SOC 2 Availability Criteria: Building for the A1 Trust Service Category
What the SOC 2 Availability criteria require, how to implement them, and why they matter for customer-facing SLAs.
The Availability Trust Service Category
The Availability criteria (A1) are an optional addition to your SOC 2 scope, but they're worth including if your customers care about uptime — which they do. The A1 criteria require that you:
A1.1: Performance and Capacity Management
You must have processes to monitor performance and capacity to ensure the system can meet availability commitments.
Evidence for auditors:
- Monitoring dashboards showing system resource utilization
- Capacity planning documentation
- Auto-scaling configuration records
- Performance metrics over the observation period
- Set up cloud provider metrics (CloudWatch, Azure Monitor) for CPU, memory, storage, and network utilization
- Configure alerts when utilization exceeds thresholds
- Document your capacity planning process (even if it's quarterly review of metrics)
A1.2: Environmental Protection
You must have controls to protect against environmental threats to availability: power outages, cooling failures, network disruptions.
For cloud-native companies: This largely inherits from your cloud provider's controls. Document your use of multi-AZ (or multi-region) deployments and reference your cloud provider's SOC 2 report as complementary coverage.
Evidence:
- Architecture diagram showing redundant components
- Cloud provider's SOC 2 or ISO 27001 report
- Your deployed region/AZ configuration
A1.3: Recovery and Business Continuity
You must have a recovery process to restore system availability after an event. This requires:
Recovery Time Objective (RTO)
How long can you be down before it breaches your SLA? Define this explicitly. Common values: 4 hours, 8 hours, 24 hours.Recovery Point Objective (RPO)
How much data loss is acceptable? This drives your backup frequency. Common values: 1 hour, 4 hours, 24 hours.Backup Configuration
- Database backups: frequency and retention matching your RPO
- Configuration backups: infrastructure-as-code in version control
- Backup encryption: backups are encrypted at rest
DR Test Evidence
Auditors want to see that you've actually tested your recovery procedures — not just that they exist. Run at minimum an annual DR test and document:
- Date of test
- Scope (which systems were restored)
- RTO achieved in the test
- Issues identified and remediated
Calculating Your SLA for SOC 2
| Uptime SLA | Allowed downtime per month |
| 99.9% ("three nines") | 43.8 minutes |
| 99.95% | 21.9 minutes |
| 99.99% ("four nines") | 4.4 minutes |
Match your stated SLA to what you can actually achieve. If you've had incidents, check whether your historical uptime supports the SLA you're promising customers.
Uptime Evidence
SecureSpect connects to your status page provider (Statuspage, Better Uptime) and records uptime data throughout the observation period. At audit time, you can export a monthly uptime report showing actual availability vs. committed SLA for the entire period.