When a tiered support model doesn’t fit
If your Tier 2 team is doing 60%+ of the work, you don’t have a tiered model — you have a misnamed flat model, and you should restructure: switch to skill-based default routing, a flat team structure, or put an AI agent in front of Tier 1. A tiered model assumes most conversations are routine and only some escalate; when most conversations need specialist knowledge instead, you’ll be reassigning constantly.
How to tell it’s breaking down
- Tier 1 reassigns a large share of its conversations rather than resolving them.
- Tier 2 or specialist teams handle the majority of total volume.
- Check Analytics → Team Performance to see how conversations actually distribute across teams.
Alternative 1 — Skill-based default routing
Skip the universal Tier 1 default. Instead, use channel-specific defaults: billing-related channels point directly at your Billing specialists, technical channels at your Technical team, and Tier 1 catches whatever’s left. Each channel’s default team is set in that channel’s settings — see Assign channels to a team.
This fits when your traffic already arrives pre-sorted by topic (dedicated inboxes, numbers, or widgets per subject area).
Alternative 2 — Flat team structure
No tiers at all: one or two teams that handle everything. Smaller operations often don’t need Tier 1 + Tier 2 complexity, and a flat structure removes the constant reassignment overhead. If the volume later grows and a routine/complex split emerges, you can reintroduce tiers.
Alternative 3 — An AI agent in front of Tier 1
For high-volume tenants, an AI Specialist Agent can handle the frontline directly — answering common questions, gathering context, and only handing off to human teams when AI confidence drops or the customer asks for a person.
This isn’t part of the team setup itself — it’s an Agent Stack feature that lives upstream of human assignment. The AI agent handles initial triage; team assignment happens at handover time. See Graceful handover to a team.
“Specialist Agent” in Atender refers to an AI agent inside an Agent Stack. It’s not the same as a human specialist team. Don’t conflate the two — they’re different layers.