The numbers behind a creator business, including the ones nobody posts

The publicly visible examples are selected for being remarkable. That is what makes them worth posting. Planning around them is like planning a career in film around the people who won awards — not wrong exactly, but a poor basis for arithmetic.

The numbers that matter are unglamorous and specific to you: what each activity earns per hour, how reliably it recurs, and how much of it survives costs and tax. None of that is knowable from someone else's post.

What follows is not a set of figures. It is the set of questions whose answers determine whether the business works.

• Client work. Highest hourly rate, entirely non-recurring, capped by your available hours. Reliable while you are doing it and zero the month you stop.
• Products. Low hourly rate while building, then income without further hours. Slow to start, compounds if you keep publishing.
• Subscription. The most stable, the hardest to start, and the most demanding — it commits you to producing indefinitely.
• Platform income. Advertising, affiliate, revenue share. Unpredictable, dependent on decisions you do not control, and generally worth less per hour than any of the above.

Most sustainable creator businesses are client work funding product development, with the products gradually replacing the hours. The order matters: products rarely start fast enough to live on.

For one month, log every hour against an activity. Then divide the income each activity produced by the hours it consumed.

The results are usually uncomfortable in a specific way: the activity that feels most productive is often not the one earning most, and a large share of hours belongs to activities earning nothing directly — posting, replying, admin, learning.

Those unpaid hours are not waste, but they must be counted, because they are the difference between a rate that looks good and a month that does not add up. A day rate that ignores the two unpaid days around it is not a day rate.

Software subscriptions, model credits, storage, hosting, marketplace fees, payment processing, hardware amortisation. Individually small, collectively substantial, and almost never counted when people quote their income.

List them all once with an annual figure next to each. The total is generally larger than expected, and a few items are usually redundant — two tools doing the same job, a subscription for a project that ended.

Then set aside for tax from every payment as it arrives rather than at the end of the year. This is the single most common way otherwise viable creator businesses fail: the money was earned, spent, and then owed.

Take a month off in your head. What still arrives?

If the answer is nothing, every hour of income is bought with an hour of work and the arrangement fails the moment you are ill, busy, or tired. That is not a criticism — it is where nearly everyone starts — but it is worth knowing explicitly rather than discovering.

The goal is not to eliminate client work. It is to have the base of recurring income cover the fixed costs, so that a slow month is inconvenient rather than dangerous. That threshold is far lower than replacing your income entirely, and reaching it changes how it feels to do the work.

Every three months, sit down with actual numbers: income by activity, hours by activity, costs, and what recurs. Not estimates.

The review usually surfaces one thing worth stopping and one worth doing more of, and those two decisions are worth more than any amount of working harder at the current mix.

Keep the history. Comparing four quarters shows a direction, and direction is the only thing that reliably tells you whether the strategy is working — a single good month tells you almost nothing.