Solo founders need to fire customers more often than they do. Holding on to bad customers is a hidden tax on your business and your sanity — one that compounds quietly until you can’t see why you’re burnt out.
Methodology. This guide draws on customer-firing patterns described by Mike Michalowicz in The Pumpkin Plan and Marc Benioff’s “stay friends” quote from Salesforce’s blog on the topic. We have fired customers from each of our own products. How we research.
Every solo founder eventually has the experience: a customer who consumes 5x the support of a normal account, complains about everything, and pays the same $29/month as the customer who’s never sent an email. Most founders endure it. The right answer is to fire them. The hard part is figuring out which customers actually meet the bar — and how to do it without burning the relationship beyond repair.
Use the following as a diagnostic checklist. If two or more are clearly true, the customer is costing you more than they’re worth. If three or more are true, you should send the offboarding email this week. The order matters: signal 1 is the most common and the most expensive.
You can usually feel this before you measure it. The customer’s name in your inbox triggers a sigh. They’re responsible for 30% of your support tickets while paying 1% of your revenue. The math fails immediately, and worse, the time they consume isn’t replaceable — it’s the time you’d otherwise spend acquiring better customers.
Some customers want a different product. Not a missing feature on your existing product — an entirely different shape of thing. They’ll lobby for changes that would alienate the core audience, or insist on a one-off integration that’s economically irrational to maintain. The polite version is to say no. The honest version is that they’d be happier elsewhere.
Some customers are simply unhappy with the product no matter what you do. They report it in surveys, they vent in support tickets, they leave it as 1-star reviews on G2. The signal is a pattern, not a single bad day. If three feedback events in a row are negative and the underlying complaint is something you’ll never build for, the kindest move is to release them.
This is the unambiguous one. A customer who calls you names, threatens chargebacks at the first hint of friction, or demeans you in support tickets is not a customer worth keeping. There is no MRR figure that justifies it. Fire them the same day and document the messages in case the chargeback dispute lands later.
If a customer has charged back a payment after using the product, or repeatedly fails to pay invoices on the agreed terms, the relationship is broken. Chargebacks ding your processor reputation and put your account at risk. The threshold is once for chargebacks; twice for late payments without communication. Beyond that, the customer is teaching you what kind of customer they are.
The reason customer-firing is under-used by solo founders is that the cost is invisible until you remove it. You don’t notice the 6 hours a week you spent managing one bad account until that 6 hours appears on your calendar as “build time.” You don’t notice the dread you felt opening the inbox until the dread is gone. The founders we know who’ve fired their first bad customer almost always say the same thing: “I should have done it six months ago.”
“The wrong customer can prevent you from serving the right customer. You owe it to your business to politely show them the door.”
— Mike Michalowicz, The Pumpkin PlanDone well, customer-firing leaves both sides intact. The customer gets a clean exit and a refund; you get your time back. Done badly, it ends with a 1-star review and a chargeback dispute. The script below is what we’ve used — concrete and direct. Adjust the tone, but don’t soften the structure.
Subject: Closing your account on [Date 30 days from now]
Hi [Name],
I’ve been thinking about whether [Product] is the right fit for what you’re trying to do, and I don’t think it is. I’m going to close your account on [Date 30 days out] and refund your last payment in full.
To make the transition easier:
Data export: [link to export tool]
Refund: [amount] processed today, will appear in 5–10 business days
An alternative I’d recommend: [Competitor X], which I think is a better match for your use case
I appreciate you giving us a try and I wish you the best with [their goal].
[Your name]
Notice what the script does not do. It does not list grievances. It does not negotiate. It does not invite a debate over reasons. It does not blame the customer. The framing is “we’re not the right fit for you” — which is true, even if the unspoken half of the truth is that they’re not the right fit for you either.
The 30-day notice matters. Anything less reads as retaliation. The full refund matters. Anything less invites a chargeback. The competitor recommendation matters. It signals that you’re prioritising their outcome over your revenue, which short-circuits most negative-review scenarios.
The reactions cluster into four predictable patterns. Bracing yourself for each makes the experience much less stressful.
Most go quietly. Around 70% of fired customers reply with a one-line acknowledgment, take the refund, export the data, and disappear. They knew the relationship wasn’t working too.
Some try to negotiate. About 20% reply asking what they can do to stay. The answer is to thank them, restate the decision, and move on. Do not negotiate. Once you’ve decided, the energy of re-litigation is what you were trying to escape in the first place.
Some leave a 1-star review. Maybe 5–10% will retaliate publicly. The full refund and the alternative recommendation are your insurance against this; even when it happens, the review tends to be self-evidently unhinged. Reply once, calmly, and let it sit.
The energy gain is worth it. The category nobody warns you about is the one where you discover, two weeks later, that you’re sleeping better and shipping more. The customer was a much bigger weight than the MRR ever indicated. Marc Benioff’s public stance on letting customers go is that the relationship can stay friendly even when the contract ends — and across public solo-founder retrospectives, that pattern holds up surprisingly often.
“Sometimes the best thing you can do for a customer is help them realise you’re not the right vendor for them — and stay friends.”
— Marc Benioff, SalesforceThe opposite mistake is firing a customer who’s actually a future high-value relationship having a hard week. Use the checklist below before you draft the email.
This isn’t a standalone tactic. Firing customers is downstream of pricing, niche, and positioning. If you’re drowning in bad-fit customers, the upstream fixes are usually more powerful: raise prices to filter, narrow your niche to attract people who actually need the product, or rewrite the pitch so misfits self-select out at the landing page. We covered the pricing angle in our solo founder pricing playbook, the niche angle in our when to niche down or broaden guide, and the LTV implications in what is LTV.
Customer-firing is also the natural pair to a real churn-reduction effort — sometimes the best churn prevention is the controlled exit you initiated yourself. Our SaaS churn reduction playbook covers the inbound side; this guide covers the outbound side.
The most underrated fact about firing customers is that you almost never regret it. The most overrated fact is the social pressure you feel before you do it. The first one is real. The second one fades the day you hit send.
The stack, prompts, pricing, and mistakes to avoid — for solo founders building with AI.