100 One-Person Companies: A 2026 Study of Solo Businesses, Ranked by Leverage and Revenue

Introduction: When "Scale" No Longer Means "Headcount"

For the better part of a century, the business world ran on a single equation: to grow bigger, hire more people. Revenue's ceiling was set by the size of the organization, the size of the organization by the number of heads, and so "success" came to mean, almost by default, a larger office, more desks, and deeper layers of reporting. This book is about how that equation is being dismantled, piece by piece. Cloud services turned servers into a monthly bill; payment gateways turned global revenue collection into a few lines of code; distribution platforms compressed "building a sales channel" into a single act of listing a product; and AI turned the design, writing, editing, and customer support that once required a team into leverage a single person can summon on demand. Once the infrastructure shoulders the heaviest part of "scale" for you, revenue is, for the first time and at scale, decoupled from headcount. And so the company of one shifts from a fringe, faintly romantic choice into one of the mainstream paths.

1. What Is a "Company of One"

The clearest articulation of the concept comes from Paul Jarvis. In his telling, a company of one is not "a startup shrunk to its smallest possible size" but rather a business form that deliberately questions growth: it treats "should we get bigger?" as a question that demands justification rather than a default goal; it expands only when getting bigger genuinely makes the product better and life better, and otherwise holds firm at the "just right" size, trading the surplus energy for freedom, profit, and resilience.

One clarification: the "one" this book takes up is a spectrum, not a dogma. At one end sits the pure lone wolf: AJ of Carrd built a website-building tool to roughly $1.5M ARR single-handedly; Mike Perham of Sidekiq built a business on one Ruby background-job library to around $7M, and says of himself that he is "closer to $10 million than to $1 million." At the other end are the ultra-small "1 + outsourcing/AI" teams: Tony Dinh of TypingMind reached roughly $137K/month with a portfolio approach, and Pieter Levels (levelsio) single-handedly runs a string of products including PhotoAI, Nomad List, and RemoteOK, with PhotoAI alone at around $1.6M+ ARR. What they share is not "literally only one person" but rather decision-making power and value creation concentrated heavily in the individual, with an organization that does not scale by piling on people.

2. Why Now

The company of one is not a new term, but "now" is the moment it truly comes into its own, propelled by two overlapping waves of infrastructure dividend.

The first wave was the SaaS and platform infrastructure of the 2010s. Stripe globalized payment collection, cloud providers made compute pay-as-you-go, and platforms like Gumroad, Notion, Etsy, Substack, and beehiiv standardized "open a store, charge money, distribute." And so we saw: Nathan Barry reached $145,471 self-publishing an e-book in 2012; Easlo built Notion templates to roughly $779K (2024); and Thomas Frank crossed $1,000,508 in a single year from Notion templates alone. The more mature the infrastructure, the larger the business radius one person can carry.

The second wave is the AI leverage of the 2020s, which directly cracks the hardest bottleneck a lone wolf faces: the capacity to produce and deliver content. Danny Postma's HeadshotPro, generating professional headshots with AI, reached roughly $3.6M ARR; Mathis Lichtenberger's ChatPDF turned "chat with your documents" into a business of about $440K ARR; France's Tibo reached roughly $1M/month with an AI product portfolio including Revid and Outrank; and the most extreme signal of all is Israel's Maor Shlomo, whose AI website-building product Base44 was acquired by Wix for $80M in cash in 2025 — one person, one blockbuster exit.

3. What This Book Draws On

This book gathers 100 cases spanning the globe and the full spectrum, deliberately avoiding the single-flavor "Silicon Valley unicorn" narrative.

• Geographically, it spans developed and emerging markets: the United States makes up the bulk with 57 cases, but there are also 8 from the United Kingdom, 6 from China, 5 from the Netherlands, and 5 from Canada, along with Vietnam (Tony Dinh), Israel, India (Louis Pereira of AudioPen), Georgia, and founders of Ugandan descent.
• By category, it straddles the "soft" and the "hard": 20 SaaS, 16 content, 16 knowledge, 12 e-commerce, 8 AI, 8 physical, 8 services, 6 assets, and 6 China-based. There are levelsio-style pure digital products, but also physical businesses carrying inventory and logistics, like Tabs Chocolate (roughly $11M in about 18-24 months) and Hill Vending (around $58K/month).
• By magnitude, it covers the full spectrum: from the top tier of Base44 ($80M exit) and Stratechery (around $5M+), to the million-dollar core cohort (40 cases scoring 7-8), to the six-figure middle (44 cases scoring 5-6), down to early-stage samples like Firebean Coffee (around $90K/year). What the reader sees is not a curated set of survivors but a real path that can be climbed step by step.

Each case carries, as far as possible, a verifiable revenue basis and source (founder's own account, GetLatka/Starter Story estimates, CNBC/NYT reporting, and so on), with its nature noted. Treat the figures in this book as orders of magnitude, not precise financial statements: many are founder-disclosed or third-party estimates, and where the sources disagree the book flags it within the entry (for example, ChatPDF's two figures of $440K and $6M).

4. Three Ways to Read

To suit different purposes, the book offers three non-conflicting points of entry.

The four-dimension scores (revenue, replicability, leverage, timing) are not there to rank people by rank, but to help you judge whether a case is worth learning from and whether it is learnable at all. For instance, 25 cases score ≥7 on replicability, and these tend to be the more beginner-friendly starting points; high revenue, meanwhile, often comes paired with low replicability — seeing this trade-off clearly matters far more than memorizing the rankings.

5. A Word to the Reader

This book does not promise that "one person can make ten million dollars," but with 100 named, documented samples it tells you this: today, "small" is no longer a synonym for "can't make money." It can be a carefully chosen, sustainable business form — and what you have to do is find the "just right" that belongs to you.

Methodology: How We Chose the 100 Cases, and How We Scored Them

The 100 samples in this book are not ranked by fame, nor by how "inspirational" they sound. They are the residue of a single, consistent screening-and-scoring pipeline. This chapter lays that pipeline out in full—the inclusion criteria, the four-dimension Composite Inspiration Index, the logic behind balancing categories, and the data sources and their limits. With it, readers can judge for themselves which conclusions are trustworthy, how trustworthy they are, and where the figures should be treated as nothing more than reference.

I. Inclusion Criteria: Three Hard Gates

To make the cut into these 100, a case had to clear all three of the following at once:

• Led by one person or a very small team. The core value is created and controlled by 1 person (or a tight partnership of 2–3). Designjoy is the purest sample—Brett Williams reaching $3.1M in annual revenue single-handedly (2024, GetLatka). Tabs Chocolate, though co-founded by Oliver Brocato and Jake Lewin, still falls within the very-small-team range, so it is retained with a note. Companies that hire large teams and win on organizational scale are excluded.
• Real commercial output. There must be money, paying users, or a verifiable exit. Base44 was acquired by Wix for $80M in cash (June 2025); TalkNotes closed on Acquire.com for $200,000 all-cash (August 2024)—hard transactions like these are the strongest evidence. Sustained revenue ranks next, such as levelsio's PhotoAI at roughly $1.6M+ ARR (2025, disclosed publicly by the founder). Raw traffic or follower counts are not grounds for inclusion.
• Publicly verifiable data preferred. All else equal, samples backed by CNBC, the NYT, or a founder's public dashboard take priority. ecommemily's Etsy shop carries a CNBC-vetted back-end figure (roughly $220,300, 2024); Thomas Frank's Notion templates have a figure precise to the dollar—$1,000,508 (full-year 2022). These "verifiable" cases form the backbone of the list.

