FS-DOC-03Operations
Process Runbook / SOP
Support Ticket Triage & Routing
[YourCompany.com] · Customer Support Department · Prepared by FullSpec · [Today's Date]
This runbook describes how the Support Ticket Triage & Routing automation works, what happens at each step, and how your team interacts with it. The automation handles ticket categorization, priority assignment, and routing to the correct queue or agent. Your support team monitors the system, reviews edge cases, and maintains routing rules as your business evolves. FullSpec builds, deploys, and supports the automation end to end. You own the process outcomes and decide when to override or adjust assignments.
01Process overview
Every support ticket that arrives in your helpdesk follows the same path: it lands in Zendesk (or any supported helpdesk), the automation reads its content in real time, extracts the issue type and urgency level, matches it against your routing rules, and assigns it to the correct queue or agent. A Slack notification is sent instantly to the assignee, and the ticket metadata is logged to HubSpot for CRM context. Your team no longer manually reads and sorts tickets. Instead, agents receive pre-categorized, pre-routed tickets and can start solving immediately. The only human touchpoint is exception handling: edge cases, misclassifications, and high-priority escalations are flagged for your team lead to review.
Process
Support Ticket Triage & Routing
Trigger
A new support ticket arrives in Zendesk, Jira, or via email from any customer channel
Final output
Categorized, prioritized ticket assigned to correct queue or agent with Slack notification sent
Agents
Ticket Analyzer (extracts intent and priority); Smart Router (matches to queue or agent)
Tools
Zendesk, Slack, HubSpot, Gmail, workflow automation platform
Weekly volume
Approximately 80 tickets per week (350 per month)
Human checkpoint
Support team lead reviews edge cases and high-priority escalations daily
Process owner
[Your name], [Your title]
Process Runbook / SOPPage 1 of 4
FS-DOC-03Operations
02Step-by-step: what happens and who acts
The only human step in this process is the Support Team Lead reviewing edge cases and high-priority escalations each day. All other steps are automated.
Step
What happens
Who acts
Type
1. Ticket arrives
A new ticket lands in Zendesk from email, web form, or chat. The automation platform detects the event.
Automation platform (listening to Zendesk webhook)
Automated
2. Read ticket content
The Ticket Analyzer agent pulls the full ticket body, subject, customer info, and any attachments. It reads the message end to end.
Ticket Analyzer agent
Automated
3. Extract issue type and sentiment
The agent identifies the core problem category (billing, technical, account access, feature request, or complaint) and analyzes customer tone (frustrated, neutral, urgent).
Ticket Analyzer agent
Automated
4. Assign priority level
Based on keywords (payment failed, service down, account locked), customer tier, and sentiment, the system assigns low, normal, or high priority.
Ticket Analyzer agent
Automated
5. Look up routing rules
The Smart Router queries your routing rules table to find the correct queue, team, or individual agent based on category, priority, and current workload.
Smart Router agent
Automated
6. Update ticket in Zendesk
The system writes the category tag, priority tag, and assigned agent back to the ticket record. The ticket status moves to assigned or in-queue.
Automation platform (Zendesk API)
Automated
7. Send Slack notification
A real-time message is posted to the assigned queue channel or sent directly to the assigned agent with ticket summary, priority, and a link to the ticket.
Automation platform (Slack webhook)
Automated
8. Log to HubSpot CRM
The triage decision, category, priority, and routing choice are logged to the customer record in HubSpot for reporting and audit trail.
Automation platform (HubSpot API)
Automated
9. Review edge cases (daily)
The team lead checks the daily exception log for misclassified tickets, unusual requests, or high-priority escalations that need manual override or reassignment.
Support Team Lead
Human
From step 1 to step 8, the entire process takes approximately 45 seconds per ticket. Your agents receive the Slack notification and can open the pre-routed ticket immediately. No time is wasted reading, categorizing, or assigning.
Process Runbook / SOPPage 2 of 4
FS-DOC-03Operations
03Handling exceptions
The automation handles the majority of tickets correctly on the first try. However, some situations require human judgment or override. Below are the most common exception scenarios, what the system does, and what your team should do.
