Ticket Triage Metrics: What to Measure So the Right Tickets Reach the Right Agents

Triage is the invisible step between a ticket arriving and an agent starting to work on it. When triage works, the right ticket reaches the right agent within minutes and first reply time stays low. When triage breaks — tickets sit unassigned, land with the wrong group, or bounce between agents — every downstream metric suffers.

Most teams measure the results of triage (slow response time, high escalation rate) without measuring triage itself. This post covers the specific metrics that reveal whether your triage process is working, what causes each one to degrade, and what to fix first.

What triage actually does

In a Zendesk context, triage is the process that takes a new ticket and determines:

Who should handle it? Which agent or group has the right skills and capacity?
How urgent is it? What priority should it carry?
What is it about? Which category, product area, or issue type does it belong to?

Triage can be manual (a person reads the ticket and assigns it), automated (Zendesk triggers and automations route based on rules), AI-assisted (intelligent triage predicts intent and priority), or a hybrid of all three.

Regardless of method, the goal is the same: reduce the time between “ticket created” and “the right agent starts working on it.” Every minute spent in triage is a minute added to the customer’s wait.

The four core triage metrics

1. Time to first assignment

This is the time between ticket creation and the first assignment to an agent or group. It measures how long the ticket sits in limbo before anyone owns it.

How to measure in Zendesk: Track the time to first assignment metric. In Explore, this is the difference between the ticket created timestamp and the first assignee change. If you use round-robin or auto-assignment, this should be near zero for automated tickets.

What to look for:

Median under 5 minutes: Triage automation is working. Most tickets get assigned quickly.
Median 15–60 minutes: Manual triage queue. Someone is reviewing tickets before routing. This may be acceptable for low-volume teams but becomes a bottleneck at scale.
Median over 1 hour: Triage is broken or nonexistent. Tickets are sitting in a shared queue waiting for someone to claim them.

What drives it up:

No auto-assignment rules configured.
Triage queue relies on a single person who is not always available.
Tickets arriving outside business hours with no after-hours routing.
Ambiguous tickets that do not match any routing rule.

2. Routing accuracy

Routing accuracy measures whether the first assignment was correct — whether the ticket reached an agent who could actually handle it without reassigning.

How to measure: Compare tickets with zero reassignments (first assignment was correct) to tickets with one or more reassignments. The percentage of tickets that were resolved by the first assigned agent or group is your routing accuracy rate.

Alternatively, track group reassignment rate — the percentage of tickets that moved between groups. A high reassignment rate is a direct signal of poor routing. See the group reassignment rate report for setup details.

What good looks like:

> 85% of tickets resolved by first assigned group: Routing is working well.
70–85%: Acceptable but worth investigating the misrouted categories.
< 70%: Routing rules need significant rework. Customers are experiencing delays because their tickets are bouncing.

What drives it down:

Trigger rules that are too broad (e.g., all “billing” tickets go to one group, but half are actually account access issues).
Missing routing rules for new products, channels, or customer segments.
Agents transferring tickets because they lack training, not because the ticket was misrouted.
Zendesk auto-assignment accuracy not tuned to current team structure.

3. Escalation rate from triage

This is separate from your general escalation rate. Triage-related escalation captures tickets that are assigned, worked on briefly, and then escalated — not because the issue is complex, but because the initial routing was wrong.

How to identify triage escalations vs. legitimate escalations:

Triage escalation: Ticket reassigned within 30 minutes of first assignment with minimal or no agent interaction. The agent looked at it, realized it was not their area, and moved it.
Legitimate escalation: Ticket reassigned after the agent spent time investigating and determined it required specialist expertise.

In Explore, filter reassignments where the time between assignment and reassignment is under 30 minutes and the agent sent zero replies. This cohort represents triage failures.

Why it matters: Each triage escalation adds 15–30 minutes of dead time to the ticket. The first agent spends time reading and deciding to reassign. The second agent spends time re-reading the same context. The customer waits through both. If your triage escalation rate is 15%, you have a systemic routing problem hiding inside your first reply time numbers.

4. Unassigned ticket age

This metric tracks how long tickets sit without any assignment. It is the most basic triage failure: not routing at all.

How to measure: Filter for tickets where assignee is null and age is greater than a threshold (e.g., 30 minutes during business hours). Count these tickets and track the count over time.

What to look for:

Zero unassigned tickets older than 30 minutes during business hours: Excellent triage coverage.
1–5 unassigned tickets at any given time: Manageable, but investigate why they slipped through.
10+ unassigned tickets regularly: No effective triage process exists. Implement auto-assignment or a dedicated triage role.

Track this alongside your backlog metric. Unassigned tickets are the worst kind of backlog — they are not just unsolved, they are unowned.

