How are you keeping agents on top of aging cases? Looking for tools, views, and AI ideas

Mike Clemons
Tera Expert

Hey all,

Looking for some collective wisdom from other CSM/ITSM admins and support leaders. We've been seeing a recurring pattern that I'd love to hear how others are solving:

The problem

Customers are increasingly reporting that Support hasn't responded in X days, or they're having to ping us multiple times for an update on an open case. When we dig in, it's usually a mix of agent follow-through and the fact that our tooling doesn't make staying on top of cases easy. Manager coaching helps in the short term, but it doesn't scale, and we keep ending up in the same loop.

What we think is contributing

The out-of-the-box ServiceNow case states (New, Open, Awaiting Info, Awaiting Vendor, Resolved, Closed, etc.) are too blunt to surface "this case has gone quiet and needs a nudge." Default list views don't do a great job prioritizing aging cases against newer ones, especially when an agent has 30+ open at any given time. Time-in-state and SLA breach indicators exist, but they don't always reflect customer-perceived staleness, for example, a case sitting in "Awaiting Info" for 10 days because the agent never followed up. Coaching the behavior helps individual agents, but turnover, workload spikes, and competing priorities make it hard to sustain.

What we're considering

A custom list view or dashboard that surfaces cases by "days since last customer-facing update" rather than just state or priority. A scheduled job or notification that pings the agent (and/or manager) when a case crosses a staleness threshold. AI-assisted approaches: an agent that scans open cases daily, flags ones that look neglected, and either drafts a follow-up or surfaces them to the human. Browser extensions or UI customizations to reorder agent views based on customer-impact signals.

What I'd love to hear from you

How is your team measuring and surfacing case staleness beyond standard SLA fields? Are you using Now Assist, custom AI agents, or third-party tools to detect and act on neglected cases? Have you built custom states, list views, or dashboards that meaningfully changed agent behavior? Any workflow or notification patterns that worked, or didn't? If you tried AI for this and it didn't pan out, what got in the way?

Trying to avoid the "just coach harder" answer because it isn't scaling. Open to anything, no idea is too small. Thanks in advance.

1 REPLY 1

sanjaygupta
Mega Guru

I’d start by tracking the time since the last customer-facing update, rather than relying on case state or SLA alone. In CSM, you can flag stale cases with a report and scheduled notifications before introducing AI.

  • Create a report on the Customer Service Case table using fields such as State, Assigned to, Priority, Updated, and customer-facing activity fields.
  • Check whether your instance tracks the last customer-visible update. If not, consider adding a dedicated date/time field that updates only when an agent sends a customer-visible comment or communication.

  • Build a report or dashboard highlighting open cases with no customer-facing update for 3, 5, or 7 days, depending on your support commitments.

  • Use Flow Designer to notify the assigned agent when a case crosses your chosen threshold, and escalate it to the manager if it remains untouched.

  • Exclude cases where the customer is expected to respond, or set a separate threshold for those cases to avoid unnecessary alerts.

  • Once the tracking works reliably, explore Now Assist or AI agents to summarize stale cases and draft follow-up messages for agent review.

 

One important point: sys_updated_on tracks any record update, not just a customer-facing response. An internal work note or background process can make a case look active even when the customer hasn’t received an update.

 

I’d test this in a sub-production instance first and keep customer-facing messages under agent approval until the workflow is proven. Exact fields and automation options can vary by CSM release and configuration.

 

Hope you find this helpful, for more info you can connect at