a record of who or what approved it. That record is the one list on which nothing is disguised. A line either ran or it did not. The log will not tell you what to stop. That is still yours.

The gift

Do this Monday. Take the list from last week, or the sixteen defaults above if you did not make one.

For each line, write two words. Who notices if it stops: nobody, you, the founder, or a customer. When: by lunch, this week, this month, or never.

Cross out everything in the "nobody, this month or never" corner. That is the stop list. Stop it this week, and tell one person what you stopped and who would have noticed.

Circle everything in the "a customer, by lunch" corner. Nothing in that circle gets cut, however small it looks. If a line in the circle is also tagged "nobody owns it," it goes to the top of next week's list.

Then look at what is left in the middle. It can wait. The corners cannot.

Where this goes

Part 3, The Work That Runs Without You: the never-stop corner, line by line, and what it takes to hand each one to a rule you would trust. Which lines run on their own, which sit in Approve mode, and which stay yours.

Part 4, The Stack You Hand Over: the day hire number two, or your successor, inherits the stack, and the record becomes the onboarding doc.

Part 2 of 4 of The Growth Team of One.