There was a time when running a business alone meant pretending you didn’t. I remember setting up ten years ago, buying a London phone number and a virtual office address. It felt critical. I needed a trading name that sounded like a floor of people. “We” in every email.

That era is over. Well, for me anyway. I wear lean (read: not solo, we’ll come to this later) as a badge of honour now. I run several businesses (or income lines) with no employees, and it’s a conscious choice. Some might say a business of one is not a real business. Not a big enough swing. Too reliant on the individual. We’ll dig into this. What I can say is that every now and then I’ve been offered investment for projects and turned it down. Big swings are there to be taken, there is plenty of money out there looking for a home, it’s always tempting. But beware the hand that feeds you, and know the autonomy, potential, and time you’re trading.

Right now the whole world has opinions about these one-person armies. Sam Altman told an interviewer that his tech CEO group chat has “a betting pool for the first year that there is a one-person billion-dollar company”, a line that launched a thousand headlines about the age of the solo empire. The press wants you to believe this life is either a gold rush or a fantasy.

It’s neither. As always there is nuance to explore here, and I’m still attempting to unpick it myself. I should say up front that I’m not writing this from the summit. I’m partway on my journey, maybe halfway up the mountain, building this model in public where I can, comparing notes with peers doing the same. So no guru BS here, just sharing from my ledge - what the terrain actually looks like in my experience. And I’m not selling you a course. I don’t believe there is one that would work.

So let’s demystify it. The hype, the numbers, the people actually doing it, resources, and hopefully it gives you some ideas for your own scale attempt.

The hype has a body count

The one-person empire story followed a predictable arc.

February 2024: Fortune reports Altman’s betting pool. February 2025: Forbes runs “The Future Is Solo: AI Is Creating Billion-Dollar One-Person Companies”. June 2025: the story gets its proof point when Maor Shlomo, a genuinely solo founder, sells his six-month-old vibe-coding platform Base44 to Wix for $80 million in cash. Real company, real exit, one man.

Then April 2026: the New York Times profiles a telehealth business it describes as a one-man operation tracking to “$1.8 billion in revenue, built with $20,000 and AI tools”. Altman reportedly declared the group-chat bet won.

Within days, Techdirt took the story apart. The $1.8 billion was an unaudited projection, not a valuation. The company had received an FDA warning letter before publication. Its marketing used AI-generated doctors and manipulated before-and-after photos. Gary Marcus called it “a warning sign, not a poster child, for AI”. One prominent investor offered $100,000 against the valuation claim.

That is the state of the billion-dollar one-person company: still a WhatsApp group chat bet, now with verifiable hucksters pursuing the label.

But there are six-figure and seven-figure one-person businesses that are real, documented, growing, and far more useful to think about.

The real numbers

Looking beyond the headlines and focusing on the data.

There is a genuine boom. The US Census counts 117,060 businesses with no employees that cleared $1 million in revenue in 2023. That number roughly doubled in a single year, from around 57,000 in 2021 to almost 117,000 in 2022, the year ChatGPT launched. (Treat the timing as suggestive rather than proven; a jump that sharp may partly reflect how the data was gathered.) That number, 117k, is staggering to me. It excites me to think that this many people are out there killing it for themselves.

In the US, a record 5.6 million independent workers now earn over $100,000 a year, up 86% since 2020. Upwork’s research this month found 38% of skilled American knowledge workers now freelance, up from 28% a year ago. This is a tidal wave.

All very exciting.

But as you’d expect, most one-person businesses are modest. The average US non-employer business turns over $57,611 a year, less than the average wage. Much more sobering, and it makes you think: what is the point? May as well stick with a job, and all the benefits that accompany that - security (of a sort), lower cognitive load and stress, a kinder boss, and a steady pay cheque.

The million-dollar solo club is roughly 0.4% of 30 million solo businesses. And in the UK, businesses with no employees are not simply a trend. They are the default: 4.3 million of them, three quarters of all British businesses, and it has looked that way for a decade.

Read both truths together and you get the honest picture. This life is normal, growing, and genuinely accessible. It is also not a lottery win or a magic bullet.

So, the interesting question is not “can one person build a billion-dollar company”. A billion would be great, but I can’t imagine the personal life that sits alongside that. An outcome that exceptional means being an exceptional outlier, with all the trappings that come with it.

I’m much more interested in “how do ordinary, capable people build well-paid, autonomous one-person businesses”.

Nobody does this alone

Here is the thing missing from the headlines: almost nobody famous for running a one-person company is literally one person, on their own.

Justin Welsh, probably the best-known solopreneur in the world, has documented his wife in the business and a virtual assistant behind him.

Kieran Drew, the dentist who became a seven-figure writer, runs his whole operation with one VA at $1,300 a month.

James Schramko built his business on contractors, sixty of them at peak, before deliberately shrinking, and has a long-serving core team of exceptional people.

