The Creator Built a Business

A YouTube creator with 250,000 subscribers runs what looks like a business. Studio in a spare bedroom. A $1,200 camera, a $300 microphone, editing software at $23 a month. An editor on contract at $350 a video. Four uploads a week. Brand deals in the inbox. A business plan no one wrote down but everyone can see.

YouTube takes 45 percent of the ad revenue. The creator keeps 55 percent. That is the stated split, and it is accurate, and it is not the number that matters.

The creator spent $3,000 last month on production. YouTube spent nothing. The platform provided distribution, the algorithm, the ad marketplace, and the audience. The creator provided the labor, the equipment, the ideas, and the risk. YouTube's stated take is 45 percent. But when you account for who bears the costs, the creator keeps the work and the platform keeps the business.

The creator does not control the distribution. YouTube's algorithm decides which videos get recommended, which appear in search, which reach the homepage. The creator can optimize for it. The creator cannot set the terms.

The creator does not set the price. YouTube's ad marketplace determines the rates. A finance channel earns $9 to $11 per thousand views. A gaming channel earns $2 to $6. Same effort. Different category. The creator has no say in either number.

The creator does not own the customer. YouTube owns the subscriber list. The creator cannot email their audience. Cannot export the names. Cannot move them to another platform. When YouTube launched a pilot program for reinstating banned creators in 2025, the terms were explicit: neither old subscriber lists nor previous content would carry over to new channels. Even on the same platform, you cannot recover what you built.

And the creator cannot leave with the business. The channel is a tenancy. The building belongs to someone else.

YouTube built the platform. The algorithm, the ad marketplace, the content delivery network that serves two billion users a month. That investment is real. It cost billions. And it earned the platform something specific: ownership.

The creator also invested something real. Time, equipment, ideas, years of output. That investment earned the creator something too. The question the creator never asks is what.

The real number

Three platforms dominate three different versions of the same arrangement. YouTube connects creators with advertisers. OnlyFans connects creators with subscribers. Uber connects drivers with passengers. Each publishes a revenue split.

The stated percentage is the recruitment number. It is what the platform publishes and what the worker sees when they sign up. It is accurate in the narrow sense that the platform does take that percentage of the gross transaction. It is misleading in every other sense.

YouTube takes its 45 percent before the creator spends a dollar on production. The creator's 55 percent is gross, not net. The editing, the equipment, the software, the thumbnails, the studio. All of it comes out of the creator's half. YouTube has no production costs. The creator has all of them.

OnlyFans takes 20 percent and provides no discovery. No algorithm. No promotion. No search results. The creator drives all their own traffic, which means the creator is the marketing department for someone else's platform. B9 Agency, a creator management firm, estimated that on $5,000 in gross monthly earnings, a creator's real take-home is approximately $2,800 after the platform cut, taxes, content protection, and the cost of customer acquisition. That is 56 cents of every dollar. The other 44 cents go to the platform, the government, and the work of finding customers for a business the creator does not own.

Uber is the clearest case because the costs are the most tangible. The driver pays for the car, the gas, the insurance, the maintenance, and the depreciation. Uber pays for the app.

In May 2025, the National Employment Law Project published a fact sheet documenting what Uber and Lyft actually take. The headline: both companies take around 40 percent on average, and sometimes 65 or 70 percent on individual rides.

The Economic Policy Institute calculated the full picture in 2018. Gross passenger fares: $24.77 per hour. After Uber's commission, vehicle expenses, and self-employment taxes, the driver's net hourly equivalent was $9.21. UC Berkeley's 2024 study found California rideshare drivers earning a median of $5.97 an hour without tips. Below the state minimum wage.

Uber did not start at 40 percent. It started at 10. A promotional rate, designed to recruit drivers. Then it moved to 20, then to 25. In 2022, both Uber and Lyft switched to a system they called "upfront pricing." The name sounds transparent. The mechanism is the opposite. There is no longer a fixed commission. Algorithms set the passenger fare and the driver payment independently, with no disclosed connection between the two. The platform charges passengers more and pays drivers less, and neither side can see the other's number.

The stated percentage is the number you are meant to see. The effective percentage is the one you live on.

The curve

The take is not the entire problem. The distribution is the other half.

In 2023, a team of researchers analyzed 153,000 elite YouTube channels, the largest near-complete sample ever studied. They found that 0.43 percent of channels capture 68 percent of all ad revenue. Eighty-eight percent of creators earn virtually nothing. The study, published in Social Media + Society, described the majority as stuck on an "aspirational curve" performing what the researchers called hope labor.

