Studio Queues Team · Last updated June 24, 2026
Learn how to manage difficult clients in art commissions without losing your mind. Scripts and strategies for demanding, disrespectful, or abusive behavior.
A client has been a problem.
They're rude in messages. Pushy on the timeline. Sending updates at 2 AM asking if it's done yet. You're three weeks in, you dread opening your inbox, and you've started avoiding work on their commission because the next message is going to be more of the same.
You don't know whether to deliver fast and exit, refund and walk, or set a hard boundary and risk them posting about you on Twitter.
This article is the framework for difficult-client situations. The scripts that work. The math on deliver-fast versus refund-and-walk. How to handle the public-complaint threat (it's almost always a bluff, and even when it isn't, your record speaks louder than theirs). The structural prevention that keeps the next one from getting this bad.
Not all difficult clients are the same. The right move depends on which kind.
Demanding but engaged. They're pushy, they message a lot, they have strong opinions on revisions. They're also paying on time, signing off on milestones, and not crossing lines. This client isn't actually a problem, they're just high-touch. The fix is communication structure (defined update cadence, scoped revisions) rather than ending the project.
Boundary-testing. They're asking for things outside scope, pushing on prices, trying to negotiate the deal mid-project. The fix is the scope creep response and consistent enforcement of the deal you agreed to. Hold the line and the testing usually stops.
Disrespectful but not abusive. Messages are dismissive, condescending, or curt. They're not crossing into abuse, but the tone is consistently unpleasant. The fix is professionalism on your side and a quiet decision not to work with them again after the current project.
Abusive. Hostile, threatening, sexually inappropriate, or otherwise crossing lines you shouldn't have to tolerate. The fix is ending the working relationship as cleanly as possible and protecting yourself with documentation.
These are different situations with different answers. Lumping them together is part of why difficult-client management goes badly. Identify which one you're in before deciding the response.
A lot of "difficult client" friction comes from unstructured communication. The client messages when they feel like it, the artist responds when they feel like it, expectations on both sides drift, and irritation compounds.
A defined cadence fixes a lot of this:
Updates on a schedule, not on demand. Tell the client at intake when they should expect updates. "I'll send a sketch by [date], lineart by [date], final by [date]. If anything changes, I'll let you know early." This sets the expectation that updates come on those triggers, not in response to every "any progress?" message.
Single channel for project communication. Pick one place where commission conversations happen. If it's email, all project conversation goes to email. If it's an in-platform message thread, it lives there. Don't let the conversation scatter across DMs, Discord, and email, because that's where things get lost and the client's irritation grows.
Response time expectations set at booking. "I respond to commission messages within 48 hours on weekdays" is a fair professional standard. The client doesn't expect a same-hour reply, you don't feel guilty about not providing one.
For demanding-but-engaged clients, the structure usually resolves things. The client was anxious because the project felt opaque. Updates on a schedule reduce the anxiety, and the rude tone often softens.
When a client crosses a line and you need to address it, the script that works is direct, professional, and short.
For tone issues:
"I want to flag the tone in your recent messages. I'm committed to delivering your commission, and I'd like to keep our communication professional. Can we reset?"
This calls out the behavior, reaffirms your commitment to the work, and offers a path forward. It's harder to be hostile in response to this than to escalation.
For timeline pressure:
"The timeline I gave you at booking is the timeline I'm working to. I'll send the next update on [date] as planned. If something changes, I'll let you know early."
Same pattern. Reaffirms what was agreed, gives the client a date to anchor on, doesn't promise anything that wasn't already promised.
For unreasonable demands:
"That's outside the scope we agreed to at booking. I can quote it as an addition if you'd like, or we can keep the project at the original scope."
The same phrase from the scope creep article. Works for tone-driven demands as well as scope expansions.
For boundary crossings that you're not ending the project over:
"I need to ask you to keep our communication focused on the project. [Specific behavior] isn't something I'll continue to engage with."
This names the issue specifically without being aggressive. The client can either reset or escalate, and either response gives you useful information.
