Open-ticket count is a rough proxy for workload. A person waiting for ten customer replies may have room for new work; someone handling two complex escalations may not. A routing model needs an agreed view of active work.
The May HubSpot roundup described capacity exclusions based on ticket stage or age. The configuration should follow a deliberate workload policy.
Define the waiting states carefully
Separate waiting for a customer, waiting for an internal specialist and waiting for an external provider. Each state needs an owner and a next review point, even if it does not consume the same routing capacity.
Excluding a ticket from capacity must not remove it from follow-up reporting. Otherwise a more available-looking team can accumulate unattended customer commitments.
Agree what the model is trying to distribute
Start with a manageable definition, such as the count of actionable assigned tickets within a support queue. Avoid inventing a precise effort score before you have evidence that the categories are consistent.
Review complex work separately. If certain request types demand specialist time, document that constraint in routing and staffing rather than assuming a single numeric limit fits every role.
Plan for no eligible recipient
A ticket can meet the right language and product criteria while no suitable person is available. Define a fallback queue, a monitoring owner and an escalation expectation. The absence of capacity should create a visible operational state.
For an illustrative test, submit a request outside the normal coverage window. Confirm where it lands, what the customer is told and who sees it when coverage resumes.
Test transitions that change workload
- A new ticket enters the queue while an owner is at capacity.
- A waiting ticket becomes actionable after a customer reply.
- A ticket is escalated but customer communication remains with support.
- A person becomes unavailable during an active assignment.
- A category correction changes which team is eligible.
Inspect both the routing result and the follow-up obligation. Rebalancing work should not erase the context needed by the next owner.
Review fairness with the team
Compare assignments, active backlog, waiting work and escalation load. Ask users whether the model reflects effort, and inspect examples where it does not. A lower average ticket count does not establish a better customer outcome.
Change one part of the policy at a time where possible. Document the expected effect, observe the distribution and retain a clear way to explain assignments.
The Service Hub operating model covers ownership and SLAs; the workflow design guide covers exception handling.
Make routing match the work.
I can review capacity assumptions, queue ownership and service workflow requirements with your team.
Explore service workflow design →