OnlyFans has a Gini coefficient of 0.83. For comparison, South Africa, the most unequal country in the world by income, has a Gini of 0.63. The OnlyFans economy is more unequal than the most unequal nation on earth. Its 4.6 million creators split $5.8 billion in annual payouts. The mean is $1,300 a year. The median is roughly $150 a month. One person, the platform's majority owner, collected $497 million in dividends in 2024.

A 2025 study from University College London measured creator earnings across platforms and found they follow a Pareto distribution with an exponent of approximately 2. That is closer to concentrated capital income than to labor income. The researchers identified a specific mechanism: when a platform's algorithm tilts more attention toward the top, the gains are drawn disproportionately from the creator middle class. The algorithm does not redistribute evenly. It redistributes upward.

Goldman Sachs estimated in March 2025 that only 4 percent of creators earn more than $100,000 a year. They project the professional share will shrink to 2.5 percent by 2030, even as the total creator economy doubles to $480 billion. The market grows. The professional slice shrinks. More people enter. Fewer make it.

MrBeast earned $85 million. He is YouTube's best advertisement for YouTube.

The mechanism is structural. Every new creator who joins a platform increases the supply of content and competes for the same pool of attention and ad dollars. The platform's revenue grows because total engagement grows. The per-creator share shrinks because the denominator grows faster than the numerator. The platform profits from the influx regardless of whether any individual creator succeeds. The success stories serve as free recruiting for the next wave of supply.

The framing

YouTube, OnlyFans, and Uber run structurally identical businesses. A platform provides access to a marketplace. A worker provides labor and bears the operating costs. The platform controls the distribution, the pricing, and the customer relationship. The worker cannot leave with what they built. The economics converge. The cultural stories could not be more different.

An Uber driver is called a gig worker. The connotation is low status. The public response is sympathy and regulation. California spent years on AB5 and Prop 22. The gig companies spent $205 million to pass Prop 22, the most expensive ballot measure in the state's history at the time. The UK Supreme Court ruled unanimously in 2021 that Uber drivers are workers, not independent contractors. The Amsterdam Court of Appeal found that Uber's claim of "human in the loop" review of algorithmic firings was "not much more than a purely symbolic act."

Bernard Moses drove for Uber in Chicago for nearly ten years. Twenty thousand rides. A 4.99 rating out of 5. One unverified customer complaint and he was locked out of his account. He called the support number listed in the app. He reached an answering machine. He called multiple times a day for weeks. A month later, he received two missed calls from someone claiming to be an Uber employee. He called back. No one answered.

An OnlyFans creator is called empowered. The platform is mostly sexual content, and everyone knows it. The cultural story is direct: cut out the middleman, keep 80 percent, and what you do is your business. Why not. In August 2021, OnlyFans answered the question. The platform announced a ban on sexually explicit content, the content that built it. The CEO told the Financial Times the reason was banking pressure: "The short answer is banks." Creators who had built their entire income on the platform had no contractual protection. No equity. No governance rights. No seat at the table where the decision was made. The ban was reversed after six days. The subscribers who left during those six days did not all come back. The announcement itself was the damage.

A YouTube creator is called an entrepreneur. The connotation is aspirational. MrBeast is profiled in business magazines. Creator houses are featured in architecture publications. YouTube hosts an annual "Made on YouTube" event to celebrate its creators. In March 2017, the Wall Street Journal reported that major brand ads were running alongside extremist content. Over 250 advertisers pulled their spending. YouTube's response was mass algorithmic demonetization. Comedy CPMs fell 37 percent. Gaming fell 29 percent. Zombie Go Boom, an entertainment channel, went from $10,000 to $15,000 a month to as low as $20 a day. They sued YouTube for transparency on the algorithm. The judge dismissed the case. Another creator had been planning to quit his day job to go full-time. He watched his income vanish overnight and never went through with it. The platform made the decision. The creators bore the cost. The word for what happened to them is not "entrepreneurship."

YouTube takes the biggest cut and gets the least criticism.

The framing does not correlate with the economics. It correlates with the perceived status of the work. Driving a car is blue-collar. Creating sexual content is stigmatized. Making videos is creative. The creative label buys the platform an immunity that the other two do not receive.

