Capacity spent badly is capacity gone

Once you have more enquiries than slots, accepting is no longer free. A job taken at the bottom of your range does not just earn less, it removes a slot that a better-paid job would have filled, and that difference is the whole argument.

The hard part is not the arithmetic. It is that declining feels like losing something, and it takes about a month of doing it before your income confirms that it is not.

When selectivity starts paying

Before you are booked out, decline almost nothing. An empty week has no opportunity cost, so a cheap job is genuinely better than no job, and being fussy early is how the cold start gets longer.

The switch happens when you are turning over roughly your sustainable capacity for three or four consecutive weeks. At that point the cost of a cheap job is the difference between it and the job you could not fit, and that number is often larger than the fee itself.

An illustration, with the assumption stated: if your sustainable load is eight jobs a week and two of them are at half your normal rate, you are running at 87 per cent of the income the same hours could produce. Over a year that gap is a month of work performed for nothing. This is the arithmetic behind why volume sometimes beats price and sometimes does not, read from the other end.

The criteria, in order of usefulness

Decide these once, in writing, before an enquiry arrives. A rule applied in the moment is a preference, and preferences lose to the discomfort of saying no.

Price below your floor. Not below your list price - below the floor you set when you worked out the minimum a job can pay once fees come out. Anything under it is a job that loses money once the flat deductions land.

Scope you cannot bound. A brief that will not resolve into a defined deliverable after one clarifying question will produce a revision spiral, and revision hours are unpaid hours.

Material outside what you assess. Taking a job you are not equipped for costs the review, and the review costs several future jobs.

A buyer who has already been difficult. Enquiry behaviour predicts job behaviour with uncomfortable accuracy. Three messages of pressure before booking becomes ten after.

Timing you will resent. A deadline that requires a Sunday is fine at a rush price and corrosive at a normal one.

Everything else is not a reason to decline, including a brief that is merely dull.

The phrasing

Short, immediate, without a justification the buyer can argue with.

"Thanks for asking - that one is not a fit for what I do. Good luck with it."

That is the whole message. Do not explain the price, because explaining invites negotiation; do not apologise twice; do not offer to reconsider.

Where the reason is capacity rather than fit, give the date instead: "I am booked until the 20th - happy to take it then if that works." That converts a decline into a deferred booking about a third of the time, at no cost.

Where the reason is price, name a number rather than refusing. "That scope is a job at my full rate, which is X - happy to start today if that works" is a decline that sometimes gets accepted, and it never reads as rudeness. The mechanics of holding a number under pressure are in negotiating without lowering it, and the speed side is covered by declining politely and quickly.

Buyers, for their part, treat a fast clear no far better than earners assume - the conventions around how commissions open and close are written from their side and contain no expectation that a professional accepts everything.

The first month of doing it

Expect three things.

Your booking count falls and your income does not, which is the point but still feels wrong on the days it happens. You will decline one job you should have taken, and you will find out because the week stayed empty. And you will get faster at reading an enquiry, because you are now reading it against a list rather than against your mood.

Keep a note of what you declined and why. After eight weeks the pattern in that list tells you more about your market than any pricing post can: which format attracts the pressure, which price point attracts the scope creep, which enquiry wording predicts a dispute. Tracking the enquiries that did not convert is the same habit applied to the other side of the ledger.

What this is not

It is not a filter for buyers you find unpleasant to imagine, and it is not a way of pretending to be busier than you are.

It is capacity allocation, and capacity is the only genuinely scarce input in this trade. Precision about your own limits is worth more than precision about anything else you could measure, though the conventions for the things that genuinely can be measured are a useful reminder that "I am full" should mean a number rather than a feeling.

The floor under all of this is that generic work has already been absorbed by automation - what an automated score reliably delivers is the honest account - so the jobs worth protecting your capacity for are the ones that need a person to have actually looked. Those are also the jobs that pay.

Marketplace mechanics affect how visible a decline is: some platforms count response rate, some count acceptance rate, and the two reward opposite behaviour. Outside this trade, Airbnb's help centre shows both at once: it tells hosts to decline a request within 24 hours to protect their response rate, and warns that turning down a lot of requests can affect where a listing shows up in search. Rate Cock's judges page sets out which it uses, and that detail should shape your phrasing more than any general advice does.