FS-DOC-03Operations
Process Runbook / SOP
Complaint Escalation Workflow
[YourCompany.com] · Customer Support Department · Prepared by FullSpec · [Today's Date]
This runbook is the day-to-day operating guide for your support team. It explains exactly what the automated Complaint Escalation Workflow does at each step, what your team is still responsible for, how to handle situations where something goes wrong, and who to call when you need help. FullSpec builds, monitors, and maintains the automation end to end. Your team's role is to resolve complaints once they land in the right place, and to override or escalate when human judgement is needed.
01Process overview
When a customer complaint arrives by email or web form, the automation platform detects it instantly and begins a structured response sequence. A Zendesk ticket is created automatically, the Complaint Triage Agent scores the complaint's severity and pulls the customer's history from HubSpot, and the Escalation and Routing Agent assigns the ticket, notifies the right agent or manager on Slack, and sends the customer a personalised acknowledgement via Gmail. If the ticket is not resolved before its deadline, an overdue escalation alert fires automatically. The only step that remains with your team is the actual resolution of the complaint, where human judgement and relationship skill are needed. Everything else runs without manual intervention.
Process name
Complaint Escalation Workflow
Trigger
A customer complaint is submitted via email, web form, or support chat
Final output
Complaint resolved, ticket closed in Zendesk, resolution optionally logged in Notion
Agents running
Complaint Triage Agent; Escalation and Routing Agent
Tools involved
Zendesk, HubSpot, Slack, Gmail, Notion
Weekly volume
Approximately 22 complaints per week (~90 per month)
Human checkpoint
Complaint resolution: the assigned agent investigates and resolves the complaint in Zendesk
Process owner
[Your name] — fill in once confirmed
Process Runbook / SOPPage 1 of 4
FS-DOC-03Operations
02Step-by-step: what happens and who acts
What you actually need to do: There is ONE human step in this process. Once the automation has triaged, assigned, and acknowledged the complaint, the assigned support agent opens the Zendesk ticket, reviews the customer context and severity score, resolves the complaint, and closes the ticket. Everything else is handled automatically.
Step
What happens
Who acts
Type
1
A complaint arrives by email or web form. The automation platform detects the new message immediately and triggers the workflow without any manual action required.
Automation
Automated
2
A ticket is created in Zendesk automatically. The customer's name, contact details, complaint text, and any attachments are pulled from the incoming message and populated into the ticket fields.
Automation
Automated
3
The Complaint Triage Agent reads the complaint text and assigns a severity score of low, medium, or high. It uses keyword matching, sentiment analysis, the customer's tier, and their prior complaint history retrieved from HubSpot.
Complaint Triage Agent
Automated
4
The customer's full record is fetched from HubSpot, including purchase history, lifetime value, and prior support interactions. This context is appended directly to the Zendesk ticket so the assigned agent has everything in one place.
Automation
Automated
5
The Escalation and Routing Agent assigns the ticket to the correct agent or team in Zendesk based on the severity score and complaint category. A Slack notification is sent to the assigned agent with a direct link to the ticket and a plain-English summary.
Escalation and Routing Agent
Automated
6
If the severity score is high, a separate Slack message is sent to the support manager with the complaint summary, the customer's value, and a prompt to review within one hour. This step is skipped automatically for low and medium severity tickets.
Escalation and Routing Agent
Automated
7
A personalised acknowledgement email is sent to the customer via Gmail within minutes of the complaint arriving. The email confirms receipt, names the assigned agent, and states the expected response time based on severity tier.
Automation
Automated
8
The assigned agent opens the Zendesk ticket, reviews the triage summary and customer context, and resolves the complaint. The agent can override the severity score directly in Zendesk before responding if the automated score does not match the situation.
Support Agent
Human
9
If the ticket has not been resolved by its deadline, an overdue escalation alert is sent automatically to the support manager in Slack and the ticket priority is updated in Zendesk. No manual reminder is needed.
Automation
Automated
10
Once the complaint is resolved, the agent marks the ticket closed in Zendesk. A resolution summary can be added to the shared Notion page for future reference, particularly for recurring complaint types.
Support Agent
Human
Process Runbook / SOPPage 2 of 4
FS-DOC-03Operations
03Handling exceptions
Situation
What the system does
What you do
Complaint arrives with missing or incomplete customer data (no email address, no name, or a blank complaint body)
The automation creates the Zendesk ticket with whatever data is available and flags it with a 'Incomplete intake' tag. The Triage Agent skips HubSpot enrichment and assigns a default medium severity score. A Slack alert is sent to the support team channel.
Open the flagged ticket in Zendesk, manually look up or contact the customer to fill in the missing details, then update the ticket and adjust the severity score if needed before proceeding to resolution.
A duplicate complaint is detected — the same customer submits the same complaint more than once within a short window
The automation identifies a matching open ticket by customer email and complaint keywords. The duplicate submission is merged into the existing ticket in Zendesk with a note recording the duplicate. No second acknowledgement email is sent.
