A creator income plan becomes more useful when it distinguishes three separate problems: getting discovered, getting remembered and getting paid.

Jack Conte’s experience before starting Patreon illustrates why. In the company’s own founding account, millions of YouTube viewers translated into only hundreds of dollars. Conte and Sam Yam founded Patreon in 2013. That history is a case study in the gap between attention and support, not proof that one platform solves every creator’s economics.

Separate three kinds of evidence

Discovery: people encounter the work. Impressions, views and first-time visits can help describe this. They don’t establish loyalty or buying intent.

Return: people choose the work again. Repeat readership, recurring listeners and substantive replies can be useful signals. Compare like periods and account for what your analytics actually measure; an email open is an imperfect proxy for attention.

Support: someone makes a concrete commitment. That might be a purchase, a ticket, a donation or a subscription. Each has different costs and obligations. A request for a product is useful evidence, but still isn’t a paid order.

The practical mistake is using a discovery metric as the answer to all three questions. A viral post can be successful distribution while leaving the business model untested.

Build around the relationship you can sustain

Kevin Kelly’s 1,000 True Fans essay describes a model built on direct support from a smaller audience. The exact number isn’t universal: spending, retained earnings and the size of the team change what is needed.

For planning, separate revenue from what remains after delivery. A print seller must consider printing, packaging, payment fees, shipping problems and replacements. A digital product may have smaller reproduction costs but still requires creation, maintenance and support. Membership adds a continuing commitment.

This is business planning, not a revenue forecast. There is no reliable conversion rate implied by a follower count alone.

Match the offer to demonstrated demand

Consider a hypothetical illustrator with a popular frog character. A shareable drawing, a greeting card and a recurring community are three different products. Evidence for one doesn’t prove demand for the others.

If several readers ask for a card, a clearly scoped card offer tests a narrower question than launching an ongoing membership. If people mainly want to know when the next drawing appears, a voluntary email subscription may be the right connection first.

Write down what was actually requested, what remains an assumption and what would count as a useful result. “A few people liked the idea” and “people paid the stated price” belong in different columns.

Run a small relationship audit

Review a manageable sample of recent releases. Record how people found each piece, what prompted return visits, and any specific requests. Don’t compare raw totals across very different formats as though they represent equal attention.

Choose one next step that matches the strongest evidence. Put it where an interested person can find it, with a plain description of what they get. Keep the free work complete; the next step should add value rather than punish people for not buying.

For a paid experiment, set a delivery limit and estimate your time before accepting orders. Define the smallest result that makes the experiment worth continuing. A finite offer makes it easier to learn without accidentally promising a second full-time job.

Direct access has limits

Patreon’s February 2025 State of Create commentary describes concern about creators reaching their followers. Treat it as company-sponsored context, not an independent comparison of every business model.

A newsletter still depends on delivery systems. A membership still depends on a payment platform and on people remaining satisfied. Direct relationships can improve your options without eliminating intermediaries, churn or workload.

The aim is a system you can understand: how people arrive, why they return, what they value and what serving them actually costs. Growth becomes more useful when you know which of those questions it is meant to answer.