A service level agreement is the written promise behind your hosting. Understanding what it guarantees, and what it leaves out, protects your organisation when reliability really matters.
An SLA is a hosting provider’s written commitment to a level of service, usually uptime. It states the guaranteed percentage, how downtime is measured, what is excluded, and the credit you receive if the provider falls short.
What an SLA is
An SLA, short for service level agreement, is a contract that sets out the level of service a provider promises to deliver. For hosting, the headline commitment is almost always uptime, the share of time your systems stay online and reachable.
The agreement turns a marketing claim into a contractual duty. Instead of a provider simply saying they are reliable, an SLA states a figure, such as 99.99 percent uptime, and spells out what happens if they miss it. For an enterprise, that accountability is the whole point.
What an SLA covers
A hosting SLA usually addresses a few core areas. Knowing them helps you read one properly.
- Uptime guarantee. The percentage of time the service will be available over a set period.
- Measurement. How the provider defines and measures downtime.
- Exclusions. Events that do not count against the guarantee, such as planned maintenance.
- Remedies. What you receive if the provider misses the target, usually account credit.
- Support commitments. Response and resolution times for support requests.
What uptime percentages mean
Small differences in uptime translate into large differences in real downtime. The numbers look close, but the gap over a year is wide.
- 99.9 percent. Around 8.8 hours of downtime a year.
- 99.95 percent. Around 4.4 hours a year.
- 99.99 percent. Around 52 minutes a year.
- 99.999 percent. Around 5 minutes a year, rare and costly.
Enterprise hosting typically targets 99.99 percent or higher. Each extra nine cuts downtime sharply but raises the cost and complexity of the platform behind it. Our guide to the best hosting with an SLA shows which providers commit to the higher figures.
A common trap. A high uptime figure means little if the SLA excludes broad categories of downtime or defines availability narrowly. Read the exclusions before you trust the headline number.
How SLA credits work
When a provider misses its uptime target, the SLA sets out the remedy. In most cases this is service credit, a percentage of your monthly fee returned or applied to a future bill. The worse the outage, the larger the credit.
Credits are worth understanding for what they are. They compensate for a missed commitment, but they rarely match the true cost of a serious outage to your business. The main value of an SLA is the accountability and priority it creates, not the refund itself.
What to check in an SLA
Two agreements with the same headline figure can differ sharply once you read the detail. Focus on these points.
- Definition of downtime. Confirm whether it covers partial outages and slow performance or only a total failure.
- Measurement window. Uptime measured monthly is stricter than an annual average.
- Exclusions. Note what does not count, especially planned maintenance and third-party faults.
- Claim process. Check whether you must report downtime to claim credit, and how quickly.
- Support terms. Confirm response times for critical issues, not just availability.
SLAs and high availability
A strong SLA depends on the infrastructure behind it. Providers meet high uptime targets by building redundancy, with failover and multiple data centres so no single fault takes the platform down. Our explainer on high-availability hosting shows how that architecture supports the promise on paper.
Put simply, the SLA is the commitment and high availability is how a provider keeps it. Judge both together. A confident guarantee backed by real redundancy is far more valuable than a bold figure with thin infrastructure behind it.
SLAs beyond uptime
Uptime is the headline, but a good enterprise SLA can cover more. Some agreements commit to support response times, so a critical issue gets attention within a defined window. Others cover performance, network latency, or the time to resolve a problem, not just the time to acknowledge it.
- Response time. How quickly the provider acknowledges a critical ticket.
- Resolution time. A target for fixing an issue, not just replying to it.
- Performance. Commitments on speed or latency, where the platform supports them.
- Support availability. Round-the-clock access to engineers for serious incidents.
Read every commitment in the agreement, not only the uptime line. A provider that stands behind response and resolution as well as availability gives you far more protection when a real incident hits.
Why SLAs matter for enterprise
For a large organisation, downtime carries a heavy price in lost revenue, stalled operations, and damaged trust. An SLA gives you a contractual footing, a defined remedy, and a clear escalation path when things go wrong. Treat the SLA as a core part of your buying decision rather than a formality. When you compare providers, weigh the strength and clarity of each agreement. Our roundup of the best hosting for enterprise highlights platforms that stand behind their uptime with agreements worth reading in full.
Frequently asked questions
What does an SLA guarantee?
A hosting SLA guarantees a defined level of service, usually a minimum uptime percentage over a set period. It also states how downtime is measured, what is excluded, and the credit you receive if the provider misses the target.
What is a good uptime figure in an SLA?
For enterprise hosting, aim for 99.99 percent or higher, which allows under an hour of downtime a year. Lower figures such as 99.9 percent suit less critical systems. Always check how downtime is defined alongside the number.
Do SLA credits cover the cost of downtime?
Rarely in full. Credits usually return a portion of your monthly fee rather than compensating for lost revenue or disruption. Their main value is the accountability and priority they create, not the refund itself.
What is usually excluded from an SLA?
Planned maintenance is the most common exclusion, along with faults caused by third parties, your own configuration, or events outside the provider’s control. Read the exclusions carefully, as they shape how much the headline figure actually protects you.
Does every hosting plan come with an SLA?
No. Basic shared plans often have no meaningful SLA. A firm, written uptime guarantee with defined remedies is a feature of business and enterprise hosting, and its strength varies a lot between providers.
