A quick no is cheaper than a slow yes

The expensive mistake early on is not turning work away, it is accepting work you should not have and delivering it badly. The fee for one job is recoverable in a week; the review it produces sits on your profile for as long as the profile does.

Declining well is a two-line skill, and the wording matters far less than the speed.

Why speed beats wording

A decline sent in four minutes reads as a professional with a clear scope. The same words two days later read as someone who could not face it and is now retreating.

The buyer you decline immediately often comes back later with something you do want. The one you leave hanging has booked elsewhere and formed an opinion of you on the way.

The one-sentence decline

The whole thing:

Thanks for asking - that one is outside what I do, so I would not be the right person for it. If you ever want a written assessment to a fixed structure, I am here.

An acknowledgement, a clean statement that it is outside your scope, one line naming what you do take.

What to leave out is longer than what to put in. No apology, because an apology invites negotiation. No explanation, because a reason is something to argue with and "outside what I do" is not. No referral to a named person unless you know they want it.

The signals worth declining on

Most are visible in the first message.

The brief does not match anything you offer. Do not stretch into an adjacent format to win a first booking; the stretch is where bad reviews come from.

The scope is open-ended. "Ongoing", "as much as you can". Convert it to a defined job with a number attached, or decline it.

The buyer has negotiated before agreeing anything. Price pressure in the first message predicts price pressure at delivery. Why the cheapest end of the market attracts the most demanding buyers covers the selection effect underneath this.

They want to move off the platform. You lose the payment protection, the record and the review, and you keep all the risk.

You are already full. The healthiest reason on this list, and the one ignored until a week collapses.

What a decline costs, honestly

It costs one fee. Someone with two bookings a month feels it far more sharply than someone with twenty.

The trade is still right, for a reason specific to a reviewed marketplace. Your rating is an average over few jobs early on, so one poor review moves it further than any later one will, and the visibility loss compounds through the platform's ranking.

If declining is constant rather than occasional, the problem is upstream in the listing. A first line naming the format precisely turns most of these enquiries away before they are sent, and the piece on the opening line that does that filtering is the cheaper fix.

The decline after you have said yes

Different situation, same principle. Say so immediately, return any fee taken, and do not deliver a half-version as a compromise. A refund also beats a dispute: Stripe's dispute guidance notes that card networks count the disputes a seller receives, not how many it wins.

A clean exit costs one job; a reluctant delivery costs the review, and the review costs several.

Buyers are less offended by this than earners expect, because the norms around withdrawing from a commission are better established than they look from the earner's side - the buyer-side account of commission etiquette says so plainly. Two categories deserve more care. Requests involving identifying somebody, or holding material you should not, are a privacy problem before they are a scope problem, and the privacy side of that question is where the reasoning lives. Requests really asking for an automated number are not a decline at all: name the tools that already produce one and it ends amicably in one line.

Keep a count of declines alongside your bookings. A month with none usually means your scope is too vague rather than your luck too good, and the judge listings on Rate Cock show in public what a defensible scope looks like.