II. The Composite Inspiration Index: A Four-Dimension Scoring System

Each case is scored 0–10 on four dimensions, then weighted into a "Composite Inspiration Index" (out of 100). The weightings embody this book's stance: it isn't only about how much money was made—it's about whether anyone else can walk the same road a second time.

This weighting explains a good deal of the ranking differences. Designjoy tops the list at 87.0, on the strength of all three running high: revenue (about $4M cumulative), replicability (productized design subscriptions are an open template), and high leverage (one person, no outsourcing). At the other end, No. 100 Firebean Coffee Roasters (about $90K/year, 2021)—coffee roasting is asset-heavy and labor-heavy, scoring low on both leverage and replicability—lands at just 34.0. Across the full sample, the four-dimension averages are: revenue 6.3, replicability 5.4, leverage 8.1, timeliness 7.5—leverage is the highest (49 cases score ≥9 on leverage), confirming that the essence of the "company of one" is a leverage game.

III. Balancing Categories: Nine Categories and Their Quotas

To keep the list from being monopolized by any single hot lane (AI and SaaS above all), we applied a soft balancing across categories, ensuring readers see a sufficiently diverse set of paths. The final distribution is as follows:

A note: category boundaries are not razor-sharp. Pieter Levels's product matrix straddles AI and SaaS (this book classifies it under different entries to surface different facets); "China" is broken out as its own category for the practical reason of differing data sources (see the next section), not because of any lane-level logic. Roughly 17 cases involve AI, reflecting the current tailwind—but we deliberately did not let AI fill the list. "Boring but profitable" physical businesses like EasyLunchboxes (lunch boxes, $1M+/year, from 2013) are equally important sources of inspiration.

IV. Data Sources and Limitations

The book's data was collected in tiers by credibility, from highest priority to lowest:

1. Verified by authoritative media. Figures publicly disclosed and checked by reporters at CNBC, the NYT, Fast Company, Tim Ferriss, and the like—the highest credibility.
2. Founders' public dashboards / self-reports. Indie Hackers (IH), real-time revenue screenshots on X (Twitter), public Stripe pages. levelsio and Marc Lou ($1,032,000, 2025, self-disclosed) belong here—transparent, but with a self-presentation bias.
3. Third-party estimates. GetLatka, Starter Story, the Empire Flippers Scoreboard. These are algorithmic or market estimates; for instance, ChatPDF at about $440K ARR (Latka 2025 estimate) diverges sharply from the founder's claim of $6M ARR, and the book labels all such cases "pending cross-verification."

Every figure is tagged with a year and a credibility tier (disclosed / media / estimated). The revenue-scale score (0–10) is a banded judgment, not an audited financial statement: across the full sample, 40 fall in the "high" band (seven figures), 44 in the "mid" band (six figures), and 5 are "top-tier." Do not equate the score with an audit result.

Limitations we must own up to: First, hard Chinese-language data is scarce. Domestic indie developers rarely keep the habit of publishing revenue dashboards; figures such as Xiaomao Fill Light's Pro version at "roughly RMB 300,000–400,000 cumulative" (from 2024) and idoubi's ShipAny at "1,000+ paying" (official site, 2025) are mostly estimates or non-revenue measures, not in the same league as overseas disclosures that are routinely precise to the dollar. Second, survivorship bias. This book records only the winners; the fellow travelers who failed are invisible, so the "replicability" score is a measure of path transparency, not a promise of success probability. Third, inconsistent measures. "Sales," "ARR," and "profit" are mixed across different cases, so be careful when comparing across them—Odd Muse's £22.5M is sales (GMV), which is not the same dimension as Sidekiq's roughly $7M (2024, close to pure profit).

Paradigm One: A Taxonomy of the Nine Business Models of the One-Person Company

The 100 solo and micro-team founders tracked in this book span nine categories: SaaS, content, knowledge products, e-commerce, AI, physical businesses, productized services, digital assets, and homegrown Chinese personal IP. They appear to travel separate roads, yet they share a single underlying logic: using the smallest possible labor leverage to move a meaningful amount of cash flow. The table below breaks these nine categories apart, annotating each with its core output, its way of making money, its typical scale range, and its barrier to entry for beginners—serving as the coordinate system for the entire book.

A Cross-Category Pattern: Replicability Is the Watershed

Lining up the nine categories side by side, the first pattern surfaces immediately: software, content, and information products are inherently high-margin and highly replicable, whereas e-commerce and physical businesses are operations-heavy and low in replicability. The marginal cost of selling the 10,000th copy of a Notion template is nearly zero—Thomas Frank's template revenue of $1,000,508 in a single year, and Easlo's roughly $779K in 2024, both prove the leverage of "build once, sell endlessly." SaaS works the same way: Carrd, maintained by one person, runs to about $1.5M ARR, and Sidekiq reaches about $7M. By contrast, in e-commerce, even as Odd Muse surged to £22.5M, behind it lay the real weight of inventory, supply chains, and returns; among physical businesses, Firebean's coffee roasting brings in roughly $90K/year and Hill Vending about $58K/month—both are at heart linear businesses where "selling one more means doing one more job's worth of work." In this book's four-dimensional scoring, "replicability" averages just 5.4 while "leverage" reaches 8.1, which is precisely why what truly opens the gap is not effort but whether the output can be delivered again and again, detached from the creator's time.

Pattern Two: AI Has Given Information Products the Form of Software

The second pattern is that the boundaries between categories are merging. Roughly 17 cases carry an AI attribute, and most of them are at heart "wrapping large-model capabilities into one-click tools"—HeadshotPro turned "taking ID-style photos" into a subscription of about $3.6M ARR, while levelsio's PhotoAI at roughly $1.6M+ ARR and ChatPDF at about $440K ARR (the founder suggests it is higher) prove out the same logic. They possess both the high margins of information products and the subscription compounding of SaaS, layered on top of an "era dividend" (the book's era dimension averages 7.5). Base44 is sharper still, selling a one-person project for $80M in cash—the most striking exit in the entire sample. For beginners, this means AI has partially lowered the SaaS barrier, once requiring deep engineering ability, to the level of "being able to call an API plus understanding a particular niche need."

Pattern Three: Productized Services Are the Bridge from Craftsperson to High Leverage

The third pattern speaks to those who have no product, only a skill: "productizing" a service is the realistic path to entering the high-replicability zone through a low barrier. Consulting is the classic time-for-money business, yet Designjoy used a fixed-monthly-fee "design subscription" model to reach about $4M cumulatively as one person, and roughly $3.1M in 2024; 180Sites packaged website building into a standardized delivery of about $950K ARR; and Hilvy applied the same thinking to pull its Webflow service from $74K in 2022 to about $318.8K ARR in 2024. Their shared move is to rewrite "hourly-quoted, negotiated-per-order" non-standard services into "fixed price, fixed process, fixed deliverable" quasi-products, thereby obtaining software-like predictable cash flow.

Pattern Four: National Context and the Particularities of China

The final pattern concerns geography. The United States accounts for 57 seats in the sample—the English-language market is the dominant home field, with mature content-subscription (Stratechery ~$5M+, Lenny's $2M+) and SaaS ecosystems. The 6 homegrown Chinese cases form a class of their own: they depend more heavily on platform traffic from WeChat, Douyin, and Dedao, with outputs centered on knowledge products and in-tool purchases—Li Yizhou's AI course, at roughly ¥50M in scale, proves the explosive power of traffic, while Ruan Yifeng's Weekly (399+ issues), V2EX (once valued at over ¥25M), and idoubi's ShipAny (1,000+ paying users) represent a more restrained, more engineering-minded long-termism. With the top ten on the composite index averaging 77.8, the leader Designjoy at 87.0, and the 100th-ranked case at 34.0, this continuous spectrum from peak to entry tells beginners this: your choice of category determines your ceiling and your gradient, not absolute success or failure—choosing the right leverage matters more than choosing the right industry.

