Nobody has any hands

The owner of a small company is usually also the dispatcher. And the person who does the invoicing. And the one who answers the phone at 6:40 in the morning when somebody calls out. And the hiring manager, and the safety officer, and the one who remembers that the truck with the bad alternator should not go north.

Every one of those is a real job. He is doing all of them, and he is doing them badly, not because he is bad at them but because eleven jobs do not fit in a day.

When people in my industry talk about this, they call it a technology gap. It is not. It is a hands problem. There are not enough of them, there is no budget to hire more, and every hour spent learning something new comes directly out of the hours that keep the lights on.

The thing almost everybody has already been through

Before I get to what I build, I have to say the part that gets skipped, because it explains why the first ten minutes of most of my conversations are spent on something other than the work.

Most of these owners have already bought something. Somebody came in, ran a very good demo on a clean dataset that was not theirs, everything worked, they signed, they got a set of logins and a PDF, and then that person was gone. What is left is a subscription nobody cancels, because cancelling it means saying out loud that it was a waste.

That is the thing I am actually competing with. Not another firm. A scar. And the objection I get is almost never "will this work." It is a quieter one, and it is the right question: am I going to end up holding something I do not understand.

Anyone who has been through that once is correct to be suspicious the second time. I would be.

Why generic software keeps failing a four-truck shop

Most business software is designed for a company that has somebody whose job is the software. An administrator. A person who configures it, maintains it, trains people on it, and calls support when it breaks.

A four-truck shop does not have that person. It has the owner, at 9:30 at night, after everyone else has gone home.

So the software gets bought and about eleven percent of it gets used. The setup that would make it worth the money requires an afternoon nobody has, and it never gets that afternoon, and the tool becomes another tab in the browser that means guilt.

That is not a failure of discipline on the owner's part. It is a product built for a company shape he does not have.

What a second brain actually does, stated plainly

The phrase has been worn out by note-taking apps, so here is the version I mean.

It takes one job off a person, permanently, in the shape the business already works. Not a platform. Not a category. One job.

A concrete one, from a field services company running several crews. Before: a printed schedule made the night before, a stack of phone calls at 6:40 in the morning when it changed, and a foreman working from a sheet that went out of date at 7:15. After: the schedule lives in one place, each crew sees only the day in front of them on the phone in their pocket, and a change made once is a change everywhere.

That is not impressive technology. It is a small amount of ordinary software pointed at the exact thing that was eating an hour every morning, and built around the way that company already runs rather than the way software usually wishes companies ran.

What it does not do is have the judgment. It does not decide which crew is short, or which customer gets moved, or who is having a bad week and should not be sent to the difficult property. That was always the owner, and the point of the whole exercise is that the judgment becomes the part he has time for.

Three questions before anything gets built

I ask these in the first conversation, and if I cannot get clean answers I say so rather than quoting.

What happens every week that everybody dreads. Not the biggest problem. The most repeated one. Repetition is what makes something worth building, and dread is a reliable signal that a process is fighting the people in it.

Who would run this. A name. If nobody has time to be that person, then building the thing makes their week worse, and the honest answer is to build something smaller or nothing at all.

What breaks if it is wrong. A misfiled note is a shrug. A crew sent to the wrong address at 7:00 in the morning is a customer. That answer sets how much checking gets built in, and it is the difference between something useful and something dangerous.

The handover is the product

Here is the part I would want to hear if I were the one who had been burned.

When the thing is built, I do not send logins. I sit with the person whose name came out of question two and we run it together on real work. Then I break it on purpose, in front of them. I put in the bad address. I make it produce the wrong answer, and we look at the wrong answer together, so they know what failure looks like before they meet it alone.

Then I leave, and it keeps working, and that is what finished means. Not the day it works. The day it works without me.

The engagement is over when they can run it and can say where it breaks. Not when the software is live. Those are different days, and the gap between them is where most of this industry loses people.

If you are the person with eleven jobs

You do not need a platform and you do not need a strategy. You need one of the eleven to stop being yours, and you need to be able to say what it does after the person who built it goes away.

That is a smaller ask than what is usually sold, and it is the one that actually gives somebody their morning back.