The Uber driver knows something is wrong. The regulatory system agrees. Classification battles, minimum wage rulings, court orders. These are the responses of a system that recognizes the worker as a worker.

The YouTube creator believes they are building a business. That belief is the platform's greatest product.

The precedent

The structure is not new, only the language.

In the cotton South after the Civil War, the standard arrangement between landowner and sharecropper was a 50/50 split on the crop. The sharecropper provided the labor. The landowner provided the land. Fifty-fifty sounds fair.

The sharecropper also bought supplies on credit from the landowner's store, at prices marked up 40 to 70 percent above cash rates. Seed, tools, food, clothing. All on credit, all at inflated prices, all secured against the next harvest. Roger Ransom and Richard Sutch documented this in One Kind of Freedom, the foundational economic history of emancipation and its aftermath. By the time the crop came in and the debts were settled, many sharecroppers owed more than their share was worth. The effective take, after the furnishing costs, was often zero. Sometimes negative. The sharecropper started the next year deeper in debt than the year before.

The landowner owned the land, the store, the accounting books, and the relationship with the buyer. The sharecropper owned their labor. Nothing else. The 50/50 split was the stated percentage. The furnishing costs were the effective one.

Nicholas Carr coined the term "digital sharecropping" in 2006. "One of the fundamental economic characteristics of Web 2.0," he wrote, "is the distribution of production into the hands of the many and the concentration of the economic rewards into the hands of the few."

The parallel is structural, not experiential. Nobody is comparing a YouTube channel to the violence and racial subjugation of the postbellum South. The comparison is narrower: the economic mechanism. A stated split that attracts labor. Shifted costs that capture the margin. An accounting system the worker cannot audit. An exit the worker cannot afford.

The mechanism predates sharecropping by centuries. The putting-out system, common across Europe from the late Middle Ages through the Industrial Revolution, worked the same way. A merchant distributed raw materials to workers who produced goods at home. The merchant owned the materials, the customer, and the pricing. The worker owned a loom and their hands. Payment was per piece. No minimum. No floor. No guarantee of work next month. In 2020, sociologist Bartosz Mika published a paper arguing that platform work is "strikingly congruous" with this 18th-century system. The workers are formally free. They bear their own costs. They own nothing of the output.

Franchise economics follow the same pattern. The franchisee bears the build-out, the inventory, the labor, the rent, and the insurance. The franchisor owns the brand, the system, and the customer acquisition model. When the franchisee leaves, a post-term non-compete prevents them from operating a competing business. The FTC identified this in 2024 as one of the top twelve franchisee concerns. You build the local business with your capital. The franchisor retains ownership of everything that makes it possible.

The structural pattern has been consistent for four centuries. The worker provides labor and bears costs. The owner provides access and controls the customer. The owner controls the accounting. The worker cannot leave with what they built. The nominal split obscures the effective take. Only the language changes.

The sharecropper was called a sharecropper. The pieceworker was called a pieceworker. The gig worker is called a gig worker. The YouTuber is called an entrepreneur.

This is not a critique of platforms.

YouTube created something real. A person in Lagos with a phone can reach a global audience. An animator in Manila can earn more from a YouTube channel than from a local studio. A musician in Nairobi can distribute a record without a label. The opportunity is genuine. The access is unprecedented. The tools work.

The observation is narrower. The opportunity exists inside a structure the worker does not own. The distribution belongs to the platform. The pricing belongs to the platform. The customer belongs to the platform. The worker can be removed at any time, by an algorithm, with no hearing, no evidence, and no recourse. The audience stays behind.

Three academic concepts name what happens at the bottom of this structure. Hope labor: work without pay, in the expectation of future reward. Aspirational labor: creative work invested in building visibility, in the hope that the visibility converts to income. Venture labor: the shift of economic risk from capital to the individual worker. All three describe speculative bets with time. The platform collects its margin whether the bet pays off or not.

The most successful labor reclassification in modern history did not happen through legislation. It did not happen through a court ruling or a corporate policy or a union negotiation. It happened through language. The platform called the worker an entrepreneur. The worker agreed. The economics did not change. The framing did. And the framing was enough.

YouTube takes 45 percent and the worker celebrates the opportunity. Uber takes 40 percent and the worker demands regulation. The structure is the same. One is called a business. The other is called a gig. The difference is in the story the economics tell about themselves, not in the economics.

The creator built a business. The platform gave them the tools. The platform kept the business.