Uptime and Availability Commitments: What to Hold the Vendor To

Availability clauses are written to look reassuring and behave conservatively. A vendor can promise 99.9 percent uptime and still owe you almost nothing, because the same contract narrows what counts as downtime, excludes the most common causes, and caps the credit at a token figure you have to chase. The buyer's job is to read past the percentage to the mechanics. This clause sits in the wider ITSM contract clause library; here we focus on what to actually hold the vendor to.

What uptime should an ITSM SLA commit to?

Many ITSM SaaS vendors publish a 99.9 percent monthly availability target as standard, which permits roughly 43 minutes of downtime in a 30-day month. Higher tiers such as 99.95 percent exist, often at a premium, and they roughly halve the permitted downtime. But the percentage is a ceiling on the vendor's exposure, not a floor on your experience, and a higher number bought without tightening the definition buys very little. Treat 99.9 percent as a baseline to test, not a finish line.

The downtime figures above are arithmetic from the percentage, not vendor promises. They show why the jump from 99.9 to 99.99 percent is a large operational difference, but only if the contract counts downtime the way you do.

How is downtime defined, and why does it decide everything?

The definition of downtime is where the real negotiation lives. Vendors commonly exclude scheduled maintenance, problems caused by third parties such as cloud infrastructure or integrations, and degraded performance that stops short of a complete outage. Under a narrow vendor definition, a service that is unusably slow for an afternoon may register as fully available. A buyer-side definition counts the time the service is unusable for its intended purpose, brings recurring scheduled maintenance into a reasonable cap, and treats severe performance degradation as downtime. The percentage is identical; the protection is not.

Book a contract negotiation review →

Are SLA service credits worth negotiating?

Service credits are usually modest and capped, often a single-digit percentage of the monthly fee per breach with an overall ceiling, and they typically must be claimed within a short window or they lapse. As cash, they rarely make you whole. Their real value is structural: a credit regime forces the vendor to report incidents and creates a record of repeated failure that supports stronger rights. So negotiate credits to be automatic rather than claim-only, lengthen the claim window, and, more importantly, attach an escalation path and a termination trigger after repeated or sustained breach. The credit is the symptom; the exit right is the cure.

What should you hold the vendor to beyond the number?

1. A tight downtime definition. Count time the service is unusable for its purpose, including severe degradation, not only total unreachability.
2. Narrow, capped exclusions. Limit scheduled-maintenance carve-outs to a stated number of hours in agreed windows, and resist blanket third-party exclusions for components the vendor chose.
3. Automatic credits. Have credits apply on the vendor's own monitoring data rather than only when you notice and claim within days.
4. Transparent incident reporting. Require timely root-cause reporting and access to availability data you can audit.
5. A termination right. Secure the right to exit, without penalty, after repeated monthly breaches or a single sustained outage beyond a defined threshold.

How does the availability clause connect to the rest of the deal?

Availability rights are only as strong as your ability to act on them, which is why they pair with your exit terms. A robust uptime clause with a termination trigger is worth far more when the contract also lets you leave cleanly, the mechanics of which are set out in termination for convenience in ITSM agreements. Read together, the two clauses turn poor service into leverage rather than a sunk cost. Across ITSM engagements the firm averages a 30 percent reduction, and while service levels are not where the headline saving comes from, a weak SLA is where a good price quietly erodes into a bad outcome.

What to do next

Stop negotiating the percentage and start negotiating the definition, the exclusions, the credit mechanism and the exit right, then confirm the published 2026 SLA tiers against your order form rather than the vendor's marketing page. When you want the service-level clause built to be enforceable alongside the rest of the contract, our contract negotiation service runs the engagement on fixed fee or gainshare, vendor-neutral throughout.

Frequently asked questions

What uptime should an ITSM SLA commit to?

Many ITSM SaaS vendors publish 99.9 percent monthly availability, roughly 43 minutes of downtime a month. Higher tiers exist, but the figure matters less than the downtime definition, exclusions and remedy.

How is downtime defined in an ITSM contract?

Vendors often exclude scheduled maintenance, third-party failures and degraded performance short of a full outage. A buyer-side definition counts time the service is unusable for its purpose, not only total unreachability.

Are SLA service credits worth negotiating?

Credits are usually small, capped and claim-bound, so they rarely make you whole. Negotiate them to be automatic, widen the claim window, and attach escalation rights and a termination trigger after repeated breach.

What should a buyer hold the vendor to beyond the uptime number?

A tight downtime definition, narrow exclusions, automatic credits, transparent incident reporting, and a termination right after repeated or sustained breach. These decide whether the commitment is real.