Secondary triage metrics

Beyond the four core metrics, these add context:

Priority accuracy

If your triage process sets priority (manual or via intelligent triage), check whether the initial priority was correct. Tickets where priority was changed after assignment suggest the triage step is not accurately assessing urgency.

Track tickets where ticket priority was changed within the first hour. A priority change rate above 20% means your priority-setting criteria need revision.

Channel-to-group routing match

Different channels often require different routing. Chat and phone should route to agents with availability for synchronous work. Email can route more broadly. Track routing accuracy by channel to find channel-specific gaps.

After-hours triage gap

Tickets created outside business hours often have different triage patterns. If your team operates 9–5 but customers submit tickets at midnight, those tickets may sit unassigned until morning. Measure first-assignment time segmented by business hours vs. after hours to quantify the gap.

Diagnosing triage problems

Use this diagnostic framework when triage metrics degrade:

Symptom	Likely cause	Where to look
High assignment time, low reassignment	Slow manual triage queue	Triage staffing, automation gaps
Low assignment time, high reassignment	Fast but inaccurate routing	Trigger rules, tag accuracy
Rising unassigned count at specific hours	After-hours coverage gap	Business schedule, automation rules
High reassignment for one group only	Group scope mismatch	Group definitions, trigger conditions
Priority changes after assignment	Poor initial classification	Triage criteria, intelligent triage config

Start with the symptom, trace to the cause, and fix the root. Treating the symptom (e.g., pressuring agents to respond faster) without fixing triage just masks the delay.

Fixing common triage failures

Problem: tickets sit unassigned

Fix: Implement auto-assignment. Zendesk’s round-robin assignment distributes tickets across available agents in a group. If round-robin is too blunt, use triggers to route to specific groups based on form, tags, or subject line keywords, then let round-robin assign within the group.

For teams using Zendesk intelligent triage, configure auto-routing based on predicted intent. This is more accurate than keyword-based triggers for ambiguous tickets.

Problem: tickets land in the wrong group

Fix: Audit your routing triggers. Export the last 100 reassigned tickets and categorize why they were reassigned. Common findings:

Triggers match on outdated keywords.
A catch-all rule sends too many tickets to a default group.
New products or services lack routing rules entirely.

Update triggers quarterly. Every product launch, org change, or new support tier should include a routing-rule review. See the Zendesk triggers audit guide.

Problem: triage queue bottleneck

Fix: If a single person triages all tickets, you have a single point of failure. Options:

Distribute triage: Rotate the triage role daily among agents. Each person spends 1 hour triaging before working their own queue.
Automate the easy tickets: Use triggers to auto-assign tickets that match clear patterns (specific forms, known tags, specific requestor organizations). Only tickets that do not match any rule hit the manual triage queue.
Set a triage SLA: Tickets should not sit in the triage queue for more than 15 minutes during business hours. If they do, the triage process is under-resourced.

Problem: priority is wrong at assignment

Fix: Simplify your priority criteria. If agents regularly change priority after assignment, the criteria are either too complex or too subjective. Define clear, observable rules:

Urgent: Production outage affecting multiple users. System down.
High: Functionality broken for a single user. Workaround not available.
Normal: Feature question, how-to, or minor issue with workaround.
Low: Enhancement request, feedback, non-blocking question.

Train triage staff on these definitions. If using intelligent triage, review the AI’s priority predictions monthly and correct misclassifications to improve the model.

Connecting triage metrics to downstream performance

Triage metrics are leading indicators. Downstream metrics are the lagging confirmation:

Triage metric	Downstream impact
High assignment time	Higher first reply time
Low routing accuracy	Higher resolution time (more handoffs)
High triage escalation	Higher reassignment rate, lower CSAT
Unassigned tickets	Growing backlog, SLA breaches

If your first reply time is trending up but agent performance has not changed, check triage first. The delay may be happening before the agent even sees the ticket. See reduce first response time for the full playbook.

What to review and when

Cadence	Metric	Purpose
Daily	Unassigned ticket count	Catch triage gaps same day
Weekly	Median assignment time, reassignment rate	Spot triage degradation trends
Monthly	Routing accuracy rate, triage escalation rate	Strategic routing improvements
Quarterly	Full trigger audit	Align routing with current team and product structure

Add triage metrics to your weekly support ops review. They belong alongside first reply time and backlog because they explain a large share of what drives those numbers.

Key takeaway

Triage is not an administrative step — it is the gate between a ticket arriving and a customer getting help. If triage is slow, every metric downstream is slower. If triage is inaccurate, agents waste time on tickets they cannot handle and customers wait through handoffs they should not experience.

Measure assignment time, routing accuracy, triage escalation rate, and unassigned ticket age. Fix the root causes — automate what is predictable, train on what is not, audit triggers quarterly — and watch your first reply time and resolution time improve without any change to agent performance.

See where your tickets stall before agents even start — start free