Most people starting a one-person business begin by shopping for tools. I think that's the wrong first move because the software is the inexpensive part now. But, the system that is built around it is what can break.
The software is the small number
Justin McKelvey puts the software costs of a one-person business at $100 to $250 a month. He speculates that the AI costs are under $50 for most people. His model has five systems: content, sales follow-up, delivery, admin and books. And, an AI draft-and-approve layer.
Keep in mind that those figures are his own and the post is promotional. Read them as one operator's accounting, not a best practice for everyone.
Even so, the order of magnitude matters. If the tools cost less than a gym membership, tools are not what separates a business that runs from one that stalls.
Delivery is mostly documents
McKelvey describes delivery as "mostly documents, not software." He means a written scope you reuse, a kickoff checklist, and a standard handoff doc. For productized work it also means a fixed scope, a published price, and the same delivery every time.
None of that needs a subscription, but all of it needs to be written down.
That is why his warning makes sense to me. Automating an undocumented process just makes the mess faster. He suggests writing the workflow down and running it by hand before you automate it.
Where one person breaks
The failures he names are operational, not technical. When you leave, a business of one goes into a holding pattern. Drafts keep coming, but nothing goes out without an approver. Automate enough and you wake up to 40 drafts waiting on you, so the approval queue becomes the new bottleneck.
Then there is the quiet failure. When something breaks in a solo business, it just quietly stops working, and you find out weeks later.
In each of these cases the tool did its job. What was missing was a decision that was never written down.
More people are running this way
This is not a small group. JPMorgan Chase's June 2026 SMB Pulse reports that about 46% of Chase businesses are solopreneurs, and their number grew 44% from 2021 to 2026. Roughly 1 in 30 has crossed $1 million in annual revenue, and more than 1 in 10 million-dollar businesses are run by a single person.
Now, be careful with what that report is. It counts who exists and how much they earn. It does not cover AI adoption, and nothing in it describes how those operators run day to day. McKelvey describes how one person runs, and he offers no revenue outcome beyond his own account.
Put the two together and there is a gap. Plenty of solo operators are succeeding, but neither source I checked shows how any of them get through a week when they are away from the routine.
Run the test
Find the holes in your own calendar. Be unreachable for a week and do not prepare. Afterward, write down every point where work stopped, every approval that sat, and every automation that failed without telling you.
Each item on that list is a missing document. Write it, then decide whether a tool is worth buying or making.
What the test shows
Expect the first list to be longer than you want. Examples could be: a client reply that waited on your approval, an invoice nobody sent, or a draft that sat unchecked because no rule was created that said who could release it.
These are decisions you make without noticing because you are always there to make them. Once one is written down, a person or an agent can follow it. Until then, the business only works while you're in the room.
If you run the test, start with the first item on your list and write that one down this week. Then go from there.