In all of these, you're not arguing. You're stating the position and the expectation. Arguing invites a debate. Stating invites a yes or no.
For clients you've decided not to work with again but who haven't crossed into "refund and walk" territory, the right move is usually deliver-fast-and-exit.
The reasoning: refunding and walking creates a confrontation. Delivering fast ends the relationship on terms where you've completed your obligation. The client gets the work, you get the rest of the payment, and you decline future bookings without drama.
What this looks like:
The client probably doesn't realize they're being managed out. That's fine. The goal isn't punishment, it's exit without escalation.
A smaller category. The triggers:
In these cases, the math on continuing is worse than ending. The hours you'd spend completing the work, plus the emotional cost of the messages, outweigh the remaining payment.
The script:
"After reviewing the project, I've decided not to continue. I'm refunding the [unworked portion of the deposit / milestone payments that weren't earned]. The work I've completed is yours to use as you wish. I won't be available for future work."
Refund what you owe, keep what was earned, end the project, document everything. Save the message thread (or platform records) in case the client tries to dispute the refund or escalate publicly.
If the client owes you payment that hasn't been made (you delivered work that wasn't paid for), the calculus is different. Usually you write off the loss and end the relationship anyway. Small claims is available for larger amounts but rarely worth the time for typical commission values.
A small percentage of difficult clients escalate to "I'm going to post about this." Sometimes they actually do. The fear of this prevents many artists from holding boundaries.
A few honest observations:
Most threats are bluffs. The client is trying to scare you into giving them what they want. If they were actually going to post, they probably wouldn't telegraph it first.
The clients who actually post often discredit themselves. A long screed about an artist that includes details the writer is clearly wrong about, or that exposes bad-faith behavior on the writer's side, doesn't land the way the client thinks it will. People reading it can usually see who's the unreasonable party.
Your record speaks louder than theirs. If you have a body of work and a history of normal client relationships, one disgruntled customer doesn't define your reputation. The audience that knows your work judges based on your work, not a Twitter thread.
You have your own record of the conversation. If you've been running messaging through a platform that timestamps everything, or have email archives of the conversation, you can produce the actual record if the dispute goes public. The platform records on Studio Queues and similar tools mean the timeline isn't just your word against theirs.
Don't change your behavior to prevent a hypothetical Twitter thread. Behave professionally, hold your boundaries, document the interactions, and if the post happens, respond once with the facts (or don't respond at all) and move on.
Most of the worst difficult-client situations could have been prevented at the intake stage. The red flags were there. The structure to filter them wasn't.
The protective layers:
A proper intake form that surfaces problem clients before you've replied. Vague briefs, no references, urgency without budget, refusal to use the form at all. The red flags article covers the signals specifically.
A click-to-accept TOS that defines payment, scope, revisions, and behavior expectations. When the client crosses a line later, the agreement is the reference point, not a vibes-based argument.
Milestone payments so the financial exposure on any difficult client is bounded. If things go wrong at week three, you've been paid for the milestones up to that point, not just deposit.
In-platform messaging that creates a documented timeline. Discord screenshots are weak evidence in disputes. Platform records with timestamps are stronger.
None of these layers eliminates difficult clients entirely. They make the worst cases rarer and the difficult cases more manageable.
You will have difficult clients. Every working artist does. The skill isn't avoiding them, it's recognizing the category, applying the right response, and protecting yourself when the relationship degrades.
The communication structure handles most cases. The scripts handle the ones that need direct addressing. Deliver-and-exit handles the ones you don't want to work with again but who don't cross lines. Refund-and-walk handles the ones who do.
The Twitter threat is mostly bluff. Your record is mostly fine. The next client probably won't be like this one.
When it gets hard, the rule that's saved me is: the goal isn't to win the relationship, it's to complete the project (or exit) cleanly. Stop trying to manage the client's emotions. Manage your own behavior, document the interactions, and finish or end as the situation calls for.
You will lose a few. You'll get better at spotting the signs at intake. The bad ones will get rarer.