Paradigm Two: The Seven Levers

In How to Get Rich (Without Getting Lucky), Naval Ravikant sorts the levers of wealth into four kinds: labor, capital, products (code), and media. He places special emphasis on the latter two as the new, "permissionless" levers — the marginal cost of copying a line of code or a piece of content approaches zero. That framework explains the wealth-creation logic of the software and internet age, but inside the lived reality of the company of one, it is too coarse. Someone selling digital downloads on Etsy, someone writing a paid newsletter, and someone sniping expired domains are not pulling the same lever at all — yet all three get lumped under "products/media." The 100 cases in this book force a finer cut: under the specific constraints of the company of one (no team, no funding, time as a hard ceiling), only seven levers are used over and over. Understanding them is like holding a cross-reference chart of "what moves what."

1. The Code / Product Lever

Definition

Write software once, sell it to an unlimited number of users. This is the lowest-marginal-cost lever there is: the cost of a server carrying one more subscriber is negligible.

How it compounds

Subscriptions plus a product matrix. One underlying capability can splinter into many products. Pieter Levels (levelsio) built a matrix out of PhotoAI, Nomad List, RemoteOK, and InteriorAI, totaling roughly $3.1M+ ARR in 2025; Marc Lou strung ShipFast, DataFast, CodeFast, and TrustMRR into a single assembly line, self-reporting $1,032,000 in 2025.

Representative cases in this book

Designjoy (Brett Williams, around $4M in 2024), Carrd (AJ, roughly $1.5M ARR), Sidekiq (Mike Perham, around $7M in 2024), GymStreak (roughly $2.5M ARR). The extreme value is Base44 (Maor Shlomo), acquired by Wix for $80M in cash in 2025 — the exit form of the code lever.

Risks

Ongoing maintenance, customer-acquisition cost, and the platform or tech stack being disrupted. Code does not sell itself, which is why almost no pure code lever runs without a content or audience lever stacked on top.

2. The Content / Attention Lever

Definition

Create text, video, or audio once, distribute it to an unlimited number of readers, then monetize the attention (subscriptions, sponsorships, advertising).

How it compounds

The compounding lives in the "reader list" — it only grows, never shrinks, and the larger it gets the more pricing power it carries. Stratechery (Ben Thompson), estimated at roughly $5M+/year on 40,000 subscribers × $120; Lenny's Newsletter at $2M+/year for the newsletter and $500K+/year for the podcast. Not Boring (Packy McCormick) cleared $1M from sponsorships alone in 2021.

Representative cases in this book

Letters from an American (Heather Cox Richardson, 2.6–2.9 million subscribers) is the largest subscriber count in the entire sample; Justin Welsh reports $12.5M+ in cumulative earnings; plus Stratechery, Newcomer ($1.6M in 2023), and Feed Me (Emily Sundberg, subscriptions of roughly $400K+).

Risks

Tightly bound to the creator's personal output capacity — stop publishing and the flow stops. Platform algorithms and distribution channels can change face at any moment.

3. The Audience / Community Lever

Definition

Upgrade "readers" into a community that gathers and serves one another. The difference from the content lever: value comes not only from you, but from the members among themselves.

How it compounds

Network effects plus membership repurchase. Small Bets (Daniel Vassallo) accumulated $824,409 from November 2021 to October 2023, selling a lifetime ticket into the community; The Lab from Creator Science (Jay Clouse) reached around $830K in 2024, roughly half of it from community subscriptions.

Representative cases in this book

Ship 30 for 30 (around $1M in its 2021 launch year), Part-Time YouTuber Academy (Ali Abdaal, around $4.5M); on the Chinese-language side, V2EX (Livid) and Ruan Yifeng's Weekly for Tech Enthusiasts both retain people through community density rather than one-way broadcasting.

Risks

A community needs tending, so its marginal cost is not zero; once the atmosphere sours it is extremely hard to repair, and it relies heavily on the operator's personal credibility.

4. The Automation / AI Lever

Definition

Use AI or workflows to replace labor that would otherwise require hiring, letting one person shoulder team-scale output. This is the addition absent from Naval's four levers, yet the best fit for today's company of one.

How it compounds

AI pushes the marginal cost of both "content production" and "fulfillment" toward zero at the same time. HeadshotPro (Danny Postma) at roughly $3.6M ARR and $300K MRR; Tibo's Revid/Outrank/SuperX bundle self-reporting around $1M/month in 2025; StoryShort plus useArtemis (Samuel Rondot) reaching $35–41K/month. Across the full sample, around 17 cases have AI directly at their core.

Representative cases in this book

PhotoAI, ChatPDF (roughly $440K ARR), Formula Bot ($500K ARR), AudioPen ($73K in its first two months). On the Chinese-language side, Xiaomao Buguang Deng (Chen Yunfei) used an extremely light product to drive its Pro version to roughly ¥300,000–400,000 cumulatively through automation.

Risks

Models and APIs are controlled by others; one shift in the underlying layer and the moat evaporates. Homogenization is brutally fast — today's AI wrapper has ten competitors tomorrow.

5. The Capital / Asset Lever

Definition

Use money (or appreciating assets) to make money, rather than trading time for money. In the company of one it often appears as "buying assets that generate cash flow or royalties."

How it compounds

An asset portfolio plus the long tail of royalties. Spencer Haws's portfolio of niche sites at roughly $5M/year; Mike Carson's Park.io, sniping expired domains, peaking at around $125K/month; Sahil Lavingia's personal Rolling Fund starting at roughly $5M/year. The royalty type, such as Hugh Howey's Silo series — millions of copies, 40+ languages — is written once and earns for the long term.

Representative cases in this book

Niche content site flips (the Empire Flippers marketplace sample, roughly $589M in cumulative transactions in 2025), independent type foundries (Set Sail Studios at roughly $7,000/month in royalties), and self-published KDP royalties.

Risks

Requires upfront capital or upfront work; asset-price volatility and platform-policy changes (domains, Kindle revenue splits) hit cash flow directly.

6. The Brand / IP Lever

Definition

Deposit "trust" into a brand or IP that commands a premium and extends — so the same product sells for more, and a new category directly inherits the momentum.

How it compounds

A brand decouples order value and repurchase rate from the product itself. Odd Muse (Aimee Smale), roughly £22.5M in sales in 2024, riding on a DTC brand narrative rather than any single item; Tabs Chocolate reached around $11M in 18–24 months, essentially by turning chocolate into a content-driven IP.

Representative cases in this book

Visualize Value (Jack Butcher, around $180K/month in 2020) turned "visualizing ideas" into licensable IP; Mark Manson (roughly $2–2.5M/year) and Dan Koe (self-reporting roughly $4M+ in 2024) both use personal IP to feed multiple product lines.

Risks

Brand-building is slow and its payoff lags; personal IP is deeply bound to the founder, so reputational risk is business risk.

7. The Platform / Network Lever

Definition

Borrow the traffic and trust of an existing large platform to distribute your own thing — or become a small platform on which others transact.

How it compounds

Borrowing scale means reusing the attention a platform has already pooled, at zero acquisition cost to you. Easlo and Thomas Frank (full-year 2022 of $1,000,508) sell templates within the Notion/Gumroad ecosystem; ecommemily and Emily's POD store run entirely on Etsy's ready-made buyer traffic (around $236K in 2024). Those who become the platform include idoubi's ShipAny + MCP.so (1,000+ paying customers and 2,000+ live sites by 2025).

Representative cases in this book

Bannerbear (Jon Yongfook, crossing $1M ARR in 2025, deeply embedded in developer workflows), Closet Tools (parasitic on Poshmark), WIP/BetaList (Marc Köhlbrugge, nearly $1M cumulatively over ten years).

