“I don’t manage anyone” is the most common thing solo founders say, and it is usually wrong. You manage you, which is hard enough. You probably also manage a bookkeeper, a freelance designer, a virtual assistant, and, these days, a handful of AI tools that will cheerfully do whatever you ask, including the wrong thing.
Management is not about having a corner office. It is the craft of getting reliable results from something that is not you. The old masters of the field, people like Peter Drucker, W. Edwards Deming and Andy Grove, were obsessed with exactly that. Their ideas were built for factories and corporations, but they port beautifully to a team of one. Here is what to steal.
Why the classics still apply
Contractors and AI assistants share two traits with every employee in history: they are capable, and they do not know what you know. They cannot read your mind about what “good” looks like, which client is touchy, or which corner you never cut. Left alone, they will do what is easiest, not what is right.
Classic management is a toolkit for closing that gap. The trick is to keep it light. You are not building a bureaucracy. You are writing down the minimum needed so you stop re-explaining yourself and stop being surprised.
1. Objectives before activities
Britannica credits Drucker with inventing management by objectives: agree on the results you want, then let people (or tools) work out how to get there. For a solo founder the translation is simple. Do not hand over a task. Hand over an outcome, the constraints and a definition of done.
“Write a newsletter” is a task. “Draft a 400-word newsletter that gets readers to book a call, in our usual voice, using the three points below, with no claims I cannot source” is an objective. Contractors and AI both do dramatically better with the second. If you struggle to write the brief, that is useful information: you have not decided what you want yet. A decent job description is the same exercise, aimed at a human.
2. Written standards (your one-page SOPs)
Michael Gerber’s The E-Myth Revisited argues that a business should deliver consistent results through ordinary, documented processes, not because one talented person (you) is hovering over everything. Our E-Myth Revisited review has the details.
For a team of one, “standard operating procedure” means a page, not a binder. For each recurring job, note what triggers it, what the steps are, what good looks like, and what to do when it goes wrong. Add a style guide for your voice, a list of banned phrases, and two or three finished examples. Those pages work as a brief for a new contractor and as the instructions you paste into an AI assistant.
3. Checklists for the things that can’t fail
Atul Gawande’s The Checklist Manifesto looks at how a 19-item surgical checklist, built around pause points before anesthesia, before the first incision and before the patient leaves the room, spread into hospitals worldwide. The World Health Organization says it has shown significant reductions in both illness and death among surgical patients. See our Checklist Manifesto review.
Your work is not surgery, but the lesson carries: experts skip steps, and not because they are careless. Routine breeds blind spots. Our review of the book notes that effective checklists are short, roughly five to nine items, and focused on the “killer items” that cause the real damage.
A good solo example is a pre-send check for anything a client sees:
- Names, dates, figures and links verified against the source.
- Matches the brief and the agreed scope.
- Read once on a phone screen.
- Anything AI-drafted has been read in full by a human, with every claim checked.
- Invoice or next step attached.
4. Reviews: close the loop
Deming is associated with the Plan-Do-Check-Act cycle, sometimes called the Deming or Shewhart cycle: plan a change, test it on a small scale, check the results, then act on what you learned. According to the American Society for Quality, you then repeat it continuously.
Solo founders are brilliant at Plan and Do and dreadful at Check. Put it on the calendar. A weekly 30-minute review is enough: what shipped, what broke, what did I fix by hand that a template or prompt could handle next time? When AI is producing volume, do not review everything forever. Sample it: check a few outputs each week, and update the instructions when you find the same mistake twice.
5. Match supervision to skill
In High Output Management, Andy Grove describes “task-relevant maturity”: how closely you should supervise depends on how skilled someone is at that specific task, not how senior or smart they are. He also champions “indicators,” early-warning measures placed along a process so you catch problems while they are small. Our High Output Management review explains both.
This is the best rule we know for delegating to AI and contractors. A new task, whether to a new freelancer or a new prompt, gets close supervision: you read every word. Once it has a track record, you move to spot-checks. When the stakes go up (a bigger client, a legal or financial claim), you move back to close supervision. Skill is per task. A tool that writes decent summaries may be hopeless at math. If you are not sure how comfortable you are letting go, try our delegation self-quiz.
6. Manage the scarce resource: you
Drucker’s The Effective Executive (1967) argues that effectiveness is a learnable habit built on a few practices: knowing where your time goes, focusing on contribution rather than effort, building on strengths, and doing first things first. Our Effective Executive review covers them.
The founder is the bottleneck of a one-person business, so the first employee you should manage is yourself. Track your time for two weeks. Mark what only you can do (sales conversations, final judgment, relationships) and what anyone, or anything, could do. Then move the second pile off your plate in order of pain.
What the “org chart” looks like
Here is a believable team of one, with a clear split of who does what:
- You: decide what the business does, own client relationships, set standards, approve anything that leaves the building.
- Contractors: specialist work where a person’s judgment and accountability matter (accounting, design, legal review). Hire quality; this is where cheap gets expensive.
- AI assistants: first drafts, research summaries, formatting, data cleanup, brainstorming. They produce, and you verify.
- Your documents: SOPs, checklists and templates, which are the closest thing you have to a manager.
Take a monthly client report. You decide what the client needs to hear. AI drafts the narrative from numbers you paste in. You check every figure against the source and run the pre-send list. A contractor handles the design template. One person, one afternoon, repeatable. Paul Jarvis makes a similar point in Company of One: automate repeatable work before you hire anyone, which our Company of One review summarizes.
Where old-school goes wrong
Do not import the bureaucracy. No weekly meeting with yourself, no 40-page manual nobody reads, no metrics you track because a book said so. If a standard does not save you time or prevent a mistake within a month, delete it. The goal of all this structure is fewer decisions at 4 p.m. on a Friday, not a more impressive-looking business.
The takeaway
Write the objective, write the standard, use a short checklist, review a sample, and supervise in proportion to proven skill. That one loop covers contractors, AI and, honestly, your own tired self. Start with the task you repeat most often and give it a one-page SOP this week. Want a head start on the rest of the setup? Grab the free “One-Person Business AI Stack” checklist.
Sources
- Britannica, “Peter F. Drucker” (management by objectives): britannica.com
- American Society for Quality, “Plan-Do-Check-Act (PDCA) Cycle”: asq.org
- World Health Organization, Safe Surgery tools and resources (Surgical Safety Checklist): who.int
- Book summaries and details (Gerber, Gawande, Grove, Drucker, Jarvis) checked against our reviews at The Growth Reads