How to spot problematic clients before you write a proposal

Many studios vet clients after the fact. The project slips, the feedback turns into a committee, the final invoice sits unpaid, and only then does someone say the signs were there from the first call. They usually were. The problem is that nobody wrote them down, and nobody decided in advance what they would do about them.

A proposal is expensive to write well. It takes research, a scope, a timeline, pricing and often a few hours of design. Spending that effort on a client who was never going to sign, or who will cost you more than they pay, is the most avoidable loss in a creative business. Vetting belongs before the proposal, because that is the last moment when walking away costs you nothing.

The method here has three steps. Start by running a short intake conversation built to surface specific signals. Then score what you heard against a simple rubric instead of trusting your mood that day. Finally, choose a response that matches the score. A middling score is rarely a reason to say no; it is usually a reason to write a different proposal.

Forrester's State of Business Buying research found that 86% of B2B purchases stall during the buying process.

Even good clients are slow to buy, so a stall on its own is normal. A stall with no budget, no named approver and no clear goal is a pattern worth scoring.

A red flag only helps if you can ask for it. Bad clients rarely announce themselves; they answer ordinary questions in revealing ways. The trick is to ask questions that leave room for the revealing answer, then write it down before you have time to talk yourself out of it.

The table below pairs each signal with the question that surfaces it and a weight. Weight matters because signals are not equal. A request for free spec work tells you far more about the relationship ahead than a tight deadline does. Score each signal from nothing (no concern) up to a heavy weight (a clear problem), and add the totals at the end of the call.

Forrester's State of Business Buying research reports that, on average, 13 people within an organization are involved in the buying decision, with 89% of purchases involving more than a single department.

That is why the decision maker question deserves the extra weight. If your contact cannot name who signs, you are writing for an audience you have not met.

Signals that look bad but are not

Some items on popular red-flag lists are noise. An NDA before the first call can simply mean the client needs to show you confidential files. A formal RFP often comes from a well-run organisation with a procurement rule, not from a bad client, even if it means more paperwork for you. A first-time buyer of design needs more guidance, which is a pricing question rather than a warning. An urgent deadline tied to a real launch or event is a constraint, not disrespect.

Score these at zero unless they come with something else: an NDA plus a refusal to share a budget, or an RFP with a scoring sheet that only weighs price. On their own, they describe the client's situation. The heavy signals describe how the client will treat you.

Illustrative example with fictional clients: both briefs ask for a brand identity for a small consumer brand. Each signal scores 0 (no concern) to 3 (clear problem). Heavy signals carry extra weight and can add up to 6 points each.

Brief A, a coffee roaster. On a 15-minute call, the founder gives a budget range, says she and her co-founder sign off together, and describes success as getting stocked by independent grocers. The launch date is tied to a trade fair. She worked with a freelancer before, who moved abroad. She accepts the deposit and the discovery step without fuss.

Brief B, a skincare start-up. The marketing manager will not share a budget ("just tell us what it costs"), says the board will review "at some point", and asks for initial concepts before signing anything so they can compare studios. Success is described as "something high-end". The previous agency "didn't get it". The deadline is in three weeks, with no reason given.

Illustrative score bands for this example: 0 to 5 means proceed with your standard proposal. 6 to 11 means proceed with protections. 12 to 17 means reprice or restructure, for example by selling a paid discovery phase first. 18 and above means decline politely.

Brief A gets a normal proposal. Brief B scores well into the decline band. If the studio wants to keep the door open, the honest offer is a paid discovery phase with a named approver from the board, no free concepts, and a full deposit before any work starts. If the client refuses that, the score was right.

For Brief B, the vague approval chain should worry you more than the budget. A client who cannot say who decides is a client who may never decide.

April Dunford makes the point in her guide to building a sales pitch: companies that stuck with the status quo usually did not decide it was better than the other options.

A score is only useful if it changes what you send. The response options are simple, and you should decide which applies before you open your proposal template.

Decline when the heavy signals stack up and the client has already pushed back on a paid first step. Keep it short and kind. Thank them, say the project is not the right fit for how you work, and if you can, suggest another studio or freelancer who might suit them. Turning work down well protects your reputation; clients talk.

Reprice when the brief is fine but the budget is not. Offer tiered scope: a core option that fits the money they have, a recommended option, and a fuller option. That turns a budget fight into a choice the client controls.

Restructure when the goals are unclear. Sell a paid discovery phase first, with its own short proposal, and quote the main project when you both know what it is.

Proceed with protections when the client is good but a few signals worry you. The list below maps each worry to a line in the proposal.

Data from IPSE's Freelancer Confidence Index survey, covering Q1 2024, found that 32% of freelancers had experienced a delay in payment from a client in the preceding 12 months.

Put these protections where the client will actually read them: next to the scope and the price, not buried in terms at the end. A deposit and milestone payments are far easier to agree before work starts than to chase once an invoice is late.

McKinsey's research on how business customers buy shows that, throughout the decision journey, almost 65 percent of B2B customers will come away from it frustrated by inconsistent experiences.

Vetting also continues after you send. A proposal that lives as a link rather than an attachment can tell you who opened it, whether it was forwarded, and where the reader spent time. Treat that behaviour as late intake data. A proposal that is never opened after a warm call says the urgency was not real. A forward to people you have never heard of says the approval chain is bigger than the client admitted. Long, repeated reads of the pricing section followed by silence usually mean a budget conversation is coming, and you can open it first.

Sometimes the warning signs slip through, or a good client changes after the work starts. You are not alone in a relationship that soured.

Forrester's State of Business Buying research found that 81% of buyers express dissatisfaction with their chosen providers.

Unhappiness on the client side is common, which is why it pays to name the problem precisely rather than tolerate it.

The cleanest reset starts with a document the client can read and approve. If you send proposals as Formlio documents, a change order can be its own short Formlio proposal with the new scope, the price in an interactive quote and an agreement block, so the client approves with Slide-to-Approve or signs with an e-signature before work restarts. You get a notification when it is opened and when it is approved, and version history keeps a record of what changed and when. The same applies at the start: the scope, the named approver, the deposit and the revision cap sit in blocks the client reads, not in an attachment they skim.