What it does
Defines how many and what type of work items a rep can hold simultaneously. Profiles replace the older unit-based capacity model. Multiple profiles can be assigned to a single rep (e.g. separate daily caps for high-priority and normal-priority work, plus an overall total cap).
Key facts
- Settings per profile: work item limit, reset frequency (Immediate or End of day), assignment blocking (Yes/No)
- Reset frequency cannot be changed after creation: must delete and recreate to change it
- Assignment blocking = Yes sets the rep's presence to Busy-DND when limit is reached; blocks all further auto-assignment including unit-based
- Custom per-rep limits override the profile default: set via Enhanced User Management → Update user attributes; changes take up to 15 minutes to sync
- Profile attaches to work items via workstream default or via classification ruleset output (enables multi-profile matching)
- Escalation profile: system-managed, activates automatically when a chat is escalated to voice/video; cannot be edited
- Capacity release timing: conversation: when rep closes and ends session; case: when resolved/cancelled or removed from Worked By; record/activity: not automatic, must deactivate queue item or clear Worked By
- Supervisor can manually assign beyond capacity; negative capacity values are valid
When to use / skip
Use capacity profiles for all new implementations. Unit-based capacity is the legacy model, don't mix them in the same org.
Configuration decisions
- Reset frequency: Immediate for real-time channels (chat, voice) where throughput matters; End of day for case-based work where daily volume caps are meaningful
- Whether to use multi-profile matching (Total + Specific profiles): powerful for tiered service but adds setup complexity; worth it if the client has VIP/Standard or different case complexity levels
- Custom per-rep limits: useful for ramp-up/onboarding scenarios without creating separate profiles per tier
Gotchas
- Reset frequency is permanent at creation. If the client confirms the wrong type and you miss it, you're recreating profiles, reassigning users, and retesting. Confirm in UAT before assigning to live agents.
- Record/activity capacity doesn't release automatically. Reps and supervisors must deactivate the queue item or clear Worked By: rarely intuitive and often missed in training.
- Don't mix capacity profiles and capacity units in the same org. Behaviour becomes unpredictable.
Consultant notes
- Reset frequency confirmation needs to be on the sign-off checklist before any profiles are created. It's permanent, the options sound similar until you explain the difference with a concrete scenario, and recreating profiles after agents are assigned to them is painful.
- Record and activity capacity not releasing automatically is counter-intuitive and catches clients every time in UAT. "I closed the case, why is my capacity still full?": train on it explicitly and put the correct close process in the ops runbook, not just the handover document.
- Don't mix capacity profiles and capacity units in the same org. If the client has any legacy unit-based setup, clean it up before implementing profiles. The unpredictable behaviour when both coexist is difficult to explain to anyone.
Source last updated: 2026-02-03 | Review when: Reset frequency becomes editable or new capacity management options released