Stale ticket detection

What makes an open ticket stale, where stale tickets show up, how the AI stale analysis works and where to change the rules.

Written By Chris Scaminaci

Last updated About 3 hours ago

A ticket is stale when it has gone quiet or has stayed open too long. QuantumOps marks stale tickets so that nothing sits unnoticed. For each stale ticket, AI explains why it stalled and how to restart it. Stale detection reads from HaloPSA and writes nothing back to it.

Before you start: everyone who can open Home can see stale tickets. Changing the rules needs the Administrator or Q-Director Configuration role. The rules are in Q-Director, on the AI Processing tab, in Stale Ticket Detection.

What makes an open ticket stale

A ticket is stale when any one of these is true. The numbers are the defaults, and an administrator can change every one.

Rule	A ticket is stale when
Inactivity	Nothing has happened for more than Inactivity Threshold (business days). The default is 2, and weekends and the main US public holidays do not count. Inactivity Trigger chooses what counts as activity: No Agent Action (the default, only an agent's action counts), No User Action, No Any Action, or Agent Awaiting Response (stale when the agent acted last and the customer has not replied).
Follow-up overdue	Track Agent Follow-Up Requirements is on, an agent acted last, and the agent has not followed up within Follow-up Required Within (days). The default is 3, and days are counted the same way as for inactivity.
Age	The ticket has been open longer than Max Ticket Age (days), whatever the activity. The default is 7.
Late assignment	It took longer than Assignment Delay (hours) to assign the ticket. The default is 2.
SLA hold	A ticket with an SLA has been on hold longer than SLA Hold Time Threshold (hours). The default is 2.

Each problem also lowers the ticket's health score; the next section says how. A missed SLA is tracked on its own, not as stale, when Track SLA Breaches Separately is on.

Two things are never stale: a closed ticket, and a ticket of a type that stale analysis leaves out. Ticket Types to Include in Stale Analysis limits the check to the types you pick. Leave it empty to check every type. Configure Per-Ticket-Type Overrides gives one ticket type its own thresholds.

How the health score is calculated

Every ticket that has been assessed has a health score from 0 to 100. It starts at 100, and each rule in the table above that applies to the ticket takes points off:

Inactivity and an overdue follow-up take off points for every day past the limit.
Age, a late assignment and a long SLA hold each take off a fixed number of points.
A missed SLA takes off a heavier penalty than the other rules. It counts only while Track SLA Breaches Separately is on and the ticket has an SLA, and the SLA hold rule is then not applied to that ticket.

The score never goes below 0. A ticket that has not been assessed has no score, and the screens show a dash or a grey bar instead.

Where the score shows

Each screen bands the score differently, and one label means different ranges on two of them.

Screen	What it shows	Bands
Home, Status column	A bar beside Open or Closed. Point at the bar to read the score.	Green from 75, yellow from 50, orange from 25, red below that. Grey means not assessed.
Ticket details dialog, Health badge	The score as a percentage, or a dash when the ticket has not been assessed.	Red below 60, orange from 60 to 79, green from 80.
Ticket Dashboard, status bar	A verdict and the score as a percentage. An open sentiment or tonality alert, or a stale ticket, sets the verdict before the score does.	EXCELLENT from 80, GOOD from 60 to 79, NEEDS ATTENTION from 40 to 59, CRITICAL below 40. NOT ASSESSED when there is no score.
Client Dashboard, Client Health gauge	The average score of the client's open tickets.	Excellent from 80, Good from 60 to 79, Fair from 40 to 59, Needs Attention below 40.

NEEDS ATTENTION on the Ticket Dashboard means 40 to 59. Needs Attention on the Client Dashboard means below 40. See Home: the QA Pipeline, Ticket details, sentiment and sanity scores, The Ticket Dashboard in HaloPSA and The Client Dashboard for each screen.

Where stale tickets show

On Home. The Stale Tickets tile, the Stale column and the stale counts on the Agent workload card. Point at the amber icon in the Stale column to read the analysis and recommendations. See Home: the QA Pipeline.
In the ticket details dialog. Stale Review lists the Issues Identified and the Recommended Actions. See Ticket details, sentiment and sanity scores.
In the Ticket Dashboard in HaloPSA. A Stale Analysis card with Recommendations. See The Ticket Dashboard in HaloPSA.
In the Customer Success Center. Stale alerts, and the Stale Insights tab with figures such as Currently Stale and Avg Days Stale. See Customer Success Center. Auto-Clear Resolved Stale Alerts in its settings dismisses a stale alert when the background scan finds the ticket closed, or finds that its type is no longer checked for staleness.

How the AI stale analysis works

For a stale ticket, QuantumOps reads the conversation and writes an assessment of why the ticket went quiet and recommendations for restarting it.

The analysis uses AI tokens, so it does not run on every check. In the background scan it runs when a ticket first becomes stale, and again only when the ticket's actions have changed and at least Min Hours Between Analysis (default 4) have passed since the last analysis. When a ticket is refreshed after an update, it runs if the ticket has never been analysed or if that many hours have passed.

An administrator controls this with Enable AI Stale Analysis and Analysis Trigger. The background scan runs the analysis whenever Enable AI Stale Analysis is on. Analysis Trigger decides where else it runs: Incremental Refresh (recommended) adds the refresh after a ticket update, and a full analysis of a ticket runs it for any value except Disabled. To stop the analysis everywhere, switch off Enable AI Stale Analysis. Stale detection itself keeps working.

The Stale Analysis entry in Module Management does not switch stale detection or the AI analysis on or off. Use the Q-Director settings above. See Module Management.

When QuantumOps checks

When a ticket changes. Every update, customer e-mail and closure refreshes the ticket, and the stale rules run as part of that refresh. See How ticket analysis works.
In the background. A scan rechecks open tickets that have not been checked for Stale Recheck (hours). It runs every Scan Interval (mins), 30 by default and never less than 5. It pauses while an import is running. The scan runs only when Enable incremental refreshes on ticket updates is on, in Incremental Lifecycle Analysis on the same tab.

Change the rules

Open Q-Director and select the AI Processing tab.
In Stale Ticket Detection, change the thresholds. Use recommended values fills the section with the values suggested from fleet data. Nothing is saved until you select Save Changes.
For one ticket type, select Configure Per-Ticket-Type Overrides.

See Q-Director: AI Processing for every setting.

Was this helpful?

Still need help? Ask the team