There is no leverage in an hour of your attention

Scaling here means raising what each hour earns, not adding hours: growth has three levers - price, format, and the share of your time that is billable - and volume stops being one past a ceiling you will hit inside a year. The product is a named person's attention, and the unit of production is your hour.

This is not a pessimistic framing. It is the reason the trade holds its prices, and it is worth understanding early because most growth advice imported from other kinds of business is actively wrong here.

The ceiling is real and you can compute it

Take the number of hours you will genuinely work in a week, subtract the unbilled third that goes on enquiries, briefs, delivery and admin, and divide by your minutes per job. That is your maximum weekly job count. Multiply by your price and you have your annual ceiling at today's rate card.

The number is usually lower than people assume and it arrives sooner. The arithmetic of a part-time week does this for ten hours; the same method at forty produces a figure that is large enough to be a living and small enough to be finite, which is the whole point.

Once you have it, growth becomes a question with only a few possible answers rather than an open-ended ambition.

The four things that actually move the number

Price. The only lever with no ceiling of its own, and the one people postpone hardest. A rise applies to every future job at zero marginal cost and does not consume any hours. The first rate rise, and when to make it is the decision that most changes an annual figure, and it is usually made too late.

Format mix. Moving your work toward the formats that pay best per minute rather than per job. These are not the same ranking and the difference is frequently large. This requires knowing your take-home per format, which requires having recorded it - tracking take-home by format is the prerequisite for any decision in this category.

Minutes per job. Falls naturally with tenure, then stops. You can push it further with rubrics and templates, and only up to the point where the deliverable becomes generic, at which point you have destroyed the thing you were selling. This lever has a floor and it is nearer than it looks.

Billable share. Compressing the unbilled third. Real, worth doing, and bounded - you cannot get overhead to zero and chasing the last few percent costs more attention than it returns.

Notice that three of the four are bounded and one is not. That asymmetry is the single most important fact about growth in this trade, and it is why the earners who do well are the ones who raised prices rather than the ones who worked more.

Why outsourcing collapses the product

The obvious escape is to have someone else do the jobs while you keep the listing. It fails, and it fails in a specific way rather than a vague one.

The buyer is paying for a named person to have looked. Disclosure research shows how much the source weighs: in Yin, Jia and Wakslak's 2024 PNAS study, recipients felt less heard once they realised a message came from AI, even though AI-written messages judged blind made them feel more heard than human ones. That is not a preference they will trade away for a lower price; in many bookings it is the entire reason a human was commissioned instead of a free automated score. The technical account of what instant scoring already produces is set out on the AI property, and reading it makes the point clearly: if the buyer wanted competent generic output produced by something other than you, it is already available at no cost.

So a ghost-written deliverable competes against free rather than against your rate card. Consistency drops, the reviews change tone within a handful of jobs, and the repeat buyers - who have a baseline and notice - are the first to leave.

There is also an honesty problem that has nothing to do with economics. A listing that says a person will assess your submission and then routes it elsewhere is a misrepresentation, and it is the kind that surfaces eventually.

What growth looks like when it works

Not more jobs. Fewer, better-paid, more specific jobs, delivered inside a week that has not expanded.

Productising a part of the work is the one genuine leverage available: a reusable rubric, a template structure, a public worked example, a written framework that does the generic portion of every deliverable so your minutes go to the parts only you can produce. This is leverage on the overhead rather than on the work, and it is real but bounded. Productising part of the work is the longer treatment.

A waiting list is the other honest signal that you have hit the ceiling, and the correct response to one is a price rise rather than longer hours. Being hard to book is a position with a value attached, which is a different claim from being unavailable.

The ceiling is a feature

A trade that cannot be scaled by anyone also cannot be scaled by a competitor with more capital, which is why the upper part of the band has held while the bottom was erased by automation. Your ceiling is also everyone else's, and that is what keeps a named person's hour worth what it is worth.

The practical instruction is to compute your own number this week, decide which of the four levers you have not pulled, and pull exactly one. The lever that looks most like scaling and mostly is not is another pair of hands - working with a partner or splitting work covers what that does to the product. Two adjacent subjects belong elsewhere: how buyers decide who to commission, and therefore what a higher price has to be justified against, is covered from their side, and the measurement conventions that make a precision-led offer checkable are documented independently.

For how listings, availability and pricing tiers are actually presented on the platform side once you have made the decision, Rate Cock's judges page is the current reference.