A service level agreement is the part of a contract that states what performance the provider commits to — availability, repair times, and often latency, jitter, and loss — and what remedy applies when they fall short. It is the only part of a provider's promises you can actually enforce.
An availability percentage with a stated measurement period; a mean-time-to-repair or response objective; performance objectives for latency, jitter, and packet loss where relevant; a definition of what counts as an outage; a list of exclusions; and a remedy with a claim process. An SLA missing the remedy or the claim process is a description, not a commitment.
This is where the substance usually is. Scheduled maintenance, force majeure, customer equipment and premises, and third-party access are commonly excluded, and reasonably so. What matters is how broadly they are drawn and how much notice maintenance requires, because a generous exclusions clause can hollow out an impressive availability figure.
Remedies are typically service credits proportional to the outage, not compensation for business loss. A credit for a few hours of service is unlikely to approach the cost of the outage to your operation. That is an argument for designing for resilience rather than relying on the SLA as protection — the SLA is accountability, not insurance.
Ask every provider, including us, for the actual SLA document before signing, and read the exclusions and the claim process rather than the headline number. Comparing two providers' marketing pages tells you nothing; comparing their agreements tells you everything.
SLA is one of the specifications these comparisons turn on.
Describe the requirement and an engineer will tell you which service actually fits — without a rate card or a sales script in the way.
Service availability depends on location, network proximity, capacity, and engineering review. Share an address and we will confirm what can be delivered there.