Situation
What the system does
What you do
Ticket category is ambiguous or multi-product
The automation flags the ticket as uncertain and assigns it to a generic support queue instead of a specialist. Confidence score is below 70%.
Your team lead reviews the ticket daily, re-reads the message, and manually reassigns it to the correct specialist queue. Update the category rule if this pattern repeats.
Customer is using abusive or threatening language
The system detects high anger/threat signals and escalates the ticket to your support manager instead of a regular agent queue.
The manager reviews the message and decides whether to respond, escalate to legal, or close the ticket. Provide clear guidelines on abuse handling to the team.
Customer tier or account status is not found in HubSpot
The Smart Router cannot determine priority based on customer tier. It assigns normal priority by default and routes to the general queue.
Check if the customer record exists in HubSpot. If the account is missing or incomplete, create or update it. Re-run the triage if needed.
Assigned agent is at capacity or unavailable
The Smart Router detects the agent is offline or has too many open tickets. It routes to the queue instead and adds the ticket to a backlog list.
Your team lead checks the backlog during peak hours and manually assigns tickets to available agents. Consider adding more agents if backlog grows.
Ticket contains sensitive data (credit card, password, SSN) that should not be logged
The system detects PII patterns and masks the sensitive data in the ticket text before logging to HubSpot. The full text stays in Zendesk.
Your agents can see the full ticket in Zendesk. No action needed. The automation does this automatically.
Routing rule has drifted or team structure changed
The automation applies an outdated rule because your team moved agents or merged queues. Tickets are routed to an old queue that no longer exists.
Update the routing rules table immediately. Test with a small batch of tickets before rolling out. Notify FullSpec support if rules are complex and need help adjusting.
04Who to contact and when
Use the table below to reach your team and FullSpec support. Contact your support team lead or manager for process questions, exceptions, and day-to-day issues. Reach out to FullSpec support for technical problems, integration failures, or suggestions to improve the automation.
Role
Name
How to reach them
Support Team Lead
[Your name]
[Email] or [Phone]
Support Manager
[Your name]
[Email] or [Phone]
Zendesk Admin
[Your name]
[Email] or internal Slack
FullSpec Support
Technical support team
support@gofullspec.com or visit gofullspec.com/support
Contact FullSpec support in these cases: automation is not triggering on new tickets, Slack notifications are not sending, Zendesk API errors appear in logs, or you need to add a new category or routing rule. Response time is 4 business hours for standard issues and 1 hour for critical outages.
Process Runbook / SOPPage 3 of 4
FS-DOC-03Operations
05Ongoing maintenance
The automation runs 24/7 with minimal intervention, but your team must perform regular maintenance to keep it accurate and up to date. Use the table below to stay on top of common maintenance tasks.
Daily
Review the exception log and misclassified tickets. Reassign edge cases. Check Slack notification backlog.
Support Team Lead
Weekly
Pull triage accuracy report from Zendesk. If accuracy drops below 90%, flag categories that are slipping. Review ticket volume trends.
Support Manager
Monthly
Audit routing rules. Confirm all queues and agents in the rules table still exist and are current. Check for new categories or queue changes.
Support Manager and Zendesk Admin
Quarterly
Review customer feedback and support metrics (response time, resolution time, CSAT). Adjust priority rules if SLA targets have changed.
Support Manager
As needed
Add new ticket categories, update customer tier mapping in HubSpot, or change team structure. Notify FullSpec support if rules are complex.
Support Manager and Zendesk Admin
Annually
Conduct a full audit of the automation with FullSpec. Review cost savings, accuracy gains, and feature improvements. Plan upgrades or tuning.
Support Manager and FullSpec support
The most common maintenance issue is category drift. As your business evolves, new types of tickets arrive that do not fit your original categories, or customer sentiment changes and the AI begins to misclassify. Review the weekly accuracy report closely. If you see a pattern (e.g. billing tickets are suddenly tagged as technical), update the category rule or add a new rule to catch edge cases. Do not wait for a crisis. Small adjustments each week keep accuracy high.
Process Runbook / SOPPage 4 of 4