Risks

Platform take rates, policies, and algorithms hold the power of life and death; over-reliance on a single platform is handing over your lifeline.

The Seven Levers at a Glance

The Truth: Top Cases Are All Superpositions

A single lever can get you started, but it rarely sets the ceiling. In this book's four-dimensional scoring, "leverage" averages a striking 8.1 — far above "replicability" at 5.4 — and the high-scoring cases almost all rely on stacking. Designjoy uses code (a self-built ordering system) + brand (Brett's personal credibility) + platform (building in public on X) all at once; Justin Welsh uses content (the newsletter) + audience (community courses) + platform (the LinkedIn algorithm windfall); PhotoAI is a triple superposition of AI + code + founder IP (levelsio's public figures).

A single lever decides whether you can start; the way you stack levers decides how far you can go. First use one lever that pushes marginal cost to its lowest to prove the model, then stack a second and a third to raise the ceiling — this is the same growth curve that recurs across all 100 cases.

So choosing a lever is not a multiple-choice question. First ask what you already hold: if you have technical skill, pick 1 or 4; if you have the urge to express, pick 2 or 3; if you have capital or older work, pick 5; if you have taste and personality, pick 6; if you spot a platform windfall, pick 7. And the second step is always the same: stack another one to shore up the weakness of the first.

Paradigm Three: The Growth Playbook

A one-person company has no marketing department and no sales team. Growth can't come from throwing more bodies at the problem; it can only come from choosing the one playbook that matches your own native strengths. A playbook isn't mysticism. It's a repeatable, testable causal mechanism: the actions you take produce a particular kind of traffic and trust, which in turn convert into revenue. Among the 100 cases in this book, nearly every high-index example slots cleanly into one of the six paths below — and these are not six matters of stylistic taste, but six structurally different customer-acquisition engines. Understand the mechanism first, then find your place; only then will you avoid forcing someone else's success onto a frame it was never built for.

1. Build in Public

The mechanism

Treat the process of building the product as content in its own right: post your revenue in real time, post your code, post your failures. Transparency pays a double dividend — algorithms favor frequent, authentic updates, and audiences who "watch it grow up" develop a sense of ownership, so conversion carries almost zero persuasion cost. Pieter Levels (levelsio) is the textbook case: he made PhotoAI's MRR dashboard public (roughly $132–138K/month, $1.6M+ ARR), and the bulk of the traffic behind his entire $3.1M+ ARR product portfolio comes from his real-time disclosures on X. Marc Lou turned "earning $1,032,000 in 2025" directly into a personal brand, then routed that attention back into his ShipFast template portfolio.

Who it suits

Extroverted, developer-type founders who don't mind exposing their numbers and who ship fast.

Common pitfalls

Treating "posting" as the goal and neglecting the product itself; once MRR stalls, building in public becomes a source of pressure instead. Transparency is an asset only when the product has a real growth curve — otherwise it's a liability.

2. Audience-first (content before product)

The mechanism

First cultivate a group of people who trust you using free content, then sell to that group. The traffic is in place before the product exists, so you have your first customers the moment you launch. Justin Welsh built an audience of hundreds of thousands through LinkedIn and X content, then sold courses and a community, for a self-reported cumulative $12.5M+. Dan Koe grew The Koe Letter into a newsletter worth roughly $4M+. Sahil Bloom's The Curiosity Chronicle earns roughly $70,000/month from sponsorships alone.

Who it suits

Writers who are good at producing opinions consistently and can endure a 6–12 month, revenue-free audience-building phase.

Common pitfalls

A mismatch between audience and product — cultivating a crowd that only wants free entertainment, then trying to sell them a high-priced course. An audience-building phase that drags on too long, breaking cash flow, is the number-one reason people quit.

3. SEO / long-tail organic traffic

The mechanism

Use one high-purchase-intent keyword, or a batch of long-tail terms, to build a tool site or template site so that Google keeps sending precise traffic for free. The compounding is extreme, but it's slow to take effect. Formula Bot (excelformulabot) rode search demand for terms like "excel formula" to $500K ARR, with 2025 MRR above $42K. ChatPDF reached roughly $440K ARR on the long-tail phrase "chat with pdf." Spencer Haws's portfolio of niche sites has been valued at roughly $5M/year. Easlo's Notion templates (roughly $779K in 2024) likewise feed heavily on template-related long-tail searches.

Who it suits

Patient people who understand content structure and are willing to do technical SEO — especially introverted founders who'd rather not appear on camera.

Common pitfalls

An algorithm update can wipe you out overnight; AI Overviews are now devouring informational long-tail queries. Over-reliance on a single keyword means handing your lifeline to one platform.

4. Productized service (standardized, packaged service)

The mechanism

Package a highly customized service into a fixed-price, fixed-delivery, subscribable "product," cutting out the quoting and negotiation steps. Brett Williams's Designjoy tops this book's ranking (index 87.0): one person, a fixed monthly fee, an unlimited design-request queue, reaching $3.1M in 2024 and roughly $4M cumulative. Hilvy (a productized Webflow service) climbed in three annual steps from $74K to $252K to roughly $318.8K ARR, proving the model can be replicated.

Who it suits

Service providers with solid delivery craft who want to monetize immediately rather than cultivate an audience; cash flow can turn positive the same day.

Common pitfalls

One person's capacity is the ceiling — Designjoy's moat is precisely that Brett refuses to scale, using systematization to push a single person's output to its limit. Taking orders blindly without building process will trap you inside delivery.

5. Wedge → Expand (enter on a single point, then extend)

The mechanism

Enter the market through one extremely narrow, extremely painful need to build trust, then once you're established, expand laterally into adjacent needs. Damon Chen first built Testimonial.to (roughly $800K ARR), nailing the narrow slot of "collecting customer testimonials," then extended into PDF.ai for a combined $1.3M+ ARR. Tony Dinh entered with the small DevUtils tool and expanded into the TypingMind portfolio at roughly $137K/month, with B2B team plans accounting for over 50%.

Who it suits

Founders who can spot a "small and sharp" pain point and have the imagination to extend a product from it.

Common pitfalls

Expanding too early — rushing to build a second product before the entry point is even profitable, ending up shallow on both fronts. The discipline of the wedge is to "drive one needle all the way through first."

6. Distribution-first (claim the channel before building the product)

The mechanism

First lock onto a distribution channel that has a structural traffic dividend (a particular platform, community, or emerging format), then build the product around that channel's characteristics. Tibo's Revid/Outrank portfolio hit the dividend in short-video and SEO-automation distribution, reaching roughly $1M/month. Samuel Rondot's StoryShort fed on the AI short-video distribution dividend of TikTok/Reels, with the combined figure rising to $35–41K. Etsy's Emily (ecommemily, $236K+) and Dylan Jahraus (Begonia Rose, roughly $1.7M) are both, at heart, cases of "eat the platform's distribution dividend first, talk about brand later."

Who it suits

People with a sharp nose for channels who move fast and can grab position inside the dividend window.

Common pitfalls

The channel belongs to someone else — the moment the platform's rules change (Etsy raises fees, TikTok throttles reach), the growth engine cuts out. Distribution-first must be paired with "converting traffic into owned assets (an email list, a brand) as early as possible."

Selection at a glance

The first-90-days cold-start checklist

A growth playbook isn't something you reach for after launch. The actions of your first 90 days decide which engine you'll end up with.

