Saying no to paid work feels wrong when you're building a freelance practice. It took me a while to notice that the projects I declined cost me nothing, and one I should have declined cost me everything.

Three things make me walk.

The budget doesn't match the ask

Someone asked me for a large system with a $100 budget.

The instinct is to read that as a client who can't afford me. It's not — it's a client who has no idea what they're asking for. Nobody who understands that a multi-role system is two months of work offers a hundred dollars for it. The number isn't a negotiating position, it's a symptom.

Which matters because that misunderstanding doesn't stay in the pricing conversation. It shows up as surprise at the timeline, surprise at what's out of scope, surprise that a change costs money. You'd be arguing about the same gap for the entire project, just wearing different clothes each time.

I don't counter-offer on these any more. If someone's estimate of the work is off by an order of magnitude, the number isn't the problem.

The timeline is already impossible

Same shape, different axis. Someone wants in three weeks what takes three months.

Sometimes that's a real deadline they can't move, and then it's an honest conversation about cutting scope to fit — that version is fine and I'll take it.

The one to avoid is when the deadline came from nowhere and there's no willingness to trade anything for it. That's not a constraint, it's an expectation that you'll absorb the difference by working nights.

Something's off on the first call

The least defensible criterion and the most reliable.

Not rudeness. Just: they won't answer a direct question about scope. They mention a previous developer in a way that makes you wonder. They're vague about who decides. They push on the deposit before you've discussed anything else.

None of those is disqualifying on its own. But you're about to spend two months in regular contact with this person, and the first conversation is the only sample you get. If it's already effortful, it doesn't get easier once money is involved.

I used to talk myself out of this one, because it isn't a reason. It's a pattern you've seen before and can't fully articulate. That doesn't make it wrong.

What this actually costs

Nothing, mostly.

The projects I've declined would each have taken two months I then spent on work that paid. Declining is cheap. It only feels expensive because the alternative is hypothetical and the empty calendar is real.

The one that was genuinely expensive was a project I took. I've written about that separately — the short version is that the estimate was fine and the terms weren't, and there were things I could see in the first week if I'd been looking.

My take

You will get better at pricing slowly. You can get better at declining immediately.

And the criterion that matters isn't whether the client has money. It's whether they understand what they're buying. Those are different, and confusing them is how you end up two months into something that was never going to work.