Design a bug SLA tiering system

Bug SLAs: Making Triage a 5-Minute Operation

Untiered bug backlogs consume disproportionate engineering attention — every bug feels urgent because no bug is explicitly not urgent. Atlassian's agile metrics guidance and The Pragmatic Engineer's bug triage writing both make the same point: explicit tier definitions with response times convert triage from a political debate into a 5-minute operation. The default-to-P2 rule is the load-bearing discipline — it forces explicit upgrade justification.

How the Design a bug SLA tiering system Prompt Works

The prompt defines four tiers with explicit triggers, response and resolution targets, routing rules, and escalation triggers. The "every bug starts at P2 by default" rule prevents the P0-inflation that happens when nobody wants to be the person who de-prioritizes a bug. Monthly SLA review keeps the tiers calibrated as the product and customer base evolve.

When to Use It

• Every bug is marked urgent and engineering cannot focus on features.
• A new on-call rotation is starting and triage rules are unclear.
• Customer escalations are bypassing the queue.
• A support team is frustrated by bugs that never get resolved.
• A new EM wants to reset engineering hygiene.

Common Pitfalls

• No default tier. Without a default, every bug gets debated into P0/P1. Default to P2.
• No escalation triggers. Chronic P3s become silent P1s over time. Count and age triggers catch them.
• No SLA review cadence. Tiers calibrated 6 months ago drift. Re-review monthly.

Sources

• Agile Metrics — Atlassian
• The Pragmatic Engineer Newsletter — Gergely Orosz
• Kanban — Atlassian
• Google re:Work — Google