1. Days 1–15 | Choose a playbook, not a style. Answer honestly against the table above: Am I willing to be on camera? How long can I go without revenue? Can my craft be monetized the same day? On that basis, lock in one primary path and permanently abandon the other three-plus.
2. Days 16–30 | Define one quantifiable, narrow-slot need. Learn from Damon Chen, who did only "collect testimonials" — write your target user and single pain point into one sentence a stranger can grasp in three seconds.
3. Days 31–45 | Build the minimum visible asset. The SEO type first builds the skeletons of 3–5 long-tail content pieces; the Audience type gets a daily posting cadence running on a single platform; the service type builds one fixed-price landing page (Designjoy-style: price, queue, and case studies, the three-piece set).
4. Days 46–60 | Get your first 10 real pieces of feedback by hand. Don't automate, don't scale; serve your first 10 people personally and record their exact words — this is the ammunition for all your later copy.
5. Days 61–75 | Start accumulating owned traffic. Whatever path you take, build an email list from day one. Distribution-first players especially must funnel platform traffic into the list, so they don't leave their lifeline in someone else's hands.
6. Days 76–90 | Make your first number public and set a cadence. Even if it's just "the first paying customer," make it public once in Build-in-Public fashion to test the audience's reaction; at the same time, set a sustainable update/delivery cadence — a one-person company's growth compounds through rhythm, not bursts of inspiration.

By the end of 90 days, what you want is not a pretty revenue figure, but the embryo of a clearly directed, self-reinforcing customer-acquisition engine. Get the playbook right, and the rest is repeating the same action a thousand times.

Paradigm Four: The Stack Map

Pull apart the hundred names on this list and one structure keeps surfacing: the doer is a single person, but the work is done by an entire suite of software. Brett Williams carried Designjoy to roughly $3.1 million in revenue in 2024 (GetLatka) on his own—no employees, no project manager, no support team. His “team” was a pipeline stitched together from Webflow, Stripe, Notion, and a scheduling tool. levelsio (Pieter Levels) built PhotoAI to roughly $132–138K in MRR (founder-disclosed, 2025) on the strength of open-source models running on Replicate, Stripe for payments, and a single VPS. A one-person company is not “do everything yourself.” It is precisely the opposite: do almost nothing yourself, and own only the one link no one can replace. This section lays out a stack map you can build straight from, then explains the leverage logic behind it.

A Stack Map You Can Build From

The table below is arranged along the eight stages an indie venture passes through from birth to growth. The monthly cost figures are rough ranges—the overwhelming majority of these stages can run on a free tier or a few dozen dollars in the early days. The one stage that truly burns money is usually AI inference, and even that scales linearly with revenue and stays controllable.

The Leverage of a One-Person Company: Outsource the Non-Core to Software, AI, or Freelancers

The traditional company solves capacity by hiring; the one-person company solves it by buying leverage. Across the hundred samples in this book, the leverage dimension scores an average of 8.1 out of 10, and 49 of them—nearly half—are high-leverage cases scoring 9 or above. That is no coincidence; it is the defining trait of the species. Leverage comes from three sources, mapping to the three classes of supplier in the table above.

Software leverage: build once, run infinitely

The essence of SaaS and automation tools is to freeze repetitive labor into a configuration that runs once. Marc Lou's ShipFast/DataFast portfolio reported $1.032 million in 2025 (self-disclosed); what he sells is the act of website-building packaged into a template—buyers save weeks using his template, and he uses tooling to turn “selling the template” into automated delivery. Carrd's AJ (@ajlkn) built a single website tool to roughly $1.5 million ARR (2024), serving tens of thousands of customers without spending an extra minute on any one of them. The hallmark of software leverage is marginal cost approaching zero: Stripe processes your first dollar and your millionth on virtually the same cost structure.

AI leverage: a tenfold multiplier on content, creative, and answers

AI is the newest lever—and the fiercest—to arrive after 2023. About 17 of the cases in this book involve AI, and the era-dividend dimension scores an average of 7.5, indicating the judges broadly regard this as the direction most worth betting on today. Danny Postma's HeadshotPro used generative models to compress “getting a set of professional headshots”—something that once required a photographer, a studio, and a retoucher—into a single upload button, reaching roughly $300K MRR (2024). Samuel Rondot's StoryShort uses AI to batch-generate short videos, with his portfolio income climbing to $35–41K per month (2025–26). AI turns the stages that once demanded hiring—copywriting, illustration, video editing, customer support—into API calls, dropping costs from a monthly salary to a few cents per thousand calls.

Human leverage: buy on demand, keep no team

Not everything can be turned into software. Stages that require judgment, taste, or live communication can be bought per task through Fiverr, Upwork, or a VA—rather than under a standing contract. Justin Welsh (The Saturday Solopreneur, self-reported cumulative earnings of more than $12.5 million) has long stressed “contractors over employees”: need editing, buy editing; need design, buy design; disband when the job is done, so the founder stays asset-light with zero management burden. This inverts the traditional startup sequence of “hire first, then find the work”—the one-person company has the work first, buys people for the work, and zeroes out the relationship once the work is finished.

How to Use AI to Scale One Person Into “1+N”

The real step up is not using AI to help you do the work, but orchestrating AI into a “virtual team”—letting one person direct N tireless digital employees. This “1+N” structure is built roughly like this:

1. 1 = you, doing only the core that cannot be outsourced. Set direction, make taste judgments, build trust with real humans. Lenny Rachitsky's newsletter reached more than $2 million a year (CNBC 2024), and at its core is his own judgment and network—something AI cannot replace and should not be handed off.
2. N1 = the AI content worker. Use an LLM to batch-draft, rewrite, and distribute across languages and platforms. Tibo (Thibault Louis-Lucas) drove his Revid/Outrank portfolio to roughly $1 million a month (2025), in essence by putting the entire content production line on AI.
3. N2 = AI support and ops. Wire common questions into the AI in Crisp/Intercom, or use a tool like ChatPDF (roughly $440K ARR, Latka 2025) to automatically digest document-based queries, so the founder need not watch the ticket queue.
4. N3 = an AI + automation delivery pipeline. Zapier/Make/n8n string “payment → AI generation → delivery → community onboarding → follow-up” end to end. AudioPen (Louis Pereira, $73K in the first two months) ran exactly such a near-unattended voice-to-text pipeline.

The end of leverage is not “doing more,” but “doing less while producing more.” When software takes over the repetitive, AI takes over production, and contractors take over the specialized, the time the founder has left is just enough to think clearly about what is actually worth doing—and that, of all things, is the one thing not a single one of these hundred people outsourced.

The real value of this map is not in telling you which specific tool to use (tools change yearly), but in establishing a default stance: at every new stage, first ask, “Can I get software, AI, or a freelancer to do this for me?” Only when the answer is “No—and this is precisely my moat” do you do it yourself. Carry that rule all the way through, and a single person can run a million-dollar business—the forty million-dollar-tier samples in this book (scoring in the 7–8 band) are forty pieces of evidence.

Paradigm Five: Risk and Moats

The most seductive feature of a one-person business is that revenue is decoupled from headcount. Designjoy, run by a single person, reached roughly $3.1M in 2024 (GetLatka); levelsio's product portfolio runs at roughly $3.1M+ ARR. Neither has a team diluting the upside. But that same structure means every risk is concentrated on a single point of failure. This chapter is not about how to get started. It is about two things only: what can take you to zero, and what can keep you from being copied. Think these two through clearly, and a one-person business graduates from "a lucky sole trader" into "a durable asset."

