Intraday Insights Dashboard

What it does

Real-time operational dashboard for supervisors to monitor queue health, agent performance, and customer satisfaction on the same day without waiting for end-of-day reports. Displays queue volume, average handle time, agent availability, SLA breach rates, CSAT scores, and conversation status. Data refreshes every few minutes.

Key facts

Data refreshes every 5–10 minutes (not true real-time, but close enough for tactical decisions)
KPIs included: queue length, average queue wait, average handle time (AHT), agent availability state distribution, SLA breach count/percentage, CSAT score, conversation completion rate
Supports filtering by queue, agent group, channel, or time period (today, last 4 hours, last hour) to drill down on bottlenecks
Agent-level and queue-level views; supervisors can pivot between team and individual performance
Differs from Real-Time Analytics (enterprise reporting tool) in scope and refresh rate; Intraday is operational, Real-Time Analytics is strategic
Accessible from the supervisor dashboard; no additional licensing required beyond Contact Center license
Supports exporting snapshots for shift handoff or escalation documentation

When to use / skip

Use during live operations to spot staffing gaps, detect SLA breaches, monitor CSAT trends, and make intraday adjustments. Check it several times per shift (start, mid-shift, end). Skip if you've got a small team (<20 agents) or stable demand with minimal intraday swings. Don't use it for compliance reporting or exec presentations, use Real-Time Analytics for that.

Configuration decisions

Dashboard refresh frequency: Accept default 5–10 minutes or adjust based on operational tempo (faster refresh = more server load)
KPI thresholds and alerts: Configure threshold values for SLA warning (e.g., 80% breach rate triggers alert), CSAT warning (e.g., <4.0 score), queue depth alert (e.g., >10 waiting)
Visible metrics by role: Decide what supervisors see (full dashboard) vs. team leads (their team only) vs. quality analysts (CSAT/SLA only)
Queue/channel grouping: Pre-define queue hierarchies so supervisors can toggle between aggregated and granular views without custom queries
Intraday staffing model: Document expected load patterns by hour (peak hours, lunch coverage) to inform supervisor decision-making
Alert routing: Configure if/how low CSAT or SLA breach alerts notify supervisors (dashboard notification, email, SMS)

Gotchas

It's a snapshot, not analysis. A spike in queue length could be staffing, outage, or demand: you still need to diagnose.
Data is 5–10 minutes stale. Decisions based on "current" numbers are already behind by the refresh lag.
CSAT skew. Only completed surveys count; incomplete ones are excluded, which pushes the score up artificially.
SLA rates improve late shift. If the SLA target is 24 hours and it's 11pm, fewer conversations can breach today anyway.
Queue depth ≠ understaffing. Could be longer AHT, system delays, or complexity shifts.
Aggregation hides imbalance. An "80% service level" across a queue might hide one agent handling all the complex calls while others sit idle on simple ones.

Consultant notes

The 5–10 minute refresh lag matters when supervisors try to use this for real-time intervention. It's an operational tool for tactical decisions, not a true live feed. Set that expectation in supervisor training or they'll assume the numbers are current when they're not.
CSAT skew from only counting completed surveys is worth flagging to whoever owns KPI definitions. If leadership benchmarks against the intraday CSAT score, they're looking at an optimistic subset. Make sure that's understood before targets are set.
"Aggregation hides imbalance" is the one to raise in supervisor training. An 80% service level across a queue can hide one agent carrying all the complex calls while others handle simple ones. Supervisors need to drill into agent-level views, not just queue totals.

Source last updated: 2026-04-30 | Worth revisiting when staffing model changes, SLA or CSAT baselines are updated, or supervisors give feedback on metric usefulness