When you miss your SLA commitment, something has to happen. The question is what. The answer your contract gives to that question has significant consequences for your business, your customer relationships, and your exposure in a bad month.
The industry standard remedy is service credits. Not refunds. Not termination rights. Credits applied to a future invoice, sized as a percentage of the fees for the affected period. Understanding why credits are the right structure, and how to build a credit schedule that’s defensible without being punitive, is what this post covers.
Why Credits and Not Refunds
The distinction between a service credit and a cash refund matters more than it might seem.
A refund returns money. It reduces your recognized revenue for the period, creates a cash outflow, and signals to the customer that the relationship has been unwound to some degree. For a subscription business, refunds also create ambiguity about whether the customer is entitled to continued service for the period they’ve been refunded.
A service credit does none of this. It acknowledges the failure, provides a meaningful remedy, and keeps the customer in the relationship. The credit is applied to a future invoice, which means the customer has to continue using your service to realize the value of it. This is good for retention. It also means the credit has no cash cost unless the customer churns before using it, in which case the liability typically expires under a well-drafted SLA.
From a customer’s perspective, credits are also the reasonable expectation. Enterprise procurement teams understand that SLA remedies are not designed to compensate for consequential losses. They’re designed to acknowledge service failures and provide a proportionate, relationship-preserving remedy. A customer demanding cash refunds for SLA misses is asking for something outside market norms, and you can say so.
Termination Rights: The Remedy to Avoid
The more dangerous ask is termination rights tied to SLA performance. Some enterprise customers will push for a provision that gives them the right to terminate for cause if you miss your uptime commitment by a certain threshold, either in a single month or across multiple consecutive months.
The problem is structural. A termination right tied to SLA performance gives a customer who wants to exit the contract a contractual mechanism to do so on your failure. If your service has a bad month and the customer has been looking for an exit, you’ve handed them one. Worse, termination for cause typically triggers immediate effect with no cure period, meaning you can’t fix the problem before the contract ends.
The provider-friendly position is that service credits are the sole and exclusive remedy for SLA failures. This language is important: “sole and exclusive remedy” means the customer cannot pursue other claims, including damages, for the same SLA breach. If you’re offering credits, make sure your SLA states clearly that credits are the only remedy available, not a floor from which customers can seek additional compensation.
If a customer insists on termination rights, the negotiated middle ground is a high threshold: termination only available after multiple consecutive months of material SLA failure, with a cure period that gives you the opportunity to remediate before the right triggers. Three consecutive months below 99% is a different conversation than one month at 99.1%.
The SLA as a Liability Cap
There is a broader structural point that founders often miss: a well-drafted SLA is one of the most effective tools in your commercial stack for capping your liability exposure for system failures.
Without an SLA, a customer who suffers losses during a service outage can argue that your general liability framework applies. Depending on how your Terms of Service is drafted, that could mean a claim for direct damages, and depending on the customer’s argument, potentially consequential damages too. The customer’s losses from a significant outage, including lost revenue, staff time, and downstream business impact, can vastly exceed your monthly fees.
A credits-as-sole-and-exclusive-remedy clause changes this entirely. It substitutes a defined, capped, predictable obligation for an open-ended damages claim. The customer gets a meaningful acknowledgment of the failure. You get certainty about your maximum exposure for any given outage. The credit cap, typically 50% of monthly fees, becomes the ceiling on your liability for service availability failures, regardless of what losses the customer claims to have suffered.
This is why the “sole and exclusive remedy” language in your SLA is not boilerplate. It is doing meaningful legal work. It forecloses damages claims that would otherwise survive alongside a credit regime, and it ties your SLA liability to a number you can actually model and budget for.
The connection to your broader liability framework matters here too. Your Terms of Service almost certainly contains a consequential damages exclusion and a liability cap tied to fees paid. Your SLA’s credit regime should be consistent with that framework: credits are the remedy for availability failures, and no other claims arising from availability failures survive. If a customer wants to argue that an outage caused them $500,000 in lost revenue and they’re entitled to that amount, your SLA and your ToS together should make that argument unavailable.
A credit schedule should be tiered: the worse the miss, the larger the credit. This reflects the proportionality principle and gives customers a meaningful remedy for serious failures without overexposing you to a bad month caused by a single incident.
A reasonable structure for a 99.9% uptime commitment:
| Monthly Uptime | Service Credit |
|---|---|
| 99.0% to 99.89% | 10% of monthly fees |
| 95.0% to 98.99% | 25% of monthly fees |
| Below 95.0% | 50% of monthly fees |
A few structural points worth noting:
Tie credits to monthly fees, not annual contract value. A customer paying $5,000 per month on an annual contract should receive credits calculated on $5,000, not $60,000. Some customer DPA redlines attempt to base credits on annual fees, which multiplies your exposure by 12.
Cap total credits per period. The standard cap is 50% of monthly fees in any calendar month. Without a cap, a series of smaller incidents in the same month can stack into an outsized credit obligation. The cap also reinforces that credits are a remedy, not compensation for consequential damages.
Set a minimum miss threshold. Credits should not trigger for immaterial SLA misses. If your commitment is 99.9% and you hit 99.89%, a 10% credit is reasonable. If you want to set a de minimis floor, a common approach is excluding misses of less than 0.1% from the credit schedule entirely.
Claim Procedures
Credits don’t apply automatically. Your SLA should require customers to submit a claim within a defined window after the incident, typically 30 days. Claims submitted after the window are forfeited.
This serves two purposes. It gives you notice that a customer believes they’re owed a credit, which allows you to verify the claim against your own monitoring data before issuing it. And it limits open-ended credit liability from incidents customers discover months later.
The claim procedure should be simple: written notice to a specified email address or through your support portal, identifying the incident date, the affected service, and the customer’s calculation of downtime. You review against your status page and monitoring data, confirm or dispute the calculation, and apply the credit to the next invoice if the claim is valid.
If your monitoring data and the customer’s data disagree, your SLA should specify which controls. The provider-friendly position is that your monitoring data, or data from a mutually agreed third-party status page, is the authoritative source. This removes disputes about whether an incident occurred and for how long.
Credit Expiry
Credits that aren’t used expire. The standard approach is that credits must be used within 12 months of issuance or they lapse. This matters because an accumulation of unused credits on a churning customer’s account creates a contingent liability you’d otherwise have to honor at termination.
Your SLA should state explicitly that credits have no cash value, are non-transferable, and expire if the agreement terminates before they’re applied. This closes the loop on the refund question: a customer cannot demand cash payment for accumulated credits on their way out the door.
The Customer Perspective
It’s worth understanding why credits work for customers too, not just providers. Enterprise buyers are not primarily looking for financial remedies when they negotiate SLAs. They’re looking for accountability and confidence that you’ll take reliability seriously. A well-structured credit schedule, combined with transparent uptime reporting and prompt incident communication, demonstrates operational maturity.
What enterprise buyers actually want is for the service to work. Credits are the fallback for when it doesn’t, and a reasonable credit schedule signals that you understand your obligations and have thought carefully about what happens when you fall short. That’s a more credible posture than either refusing to offer any remedy or agreeing to uncapped refunds you’d never be able to sustain.
No Boiler provides self-service legal document generation and educational content. This material and our service is not a substitute for legal advice. Please have a qualified attorney review any documents before relying on them. No Boiler is not a law firm, and communications with us do not create an attorney-client relationship or carry any expectation of confidentiality. Use of our platform and content is governed by our Terms of Service and Privacy Policy.