A. Risks and Ceilings

1. Platform Dependence: Algorithms and Bans

The vast majority of content and e-commerce paradigms are tenants on someone else's platform. Change the algorithm, and traffic evaporates overnight. The Etsy seller Three Bird Nest (Alicia Shaffer) hit roughly $960K in 2014–2015, then was engulfed in controversy—accused of outsourcing production and failing to meet the "handmade" definition—and was hit by a platform-policy backlash. Closet Tools (later renamed Resellbot) peaked early at roughly $38K–41K MRR, then slid back to roughly $30–40K MRR after 2023, the core cause being tightened rules on Poshmark, the platform it lived off. Contrast Pieter Levels: Nomad List, RemoteOK, and PhotoAI all run on his own domains and self-built traffic—precisely why he has survived round after round of platform turbulence.

2. Key-Person Risk: You Are the Single Point

A one-person business's greatest asset and greatest liability are the same person. The strong-personal-IP paradigm is especially exposed: for Mark Manson (roughly $2–2.5M/year), Dan Koe (roughly $4M+), and Ali Abdaal's PTYA (roughly $4.5M), the brand is the person's own face and voice. The moment they stop publishing, fall ill, or see their persona collapse, the cash flow stops with them. By contrast, the SaaS paradigm "de-persons" the business: users of Carrd (AJ, roughly $1.5M ARR) and Sidekiq (Mike Perham, roughly $7M) simply do not care who the founder is—the product runs itself. AudioPen (Louis Pereira) made $73K in its first two months; TalkNotes was acquired for $200K, all cash. SaaS can exit cleanly precisely because the value accrues to the product rather than the person.

3. Growth Ceilings

A one-person structure inevitably has a ceiling. The service paradigm hits it hardest: Double Your Freelancing reports course income of $100K+/month, but the underlying consulting time cannot be replicated. This is exactly why the high-leverage cases (this book's 49 cases scoring ≥9 on leverage) almost all swap delivery from "time" into "software or content." Designjoy used a standardized subscription plus an asynchronous workflow to give design services a SaaS-like marginal-cost structure, which is how one person sustained $3.1M. To break through the ceiling, you either productize (service → software) or assetize (labor → royalties, as with Hugh Howey's Silo series and its millions of copies in royalties).

4. Burnout

No colleagues means no buffer. FeedbackPanda (Arvid Kahl) chose to sell at roughly $55K MRR, and one of the publicly stated reasons was that the couple running it were exhausted—a rare sample within one-person businesses of "voluntarily exiting at a high point to avoid burnout." Kahl later turned that experience into a methodology that fed back into his income. Against that, levelsio's long-term, high-intensity build-in-public makes sustainability highly dependent on personal energy management. Burnout is not a soft problem; it is directly the trigger for key-person risk.

5. Compliance and Taxes

Cross-border one-person businesses routinely overlook: VAT/sales tax, platform withholding, royalty withholding tax on platforms like KDP, and digital-services taxes across different jurisdictions. At the scale of Odd Muse (Aimee Smale, roughly £22.5M) or Tabs Chocolate (roughly $11M), a single compliance error means penalties in the hundreds of thousands. The smaller you are, the earlier you should outsource your entity, bookkeeping, and tax work—one of the few places where it is worth violating the "one-person" principle to spend money.

6. The Red Line of Controversy: Inflated-Outcome Knowledge Products

The knowledge and course paradigm (this book's 16 knowledge cases) most easily crosses the red line of overpromising returns. In China, Li Yizhou's AI course (Artificial Intelligence for Everyone, roughly RMB 50 million) faced mass skepticism in 2024 over course quality and marketing controversies and was pulled from sale—a textbook negative example. Contrast Sahil Bloom (roughly $70K/month, mainly from sponsorships rather than course commissions) and Ship 30 for 30 (roughly $1M, selling a writing program rather than "get rich quick"): sustainable knowledge products sell a definite delivery of skill, not unredeemable promises of results.

B. Moats

1. Brand / IP

A personal brand is the cheapest and hardest-to-copy moat there is. The pricing power of Stratechery (Ben Thompson, roughly $5M+, about 40,000 subscribers × $120) and Lenny's Newsletter rests entirely on "who the author is." Competitors can copy the format; they cannot copy the trust.

2. SEO / Content Assets

Content is a stock asset that compounds. Spencer Haws's portfolio of niche sites runs at roughly $5M/year (Starter Story estimate), and Marketing Examples (Harry Dry, roughly $360K/year) relies on pages built up over years that keep delivering free traffic. Every old article is a salesperson who never clocks off, twenty-four hours a day.

3. Network Effects / Community

Community locks users to one another. Small Bets (Daniel Vassallo, $824,409 cumulative from November 2021 to October 2023) and Creator Science's The Lab (roughly $830K in 2024, about half of it from the community)—members stay because of the other members, not the founder, and that stickiness is hard for an outsider to replicate single-handedly. V2EX (Livid) was once valued at over RMB 25 million, with all of that value residing in the community itself.

4. Proprietary Data

Exclusive data is the hardest barrier in the AI era. The real asset of HeadshotPro (roughly $3.6M ARR), PhotoAI (roughly $1M+ ARR), and ChatPDF is not just the model but the accumulated user data, prompt engineering, and generation-feedback loop. Anyone can call the same models; no one else has your training and fine-tuning data.

5. Distribution Channels

Owned distribution is the root of resistance to platform risk. Marc Lou's ShipFast portfolio (roughly $1,032,000 in 2025) uses an owned audience to cold-start new products again and again; Tibo (roughly $1M/month) likewise channels his own traffic to new tools. Once a channel is owned, it becomes a reusable launchpad.

6. Compounding Time

The most underrated moat is simply "time in the arena." Ruan Yifeng's Tech Enthusiast Weekly has published more than 399 issues (launched in 2018), and Pat Flynn's SPI single-month figure of $167,553 in 2017 was the result of a decade of accumulation. No newcomer can buy the seven years of trust you have already walked through. Time itself is the barrier.

Risk → Hedge Reference Table

The life and death of a one-person business turns not on how fast you can run, but on whether others can copy the road you run and whether the ground beneath you will hold. Swap platform risk for owned channels, swap personal risk for product assets, then build five walls—brand, content, community, data, and time. That is the watershed that takes you from "a high-multiple one-off" to "a durable asset."

Paradigm Six: Data in Perspective

The first five chapters were about paths and methods. This chapter does one thing only: it places all 100 samples in a single coordinate system and uses the aggregated data to infer their shared patterns. Every conclusion rests on one precondition — the sample is a pool of high-scoring cases filtered for being “independent, run by one person or a very small team, and publicly verifiable.” It is not a random draw from all founders on the internet. So each pattern below holds “among already-successful independents,” not “do this and you will succeed.” That is precisely what makes each one falsifiable.

1. Category and Geography: Who Made the List

Break the 100 samples out by category, and the leaders are software and content/knowledge — not the e-commerce or physical businesses that conventional wisdom might expect.

Geographically, the United States alone holds 57 seats — more than half. The second tier is the United Kingdom (8), China (6), the Netherlands (5), Canada (5), and France (4). The density of the Netherlands and France is especially notable: the 5 Dutch cases are almost all AI/SaaS (Pieter Levels's PhotoAI, Danny Postma's HeadshotPro), while the 4 French cases are all product-portfolio players (Samuel Rondot, Tibo, Marc Lou, Nico Jeannen). This shows that the high-scoring independent is not an American monopoly, but spreads along the axis of “English-language content market plus developer culture.”

2. Revenue Scale and Team Size

Sorted by the researcher's 0–10 revenue score (an estimate, not precise financials), the distribution is as follows:

The Mid and High tiers together hold 84 seats — the overwhelming majority. This means the realistic sweet spot for an independent is $100K to $10M in annual revenue: uncapped on the upside but extremely rare (only 5 at the top, such as Base44, acquired by Wix for $80M in cash, and Sahil Lavingia's Rolling Fund), with a real floor at the bottom (only 2 early-stage cases, both still in validation). The story that “one person can casually make tens of thousands a month” is, in the data, a minority phenomenon. What is truly dense is the solid million-dollar business.