Noah Kagan wrote a book about starting a business in a weekend while running AppSumo with around a hundred staff.

Even Paul Jarvis, who literally wrote Company of One, built his software product with a co-founder.

This isn’t a debunking thread. It just means the definition everyone is using is wrong. This is why I land with lean, rather than solo. Solo is not an ideal to strive for, it’s smoke and mirrors.

A one-person company doesn’t mean one pair of hands. It means one decision-maker and no payroll. Or at least conditional or alternative payroll. Leverage everywhere: contractors for specialist work, partners for specific ventures, systems and software doing the production, and now AI absorbing a layer of work that used to require salaries. Jarvis had the definition right all along: “a company of one is simply a business that questions growth.”

That is how I run mine. I spent over a decade in recruitment building and managing teams, so this is not a theory of headcount avoidance from someone who never hired. Been there. I enjoy working with people. But with today’s tools, a payroll is not the requirement it used to be, and every employee you don’t need preserves something worth more than margin: optionality. I bring in specialists when the work demands it and partner where a venture needs more than me. What I don’t do is carry fixed costs that force the business to feed them.

Think of the explosion of freelancers. You can find any expert you need, whenever you need them, and usually on the terms that fit the moment.

Lean, not zero-human. Hold that distinction.

Who actually lives this way

The research for this piece surfaced dozens of documented one-person businesses. I’ve sorted them into four groups. Maybe you’ll see yourself, or your future self, in one of the groups.

The order matters, because the most visible group is the least representative.

The audience builders. Welsh has earned a reported $10 million cumulatively (self-reported, though unusually well evidenced) across five distinct lines: courses, consulting, sponsorships, subscriptions, community. Dan Koe runs a multi-seven-figure content business. Kieran Drew publishes his launch numbers down to the expense line. These businesses are real and their owners earned them.

Rightly or wrongly, I have wariness alongside the admiration. A striking share of solopreneur income comes from teaching people how to be solopreneurs, the same way people who get famous on LinkedIn almost always end up coaching LinkedIn. It happens time and time again. No sour grapes; it is real business. The problem comes when that experience gets packaged as lessons for people running completely different businesses, and when the content machine demands more and more material stretched further from what the teacher actually did. One gets to a point where credibility becomes manufactured and the current gig outstrips the track record that enabled the current gig.

If your window into this world is audience-builders, you are looking at the show home, not the housing stock. Be sure to keep those filters on when watching the show.

The invisible operators. This is the group that should give most people ideas, because nobody is selling you their course. Rich Rosen is a solo recruiter in Massachusetts who has billed over a million dollars a year for much of three decades; there is an invite-only society of around 75 top solo recruiters like him. Alan Weiss has run a solo consulting practice for forty years across six income lines. Sara Connell, a former advertising executive turned coach, has done over $1 million a year for four consecutive years, verified by Forbes, with zero employees and six contractors. A vending machine operator in Dallas built a five-figure monthly income from machines he bought with savings from a $9-an-hour bank job. A woodworker on Prince Edward Island earns a solid living from handmade shelving.

No audience, or at least not one as a primary revenue generator. No code, well, maybe some vibe code these days, but it’ll be far from their foundational income lines. Phones, judgement, craft, and persistence.

This is what the 117,060 mostly look like.

The fractional and freelance bench. The fastest-growing route for experienced professionals is selling leadership by the slice. Fractional CFOs and CMOs typically charge $8,000 to $20,000 a month per client in the US, £5,000 to £16,000 in the UK, with day rates of £800 to £1,500, usually holding two or three clients at once.

I work this way myself, as a fractional operating partner.

Honesty about the ladder, though. Nearly three quarters of fractional executives have fifteen or more years of experience, and over 90% of their work arrives by referral. You are renting out a reputation you already earned.

It is a second act, a career step. Not always an easy entry point. Let’s talk about that if you’re not sitting on a big CV or resume.

The accessible end of this group is under-reported and more interesting. Solo recruiting has almost no formal barrier: no licence, minimal capital, a phone and a niche. One placement at £80,000 on a 20% fee is £16,000; an employed UK recruitment consultant averages around £30,000 a year in base salary. The gap between those numbers is the price of staying employed in that industry. Commission-only sales agents and independent reps run the same trade: no credential required, a warm network and six months of runway instead. The pattern across the whole bench is that the pay ceiling and the barrier rise together, except in sales and recruiting, where the ceiling is high and the formal barrier is nearly zero. What filters people there is the unpaid ramp, not the credentials. The work is not decoupled from your time, so leverage is hard to come by. If you’re not on the dialling, you’re not making money.

One more distinction, because it defines who this article is for. An interim executive with one full-time client, or a contractor with one dominant customer, is an employee with extra admin and none of the protections. HMRC in the UK agrees; that is what the IR35 rules exist to say.