If the merge looks incorrect (two separate issues from the same customer), open both tickets in Zendesk and manually unmerge them, then re-route the second ticket using the normal Slack channel to notify the agent.
The Triage Agent assigns a severity score that the agent believes is wrong — for example, a genuinely urgent complaint scored as low
The system proceeds with the automated score until an agent or manager overrides it. No automatic correction is applied.
Open the Zendesk ticket and update the severity field directly. The Escalation and Routing Agent will reprocess the routing and, if the new score is high, will send the manager Slack alert immediately. Log the case of misscoring in Notion so thresholds can be reviewed.
A customer does not respond to the automated acknowledgement email and the ticket cannot be progressed without their input
The automation monitors ticket age against the configured deadline. If the ticket remains in an awaiting-customer-response state past the deadline, an overdue alert fires to the manager in Slack and the ticket priority is updated.
The assigned agent should send a follow-up email manually from within Zendesk, noting the follow-up in the ticket. If there is still no response after a second attempt, mark the ticket as pending-closure and check with the support manager on how long to hold it open.
An agent receives a Slack notification for an assigned ticket but does not action it within the expected timeframe
The overdue escalation logic fires automatically when the ticket passes its resolution deadline, regardless of whether the agent has acknowledged the Slack message. The manager receives an escalation alert in Slack and the ticket priority is raised in Zendesk.
The support manager should check the Zendesk ticket, reassign it to an available agent if needed, and follow up with the original agent to understand what caused the delay. No manual reminder system is needed — the automation handles the chase.
Zendesk, HubSpot, Slack, or Gmail is temporarily unavailable
The automation platform retries the failed step up to three times with a short delay between each attempt. If the tool remains unavailable after three retries, the automation pauses that branch of the workflow and sends an alert to the support team's Slack channel (or falls back to email if Slack itself is down).
Check the FullSpec error log or contact FullSpec support at support@gofullspec.com for tool-outage status. While the tool is down, monitor the shared inbox manually for new complaints, create Zendesk tickets by hand where possible, and hold any unprocessed tickets for the automation to pick up once the tool recovers. Do not duplicate-process tickets once the system comes back online.
04Who to contact and when
Fill in the contact rows for your own team once your team is confirmed. The FullSpec rows are fixed and ready to use from day one.
Role
Name
How to reach them
Process owner
[Your name]
[Your email / phone]
Support Manager
[Rep name]
[Rep email]
Support Agent (primary)
[Your name]
[Rep email]
Support Agent (backup)
[Your name]
[Rep email]
Operations Lead
[Your name]
[Rep email]
FullSpec builder
FullSpec Team
support@gofullspec.com
FullSpec support
FullSpec Support
support@gofullspec.com
Contact FullSpec at support@gofullspec.com if: the automation stops processing new complaints, a tool outage lasts more than 30 minutes, severity scores appear systematically wrong across multiple tickets, or you need to change an escalation threshold, routing rule, or acknowledgement email template.
Process Runbook / SOPPage 3 of 4
FS-DOC-03Operations
05Ongoing maintenance
When complaint intake channels change (new email address, new form, new chat tool)
Notify FullSpec so the new intake source can be connected to the automation. Do not assume it will be picked up automatically. Each new channel requires a separate connection to be configured.
Process owner notifies FullSpec; FullSpec team makes the change
When acknowledgement email templates or Slack notification wording need updating
Send the revised copy to FullSpec with clear notes on which template needs changing and for which severity tier. FullSpec will update and test the change before it goes live.
Support Manager provides copy; FullSpec team implements
Monthly spot-check (first week of each month)
Review five to ten recent Zendesk tickets to confirm severity scores look accurate, acknowledgement emails were sent correctly, and no tickets were routed to the wrong agent. Note any patterns and flag them to FullSpec.
Support Manager or Operations Lead
Weekly error-log review (every Monday)
Check the automation error log for any failed steps, retry warnings, or skipped tickets. The log is accessible via the automation platform dashboard shared by FullSpec. Escalate anything unresolved within 24 hours.
Process owner or designated agent
When team members join or leave (any time)
Update the Zendesk agent list and Slack user permissions immediately so routing and notifications go to the right people. Notify FullSpec if routing rules reference a specific agent name rather than a role or queue.
Support Manager updates Zendesk and Slack; notifies FullSpec if routing rules are affected
Quarterly volume review (every three months)
Review complaint volumes against the baseline of approximately 90 per month. If volume has grown significantly, assess whether escalation thresholds, routing capacity, or automation plan limits need adjusting. FullSpec can advise on scaling.
Operations Lead reviews with FullSpec; FullSpec adjusts configuration as needed
The most common maintenance issue for this process is stale routing rules. When agents join, leave, or change roles, routing assignments in Zendesk can become outdated faster than teams expect. A complaint routed to a departed agent's queue will sit unactioned until someone notices. Review Zendesk agent assignments whenever your team changes and notify FullSpec immediately if any routing rules reference a named person rather than a role-based queue.
Process Runbook / SOPPage 4 of 4