On team size, the samples are almost all solo founders or pairs. The very small teams worth naming number in the single digits: Tabs Chocolate (Brocato + Lewin), Ship 30 for 30 (Cole + Bush), Wait But Why (Urban + Finn), FeedbackPanda (Kahl + Simpson). Even so, Maor Shlomo's Base44 reached an $80M exit as a solo founder, and Pieter Levels single-handedly maintains a four-product portfolio — PhotoAI, Nomad List, RemoteOK, InteriorAI — with ARR exceeding $3.1M.

3. Monetization, Time to Revenue, and Stacked Leverage

The four-dimension averages give this batch its “genetic profile”: revenue 6.3, replicability 5.4, leverage 8.1, timing 7.5. Leverage runs away from the field, and high-leverage cases (≥9) number as many as 49 — close to half — while high-replicability cases (≥7) number only 25. This contrast is itself the core finding: these people win on leverage, not on being “easy to copy.”

Monetization falls into three types, and high-scoring cases routinely stack them:

• Subscription / SaaS recurring revenue: TypingMind ~$137K/month, GymStreak ~$208K MRR, Carrd ~$1.5M ARR.
• Paid content / knowledge products: Justin Welsh $12.5M+ cumulative, Lenny's newsletter $2M+/year, Thomas Frank Notion templates at $1,000,508 in a single year.
• Physical / e-commerce goods: Odd Muse £22.5M in sales, Tabs Chocolate ~$11M over 18–24 months.

The time from launch to first revenue, among samples with clear disclosures, is generally very short: AudioPen hit $73K in its first two months; TalkNotes reached $200K before being acquired all-cash; Doing Content Right exceeded $130K within 8 months of launch; Nathan Barry's self-published e-book made $145,471 in 2012, the year it launched. Slow burners exist too, but reaching the million-dollar level often takes years — Easlo went from ~$500K in 2023 to ~$779K in 2024, and Carrd from ~$600K in 2023 to ~$1.5M in 2024.

4. The Rise of AI-Native Cases

Eight cases are tagged as AI by category; counting those whose name or positioning involves AI brings the total to about 17. The point is not the count but the distribution and timing: AI cases cluster at the front of the index — HeadshotPro 77.0, PhotoAI 77.0, the StoryShort group 77.0, ChatPDF 75.0, Base44 73.0, the Tibo group 74.0 — nearly all ≥73. Against a timing-dimension average of 7.5, AI cases are precisely the group that pegs the “timing dividend” dimension to the maximum.

Taken together, these 100 samples point to a profile that is unromantic but verifiable: one person (occasionally two), building a software or content/knowledge product, monetizing through subscriptions or digital goods, stacking at least two forms of leverage, operating in the $100K-to-$10M range — while AI is now pushing the “timing” dimension to its ceiling, making the window for a solo founder to reach a tens-of-millions exit shorter than ever. A top-ten average of 77.8, a leader at 87.0, and a hundredth-place finish at 34.0 — this index curve, running from Designjoy to Firebean Coffee, has never measured effort. It measures the thickness of leverage.

Paradigm Seven: An Action Framework for China

The paradigms dissected in the previous six chapters draw overwhelmingly on samples from the United States (57/100) and Europe. Yet two further threads run quietly through this book's 100 cases. The first is six purely Chinese examples: idoubi, Cat Fill Light (小猫补光灯), Li Xiaolai, Li Yizhou, Ruan Yifeng, and V2EX. The second is the large contingent of global players who are ethnically Chinese or Asian: Chinese-American Chen Damo's PDF.ai (roughly $1.3M+ ARR), Singapore/Malaysia-based Easlo's Notion templates (about $779K), and Vietnam's Tony Dinh with TypingMind (about $137K/month). These two threads map precisely onto the two roads open to a Chinese company of one: build for the domestic market, or go global and build for the world. This chapter is not about sentiment. It answers a single question—which of the first six chapters can be copied wholesale, which will fail to take root, and which road you should walk.

1. Which Paradigms Transfer Directly, and Which Face Local Friction

The test is simple: the closer a paradigm sits to “digital product plus globally deliverable,” the lower the cost of transfer; the more it depends on local payments, platform distribution, and offline fulfillment, the greater the local friction.

The conclusion: software, services, templates, and prompts are the four lowest-cost entry points for a Chinese company of one. Their differences concentrate on “how to get paid and how to acquire customers,” not on “whether it can be built at all.” idoubi's ShipAny (1,000+ paying customers, 2,000+ live sites) is exactly the Chinese proof of localizing the “global-market template/scaffold” paradigm.

2. Domestic Channels vs. Overseas Channels: A Comparison Map

Your channel determines what your content looks like and where your money comes from. Domestic and overseas channels are two almost entirely incompatible systems; choosing the wrong one means your effort is wasted.

Three differences matter most. First, the overseas model of building an audience on X—levelsio (PhotoAI, about $1M+ ARR), Marc Lou ($1,032,000 in 2025)—maps domestically onto Xiaohongshu and WeChat Channels, but Chinese platform algorithms are more closed and outbound links are harder, so the dividend from “build in public” is markedly smaller. Second, China has no mature paid-newsletter ecosystem—Lenny's path to $2M+/year on Substack simply doesn't work domestically, and must be swapped for Official Account advertising or Knowledge Planet (compare Li Xiaolai's mature paid habit at ¥199/year on Dedao). Third, Xianyu is a uniquely Chinese low-cost validation ground, well suited to selling “virtual services/templates” to test demand before any development begins—functionally close to Gumroad overseas.

3. The Realistic Path for Going Global as an Indie Developer—and Three Pitfalls

For Chinese developers, building a global SaaS/AI business is often the higher-ceiling choice. This book's SaaS (20) plus AI (8) cases total 28 samples, densely clustered at the million-dollar level. But what the book's overseas ethnic-Chinese successes (Chen Damo, Tony Dinh, Easlo) share is not stronger technology—it is that they cleared three localization thresholds.

Pitfall One: Getting Paid

This is the first and hardest hurdle. Stripe does not directly support mainland Chinese entities, so the overwhelming majority of global builders go through an overseas corporate entity (U.S. LLC / Hong Kong / Singapore) plus Stripe, or use a Merchant of Record such as Paddle or Lemon Squeezy (which handle global tax and compliance on your behalf). TalkNotes's all-cash $200,000 exit on Acquire.com and idoubi's payment system both rest on this collection infrastructure. Until you solve payments, everything downstream is just talk.

Pitfall Two: Customer Acquisition

Overseas acquisition relies heavily on X, Reddit, Product Hunt, and SEO. The pattern in this book is “product portfolio plus building in public.” levelsio cross-pollinates traffic across multiple products (PhotoAI/Nomad List/RemoteOK); Tony Dinh reached $137K/month with the TypingMind/DevUtils portfolio; Marc Lou ships several products a year. For Chinese developers, the real barrier is sustaining English-language output about “the process of building a product” on X—a dual adaptation of language and culture, and the place where people most often give up halfway.

Pitfall Three: English Content

AI tools (this book covers roughly 17 AI cases) have sharply lowered the barrier to writing in English, but “native” fluency and a sense of trust still require polish. ChatPDF (Germany, about $440K ARR) and Formula Bot ($500K ARR) prove the point: when the product itself solves a clear, urgent need, English content only has to state the value plainly—it doesn't need literary flair. Going global with a “tool-type” rather than a “content-type” product first lets you sidestep the single biggest weakness in English expression.