Autonomy starts when no single client can end your income. Employment, and one-client contracting, are the disqualifiers here, not because the money is bad but because your destiny is controlled by someone else.

The tech archetype. Pieter Levels runs a portfolio of products making roughly $3 million a year (self-reported), no employees, vanilla code on one server. He is not alone in the category. Marc Lou runs a similar portfolio of small products past $1 million a year, publishing his revenue on open dashboards, though note what his flagship actually is: ShipFast, a starter kit sold to people who want to become him. Even in tech, a chunk of solo revenue is the gold rush selling shovels. Jason Chin, better known as Easlo, cleared a reported $779,000 in 2024 selling Notion templates he curates rather than codes, having passed $500,000 before he turned 22. Peter Steinberger, an Austrian developer working alone, open-sourced an AI agent that collected 145,000 GitHub stars in 60 days and was acquired by OpenAI inside three months. And Shlomo’s Base44, the $80 million exit from earlier, was one person for most of its short life.

These stories are real. But often they underweight skills in audience building. Lots of these creators are top-class marketers alongside their tech skills. This combo of high proficiency in coding, distribution, or both makes this archetype one of the least repeatable examples on this list. To be admired rather than benchmarked.

Also, tech is a diminishing moat. Judgement and taste of course (i.e. good tech), but there is a jump in competition in the tech sphere due to AI. Distribution and GTM often outweigh the tech itself in the grand scheme of things. I often use bad tech and wonder how I’ve landed with it!

The earning web

Look at the durable examples across all four groups and the same architecture keeps appearing. Not one product. Not one client. A web of four to six income lines with one person at the centre.

Weiss built the pre-internet version: consulting, speaking, books, mentoring, workshops, licensed IP. Schramko runs the modern one: a mentoring practice plus revenue-share stakes in businesses he advises. Welsh’s five lines are the audience-builder edition of exactly the same shape.

Mine, for the record: fractional operating retainers, consulting projects including M&A and search work, AI transformation education and builds, 1:1 coaching, and a paid community. The shape is the point. The lines feed each other. The consulting sharpens what I coach. The builds generate the lessons behind the education and consulting on offer. The community elevates it all, and gives me peers on the same journey, those who value what I value.

This is why generalists thrive here, and why the one-person company is not really a technologist’s game. A specialist sells one thing into one market and lives with that concentration. A generalist sits at the centre of the web and routes between lines as demand shifts. Charles Handy described this in 1989 as the portfolio life, long before anyone said solopreneur. The idea is not new. The tooling is.

Two properties make a web worth building. Resilience: when no client is your landlord and no platform is your pension, a bad quarter in one line is an inconvenience, not an extinction event. I lost a £30k build pipeline Monday past, and surfaced a £100k consulting gig on Tuesday. In between I was relaxed enough, and had the time, to have a water fight in the garden with my kids.

This is resilience by design. And everything compounds: lines that share a centre feed each other in a way five separate jobs never could.

There is a third property I underrated for years: a web is held together by people. Over 90% of fractional work comes by referral. Every line in my own web traces back to a relationship, not a funnel. I keep trying funnels and they keep disappointing me. Peers matter most of all, the people a few steps ahead or beside you, comparing notes honestly because they are not selling you anything. Making time to pull others along that are a few steps behind you. If you take one tactical thing from this piece: your network is not a marketing channel, it is load-bearing infrastructure. Look after it.

An idea 35 years in the making

Worth knowing how deep the roots go, because “AI created the solo business” is simply not true.

Handy sketched the portfolio life in 1989. Daniel Pink counted 33 million American free agents in 2001. Tim Ferriss weaponised the idea in 2007 with The 4-Hour Workweek. Kevin Kelly supplied the demand-side maths in 2008: 1,000 true fans is a living. Taylor Pearson flipped the risk argument in 2015: “What was once risky is now safe. What was once safe is now risky.” Naval Ravikant named the mechanism in 2018: code and media are “permissionless leverage”. Jarvis gave it a manifesto in 2019, Sahil Lavingia gave it a Silicon Valley conversion story in 2021, and then the AI wave just brought a tech-bro betting pool into the limelight.

And let’s face it, this was circulated to drive their own tool usage up. The vision is romantic - AI will do it all for you. Beyond the hyperbole, I’ve not seen that to be true… yet.

Each generation added a layer. None of them needed AI to make the model work. Which brings us to AI’s actual place in it.

AI in its place

I went back to solo operating after my business hit difficulties in 2021. We had a post-pandemic spike in trade and I used the proceeds to invest in people. Those hires all failed and I nearly lost the business. I was forced back to solo operating, not by choice. That forced me to build automations and tech to survive. I automated credit control, got rid of bloated and expensive CRMs, rebuilding my own versions, and so on. Consequently, it led to me helping other businesses become more lean. This was pre gen-AI, all no-code, low-code, and self-educated. Trial by fire.

