How SLA Response Times Work

A service level agreement (SLA) is the part of a contract that sets the level of service a provider commits to: how quickly it responds to a support request, how quickly it resolves one, and how much of the time the service is available. The same "4-hour response" can mean very different deadlines depending on whether the clock runs on business hours or calendar hours, so the definitions in the agreement matter more than the headline number.

Business hours or calendar hours

A business-hours clock only runs during the hours the agreement defines as business hours, on the days it defines as business days. A calendar-hours clock runs around the clock, weekends and holidays included.

Take a ticket opened at 4:30 PM on Friday 22 May 2026 under a 4-hour response commitment, with business hours of 9:00 AM to 5:00 PM, Monday to Friday:

• Calendar hours: the response is due at 8:30 PM the same Friday.
• Business hours, no holiday calendar: 30 minutes count on Friday, the remaining 3 hours 30 minutes start at 9:00 AM on Monday 25 May, and the response is due at 12:30 PM that Monday.
• Business hours with the US federal holiday calendar: Monday 25 May 2026 is Memorial Day, so the clock resumes on Tuesday 26 May and the response is due at 12:30 PM that Tuesday.

That spread, from Friday evening to Tuesday lunchtime, is the whole reason to read the definitions: which hours count, in which time zone, and whether the agreement names a holiday calendar. The SLA deadline calculator computes windows set in business days with a country's holiday calendar applied, and the working hours calculator totals the working hours over a date range with daily start and end times.

Priority tiers

An SLA can set different windows for different severities. Amazon Web Services' published support plans are one example: they list human response times of under 1 hour for "production system down", under 4 hours for "production system impaired", under 12 hours for "system impaired" and under 24 hours for "general guidance". Other providers set their own tiers and windows, and an SLA you sign may define them differently again, so the agreement is the only reliable reference for yours.

When the clock starts and stops

Read three definitions before relying on a response window. First, what starts the clock: the moment a ticket is submitted, the moment the provider's system logs it, or the moment a person reviews it. Second, what counts as a response: an automated acknowledgement or a reply from an engineer. Third, whether the clock pauses while the provider waits for information from you, often shown as a "pending customer" status. Each of these can move the real deadline by hours.

Uptime commitments

Uptime commitments are separate from response times and use a different calculation: the share of a period during which the service is available. The arithmetic is fixed:

• 99.9% uptime allows 0.1% downtime: 8.76 hours in a 365-day year, or about 43.8 minutes in an average month.
• 99.99% uptime allows 0.01%: about 52.6 minutes in a 365-day year.

What counts as downtime is not fixed; the agreement defines it. Amazon's EC2 SLA, for example, measures a "Monthly Uptime Percentage" for each monthly billing cycle, and Google Cloud's Compute Engine SLA measures a Monthly Uptime Percentage against a per-service target.

Remedies for missed commitments

The cloud SLAs cited here express the remedy as a credit against the bill and state the scale. Amazon's EC2 SLA gives a 10% service credit for a region-level Monthly Uptime Percentage below 99.99% but at least 99.0%, 30% below 99.0% but at least 95.0%, and 100% below 95.0%. Both Amazon's and Google's SLAs also state that the SLA sets out the customer's "sole and exclusive" remedy for a failure to meet it, unless the main agreement provides otherwise. If SLA performance matters to your business, read the remedy clause before signing.

FAQ

Does SLA response time use business hours or calendar hours?

It depends on the SLA, and often on the priority tier. Read the agreement's definition of business hours, its time zone and whether it names a holiday calendar. A 4-hour response from 4:30 PM on Friday 22 May 2026 is due at 8:30 PM that Friday in calendar hours, but at 12:30 PM on Tuesday 26 May in 9-to-5 business hours with the US federal holidays, because Monday 25 May is Memorial Day.

What is the difference between response time and resolution time?

Response time is how long the provider takes to respond to a ticket; resolution time is how long it takes to fix the underlying problem. An SLA can commit to either or both, and meeting the response commitment says nothing about when the issue will be resolved.

What happens if a vendor misses an SLA?

The agreement says. Published cloud SLAs give service credits: Amazon's EC2 SLA, for example, credits 10%, 30% or 100% of the charge depending on how far uptime fell, and both Amazon's and Google's SLAs describe the SLA remedy as the customer's sole and exclusive remedy unless the main agreement provides otherwise. Check the remedies section of your own agreement.