4. A Decision Framework: Choosing Your Company of One

Converge all of the above into one self-auditable decision path—choose your paradigm at the intersection of your strengths × available leverage × channel, rather than chasing whichever paradigm makes the most money. In this book's four-dimension average scores, leverage (8.1) and timing (7.5) rank far above replicability (5.4), which tells you: picking the right leverage and the right timing dividend matters more than simply copying someone else.

This book's plainest lesson holds just as well for Chinese readers: a small starting point is not the problem (the 100th-ranked Firebean, at about $90K/year, still works; Voicy, at only about $1,600 in 2025, is still on its way), while choosing the wrong track and the wrong channel is the problem. Locate yourself with the checklist first, then dig into this book's corresponding paradigm chapter for tactics—that, precisely, is the starting line for a Chinese company of one.

Browse by category

Appendix: A Quick Reference to Terms

The one hundred cases in this book span eight broad categories—SaaS, content, paid knowledge, e-commerce, AI, physical products, services, and assets—yet the vocabulary is remarkably shared across all of them. The table below collects only the core concepts that recur throughout the main text and that genuinely decide whether a business lives or dies. Every definition is grounded in "the context of this book"—that is, what each term actually means on the real ledger of a company of one.

Further Reading

The books and communities below are the shared intellectual soil and data sources behind the book's one hundred cases. Read the books first to build the framework, then immerse yourself in the communities to watch live specimens, and finally return to this book's index table for a side-by-side comparison.

Books: four foundational reads

• Company of One — Paul Jarvis. The direct source of this book's title and central thesis: it questions "growth equals success" and argues why staying deliberately small and beautiful is the better path. Read it to understand the value foundation of the Solopreneur.
• The Million-Dollar, One-Person Business — Elaine Pofeldt. A systematic survey of the real paths and categories by which businesses with no employees break a million dollars in revenue, echoing this book's "revenue magnitude distribution" (40 cases at the high / million-dollar tier).
• The Pathless Path — Paul Millerd. A first-person account of leaving big consulting for independent work, filling in the "why leave the system" layer of motivation that sits behind the glossary.
• The Almanack of Naval Ravikant — edited by Eric Jorgenson. The master text for concepts like "leverage," "specific knowledge," and "don't trade time for money." The book's high average leverage score of 8.1 is theoretically rooted here.

Communities: three places to watch live specimens

• Indie Hackers. The headquarters where independent developers publish revenue and post-mortems openly; a great deal of the book's SaaS/AI revenue figures (such as WIP, Park.io, and the FeedbackPanda exit data) come from interviews of this kind.
• Starter Story. Known for structured "from zero to revenue" founder interviews; several of the book's revenue estimates (such as the Long Tail Pro portfolio at roughly $5M/year and Frag Out Flavor at roughly $1.5M ARR) follow its figures.
• MicroConf. The conference and community for bootstrapped SaaS founders, the offline hub for the methodology of "no funding, small and profitable software."

The book's main categories of public data sources

Every number in this book is labeled with its basis, drawn mainly from: ① founder public disclosures (X / blogs / annual reviews, such as levelsio's and Marc Lou's self-reports); ② verification by mainstream business media (CNBC, NYT, Fast Company, The Hustle—for example, ecommemily at roughly $220,300 as vetted by CNBC); ③ third-party revenue-database estimates (GetLatka, Starter Story, Empire Flippers Scoreboard); ④ transaction / exit platforms (Acquire.com sale prices, Wix's $80M acquisition of Base44).

A note: anything marked "estimate," "self-reported," or "run-rate" is an unaudited figure, and cross-source numbers may conflict (for instance, the gap between ChatPDF's roughly $440K ARR per Latka and the roughly $6M ARR cited on the founder's side). When reading the numbers of a company of one, focus on magnitude and trend, not the decimal point.

Conclusion: Scale Is a Choice, Not an Obligation

This book profiled one hundred people. From Designjoy at index 87.0 (Brett Williams, a one-person SaaS earning roughly $4M a year) to Firebean coffee roasting at number 100 (Michael Russo, around $90K a year), revenue spans two orders of magnitude, and the categories cut across all eight types: SaaS, content, e-commerce, AI, physical products, services, and assets. If you remember only one thing after reading it, I hope it is this: not one of these people stayed solo because they couldn't scale up. They stayed solo because they thought it through and chose to.

That distinction is the axis of the entire book. Today, growth is treated as a default obligation: hire, raise money, expand, scale, as if not growing were failure itself. But Pieter Levels proved, with PhotoAI (roughly $1M+ ARR) and his whole product portfolio (around $3.1M+ ARR in 2025), that one person plus a few servers can stand in for an entire company. Mike Perham built Sidekiq to roughly $7M a year and said, "I'm closer to $10 million than I am to $1 million"—he could easily have raised money and expanded, but he didn't. Base44's Maor Shlomo single-handedly built a product that Wix acquired for $80M in cash. For these people, scale is a checkbox in the options menu, not a whip cracking overhead.

Four Weapons, Tying the Book Together

What underwrites this kind of choice are the four words that recur throughout the book. They aren't slogans; they're statistical facts drawn from these one hundred cases. Across the four scoring dimensions, leverage averaged 8.1—the highest of them all.

• Leverage—code, content, media, and capital that let a single unit of labor be reused countless times. Forty-nine cases scored 9 or higher on leverage. Marc Lou sold a single ShipFast codebase to tens of thousands of developers (self-reported at $1,032,000 in 2025); Easlo turned Notion templates—the quintessential "make once, sell ten thousand times" asset—into roughly $779K.
• Distribution—scarcer than the product itself is the ability to be seen. Justin Welsh has earned a cumulative $12.5M+ off the back of content alone; levelsio's "building in public" is itself a distribution engine. Without distribution, even the best product is just a good thing sitting in the dark.
• Compounding—time is on your side. Ben Thompson's Stratechery rolled a decade of subscriptions into roughly $5M+ a year; Ruan Yifeng's newsletter has reached its 399th issue. Compounding isn't sexy, but it's the only force that doesn't require you to start over every single day.
• The courage not to grow—the most counterintuitive, and the rarest. Carrd's AJ and ChatPDF's Mathis both held their hands still at the very moment they could have expanded. Keeping the scale small is how you keep the freedom for yourself.

An Honest Closing

I won't lie and tell you this is easy. Across these one hundred people, replicability averaged only 5.4—far below leverage. The implication is blunt: anyone can learn the tools, but Ben Thompson's judgment, the trust Pat Flynn accumulated over a decade (peaking at $167,553 a month), and the taste with which Aimee Smale built Odd Muse to roughly £22.5M in annual sales—none of that can be copied. A one-person company is not a shortcut; it loads all the weight onto a single person—and hands all the freedom back to that same person. It's an honest trade.

Big companies hedge risk with headcount; one-person companies hedge scale with leverage. The former is buying certainty, the latter is buying freedom. Neither is more noble—only one is more like you.

The First Step You Can Take Today

Don't wait until the whole board is figured out. Almost every one of these one hundred people started with something tiny: Nathan Barry wrote an e-book first in 2012 ($145,471); Damon Chen first built a small tool called Testimonial.to. Here is what you could do today—pick one:

1. Write down a single sentence—"For whom can I solve what problem, again and again?" If you can't name a concrete "whom" and "what," don't build a site or write any code yet.
2. Publish your first thing in the open—a post, a newsletter, a small tool that actually works. The compounding of distribution only starts earning interest the first time you're seen.
3. Choose one reusable asset to build, rather than yet another job that trades time for money.

Scale is a choice, not an obligation. What this book hands you is not a map to ten-million-dollar revenue—that map doesn't exist—but a permission: you can be just one person, and do it well, for a long time, and freely. Now, close the book and go do that one tiny thing.