So these days my hats include educating business owners on the ROI of AI (done with you), sometimes building AI systems for businesses (done for you), and of course my own business life is AI-enabled. So this is not a sceptic talking. But the honest description of what AI does for a one-person company is narrower and more useful than the headlines.

AI absorbs the production layer. Drafting, research, formatting, triage, admin, first-pass analysis, code. The market is pricing this in real time: Upwork’s data shows freelancers doing AI-augmented work earn 34% more per hour, and earnings on complex judgement work rose 45% year on year, while pay for commodity AI-content production fell 13% per contract even as volume soared. Payment data from Ramp shows companies quietly swapping freelance production spend for AI tools at a fraction of the cost.

What AI does not absorb is the layer that makes the business exist: judgement, relationships, trust, selling, taste. The strongest evidence comes from the movement’s own poster child. Shlomo, the man who actually did the $80 million solo exit, says plainly that AI-built products have no moat, that competitors copied his features within weeks, and that he was setting alarms every two or three hours to babysit his servers. The enterprise data agrees: MIT found 95% of corporate AI pilots produce no measurable P&L impact, and Gartner expects over 40% of agentic AI projects to be cancelled by 2027.

So the operating rule I use, and teach: AI extends machines that already work. Levels bolted AI products onto ten years of distribution. Welsh added AI to a decade of audience. The production layer only compounds when the fundamentals underneath it, the offer, the niche, the trust, were built first, usually by a human, usually slowly. I have written before about why the “AI chief of staff” framing gets this backwards; the same logic applies to the whole business.

And sometimes the highest-leverage fix has no computer in it at all. One of the best workflow improvements I’ve made as a fractional operating partner was in a bakery, and the solution was a process change you could sketch on paper.

AI is a tool in the kit. The kit belongs to the operator.

The graduation question you’ll hopefully have to answer

There is a moment most successful one-person businesses hit, usually somewhere around the million mark, where the model strains. Jay Clouse crossed it and hired; so did several others in the research, and there is nothing wrong with that choice. Rosen and Weiss crossed the same line and stayed solo. Also a fine choice.

The failure mode is not growing headcount or staying solo. It is drifting into headcount because you never decided. Schramko’s compass is the best tool I know here: your effective hourly rate, total income divided by total hours including all the admin you pretend doesn’t count. Anything that drags it down gets systemised, delegated to a contractor, or deleted. Run that honestly and the decision about headcount mostly makes itself. Staying small is not a ceiling you accept. It is a choice you keep re-making, with a number attached.

The exit question

Here is the question almost nobody in the solopreneur world asks: what happens when you stop?

Jarvis answered it personally. The man who wrote the book sold his stake in his software company at the end of 2024 and retired. The philosophy’s own author cashed out, which tells you something important: a company of one can have an ending that pays.

But only some lines can. Working in M&A has trained me to look at every income line and ask which would survive my absence. Products, niche assets, standalone tools, a genuinely systematised service: sellable. A personal brand, a coaching practice, a newsletter that is unmistakably you: those die with you, or at least discount heavily. Both kinds of line are fine. Confusing them is not. If everything you build needs your face, you have built a well-paid job with no boss, which is a good life but a poor asset. The web should contain at least one line you could hand to a stranger.

The path to follow

Strip away the hype and the path into this life is unglamorous and repeatable.

This is the version I follow and the one I see working around me.

Sell a service first. It is the fastest line to cash and the best teacher of what the market wants, whether that is recruiting, consulting, a craft, or a fraction of your professional experience.

Add lines that compound what the service teaches you: a product built from the repeated work, a community around the people you serve, a rev-share where you would otherwise consult for free.

Build systems early so production stops needing you; that is where AI genuinely earns its place. Cap any single client at a level where losing them stings but cannot sink you. Find peers on the same path, because the web runs on relationships and honest comparison is worth more than any course. And decide what enough looks like, because a business that questions growth needs an answer ready.

I am still building this. That is rather the point: the one-person company is not a summit you reach, it is an operating model you run, and it gets better the longer you run it. It is also, on the evidence, more accessible than it has ever been, to generalists, to salespeople, to craftspeople, to professionals with a reputation worth renting, and not just to developers in the timeline.

This is the operator we are breeding in the Leanpreneur community, where a small but growing group of us build exactly this model side by side. It is also what I coach on a small number of engagements each year - I can take you from zero or low income to a fully sustained, leveraged business.

One question to leave with, the same one I keep asking myself: which line would you add to your web next, and which existing one is quietly holding you hostage? Stay flexible. Filter by effective hourly rate. Be ruthless.

Sources and further reading

Enjoyed this? Join the newsletter.

One email a week. What I'm building, learning, and what's actually working. No fluff.