By the time an invoice goes unpaid, the useful moment to have noticed something was weeks earlier.
Most difficult clients do not turn difficult partway through. They arrive that way, and the signs are usually visible before any work is agreed, in things that feel too small to act on at the time. A vague brief. A push for a discount before the scope exists. A slightly odd insistence on starting immediately.
None of those things alone means much. Together, and ignored, they explain a lot of the write-ups on when a client won’t pay.
What actually predicts trouble?
Not rudeness. Rudeness is easy to notice and rarely the real problem.
The better predictors are structural, and they show up in how a client talks about the work before it exists.
Vagueness that survives a direct question. If you ask what “done” looks like and the answer stays general, that is not a communication style. It is a scope that has not been decided, and a scope nobody has decided is a scope that changes after you have started, at your expense.
Urgency with no reason behind it. “Can you start today” is sometimes genuine. It is also the standard shape of a client who has already gone through other freelancers and needs someone in the gap immediately. Ask why the timeline is tight. A real answer is fine. No answer, or a slightly irritated one, is information.
Negotiating the price before there is a price. A client who pushes back on a number you have not given yet, in the form of “what’s your best rate”, before describing the job properly, is telling you the budget is the constraint and the work will bend to fit it. That is workable if you go in with your eyes open. It is not workable if you assume the scope is fixed and the price is not.
Does asking about budget tell you anything?
Yes, both in what they say and how quickly.
A client with a real budget can usually give you a range, even a rough one, because someone has thought about what this is worth to them. “We haven’t really got a number, what would you charge” from a business, rather than an individual with no experience buying this kind of work, often means the number has not been decided because nobody wants to be the one who said it first.
You do not need to treat this as disqualifying. You do need to treat it as a reason to be the one who proposes a structure, rather than working to a budget that turns out to have been invented on the call.
What does the deposit conversation actually test?
More than the deposit itself.
A deposit is ordinary. See deposits and progress payments: how much to ask for for how to size one and what to put in writing around it. What matters here is not the money, it is the reaction.
A client who has worked with freelancers before treats a deposit request as unremarkable, because it is. A client who reacts as though you have asked for something unreasonable, for a normal deposit on a normal job, is telling you something about how they see the relationship. Someone who resists paying a fraction of the job before it starts is unlikely to become more comfortable paying the rest of it after.
This is not a trick question with a right answer that gets you the job. It is closer to a credit check that happens to double as the first invoice.
Should a new client’s own paperwork worry you?
Sometimes, and it is worth a look before you agree to anything.
A business with no website, no public trace of the people involved, and a request to be paid to a personal account rather than a business one is not automatically a problem. Plenty of legitimate small operations look exactly like that in their first year. But it removes some of the ordinary checks a longer track record would give you, so it is worth being more careful about the things you can still control: a clearer written scope, a larger deposit relative to the job, and a payment schedule with more, smaller steps rather than one large one at the end.
The absence of a reputation is not evidence of anything. It just means you cannot use someone else’s experience instead of your own judgement.
Is a client who has already fired several freelancers a warning?
Often, though it depends on why.
“Our last person just disappeared” happens, and freelancers genuinely do vanish on clients. “We’ve been through three designers this year and none of them understood the brief” is a different story, told from one side. It might be true. It might also mean the brief itself is the problem, and you are being offered the chance to be freelancer number four.
Ask what happened, specifically, rather than accepting the summary. “They stopped communicating” and “we kept changing our minds after each round” are both real answers you might hear, and only one of them is actually about the freelancer.
Can you check anything concrete before agreeing?
A few things, and they take minutes rather than hours.
Look at whether the business actually exists in the way it presents itself. A quick search for the company name alongside “review” or “complaint” surfaces real patterns more often than you would expect. If you are being hired by a person rather than a registered business, a normal social media presence is a low bar that a genuine client clears easily.
If a mutual contact exists, a two line message asking “how did that go” costs you nothing and occasionally saves you a great deal.
None of this is about building a dossier on every enquiry. It is about spending five minutes on jobs that are large enough, or unusual enough, that five minutes is clearly worth it.
Should you ever turn down a client you need the work from?
Yes, and this is the hardest part to act on rather than agree with.
A quiet month makes every enquiry feel necessary. That is exactly the condition under which a bad client is most likely to get through, because the incentive to notice the warning signs is lowest precisely when you can least afford the consequences of missing them.
The honest test is not “can we use this money.” It is “if this goes exactly as the warning signs suggest, can the business absorb it.” A slow-paying client who eventually pays is a cash flow problem. A client who disputes the invoice after the work is delivered, having never agreed a clear scope, is a different and larger problem, and the second one is the one the warning signs above are mostly pointing at.
How do you say no without damaging anything?
Plainly, and without a detailed explanation you do not owe anyone.
“This doesn’t sound like the right fit for us” is a complete sentence. You do not need to list the reasons, and doing so rarely helps: it either sounds like an accusation or invites a negotiation about the very things that concerned you.
If the work itself is appealing and only the terms are the issue, it is worth proposing the version you would actually accept rather than declining outright. A larger deposit, a shorter first stage, or a smaller opening engagement can turn a client you are unsure about into one you are testing deliberately, on your terms, rather than hoping for the best.
A pattern is worth more than a single answer
No individual question above disqualifies a client on its own. People have odd first calls for ordinary reasons, and treating every hesitation as a red flag will cost you clients who would have been fine.
What is worth acting on is the pattern: vagueness about scope, resistance to a normal deposit, a rushed timeline with no explanation, and a history of freelancer relationships that all ended the same way. One of those is a data point. Three of them, on the same enquiry, is a client telling you exactly what working together will be like, before you have agreed to find out.
Keep a short note on how a new client actually behaved once the job is done, win or lose. The pattern that matters is not visible in any single job. It is visible across several, and Ledger keeps a client’s full history attached to their record, so what happened eighteen months ago does not depend on you remembering it correctly.
For what to put in place once you do say yes, see getting clients to pay on time.