Support Tier

Support tiers are the standard way customer support orgs sort incoming tickets by complexity so each ticket lands with the right level of expertise. Tier 1 covers routine work like password resets, basic how-to questions, billing lookups, and FAQ answers. Tier 2 handles intermediate issues that need product knowledge: configuration help, integration debugging, and product-specific troubleshooting. Tier 3 covers escalated cases that need engineering, security, or executive involvement: data corruption, account compromise, breaking bugs, and contract disputes. The tier model exists because routing every ticket to a senior engineer would be expensive, and routing complex tickets to a generalist would produce bad answers and angry customers.

The volume distribution across tiers follows a predictable shape in most SaaS support orgs. Tier 1 absorbs 60 to 75% of total tickets, Tier 2 handles 20 to 30%, and Tier 3 takes the remaining 5 to 10%. The cost-per-ticket scales sharply by tier. A Tier 1 ticket costs roughly $4 to $8 to resolve. Tier 2 runs $15 to $30. Tier 3 can run $80 to $200 once engineering time is counted. This is why support orgs invest heavily in deflection at Tier 1: even modest improvements in self-service or AI deflection move thousands of tickets out of the queue and free up senior staff for the work that actually needs them.

AI deflection lands first at Tier 1 because the question patterns are repetitive and the answers live in the knowledge base. The AI Support Department standard architecture handles 30 to 50% of Tier 1 volume without human handoff, with answers grounded in RAG against the help center and citations back to source articles. Tier 2 work is harder to automate because it requires reasoning across configurations and product internals, but copilots that draft responses for human agents lift productivity 30 to 60%. Tier 3 stays human because the stakes are too high and the situations too novel for current models to handle without supervision. The tiering model is the foundation for cost-effective AI deployment in support.