Every business internet proposal seems to include a service level agreement, usually with an impressive uptime number near the top. Few people read past that number, and fewer still read the fine print that determines what happens when service falls short. Understanding an SLA before you sign can save a lot of frustration later.
This guide explains the main parts of a typical business internet SLA, what the numbers mean in practical terms and which questions to ask a provider.
What an SLA is, and what it is not
A service level agreement is a written commitment describing the performance a provider aims to deliver and the remedy if it does not. It is important to understand what that remedy usually is. In most cases, an SLA does not pay for your lost revenue or productivity. It typically offers a credit against your monthly bill.
That does not make SLAs meaningless. A strong SLA signals that a provider has built its network and support operation to meet a standard, and it gives you a documented baseline for holding them to it. But the real value lies in the provider’s ability to avoid outages in the first place, not in the credit you receive afterward.
Uptime percentages in plain terms
Uptime is expressed as a percentage of time the service is available over a period, usually a month. The differences between percentages look tiny but add up quickly. Over a 30-day month:
- 99 percent uptime allows roughly 7.2 hours of downtime
- 99.9 percent allows roughly 43 minutes
- 99.99 percent allows roughly 4.3 minutes
- 99.999 percent allows well under a minute
When comparing proposals, translate each percentage into minutes or hours. That makes the trade-offs much clearer, especially if your phones, point of sale or cloud applications depend on the connection.
How downtime is measured
The fine print matters here. Look for answers to these questions:
- When does the clock start? Many SLAs count downtime only from when you open a trouble ticket, not from when the outage began.
- What counts as down? A complete loss of service is clear. Severe packet loss or a connection running at a fraction of its speed may not qualify unless the SLA says so.
- What is excluded? Scheduled maintenance, customer-caused problems, power failures at your site and events beyond the provider’s control are commonly excluded.
- Where is it measured? Some SLAs cover only the provider’s network up to a certain handoff point.
Performance metrics beyond uptime
Availability is only one dimension. Some business SLAs also address latency, packet loss and jitter. These metrics matter a great deal if you run cloud voice or video meetings, because a connection can be technically up while delivering poor call quality. If real-time applications are important to your business, ask whether the provider commits to performance targets and how they are measured.
Response time versus repair time
These two phrases get confused often. Response time is how quickly the provider acknowledges and begins working on a problem. Repair or restoration time is how long it takes to get service working again. A one-hour response commitment is helpful, but it does not mean the problem will be fixed in an hour.
Also check when support is available. A provider with 24/7 support can start work on an overnight outage right away, while one with business-hours support may not begin until the next morning. For businesses operating around the clock, that difference can matter more than the uptime percentage.
How service credits work
Credits are usually a percentage of the monthly charge for the affected service, scaled to how long the outage lasted. Read closely for:
- Whether you must request the credit, and the deadline for doing so
- Caps that limit the total credit in a month
- Whether repeated failures give you the right to cancel without penalty
That last point can be the most valuable part of an SLA. A chronic reliability problem is far more damaging than a single outage, and the right to exit gives you leverage.
Different connection types, different expectations
Dedicated connections, such as dedicated fiber internet, typically come with stronger SLAs than shared broadband services because the capacity is reserved for your business. Wireless and satellite connections can be excellent choices, especially in remote areas, but their SLA terms may reflect different conditions. Ask each provider to explain the terms for the specific service being quoted rather than assuming one standard applies to everything.
An SLA is not a continuity plan
Even the strongest SLA cannot prevent every outage. A fiber cut from construction work can take a connection down regardless of the contract language. That is why many businesses pair a primary circuit with a diverse backup path, such as fixed wireless, and configure automatic failover.
Doug Roberts, chief technology officer of Cytranet, puts it this way: “An SLA tells you how a provider will make it right. Redundancy and good engineering are what keep you from needing it in the first place.”
For a deeper look at preparing for connectivity failures, see our business continuity playbook for cloud phone systems.
Questions to ask before you sign
- What exact uptime is committed for this specific service?
- How is downtime detected and when does the clock start?
- What performance metrics, if any, are guaranteed?
- What are the response and restoration commitments, and is support 24/7?
- How are credits calculated and requested?
- What rights do I have if outages keep happening?
Clear answers to these questions reveal far more about a provider than the headline percentage. If you would like to walk through SLA terms for your locations, the Cytranet team is happy to help through our contact page.
