Your CX Metrics Are Moving Right Now. Are You Watching?
NPS, CSAT, sentiment, topic volume — all moving in real time. Pivony monitors every metric in every segment and fires an instant alert the moment something shifts abnormally. No weekly report. No discovery lag.
Why Weekly Reports Are the Wrong Cadence for CX
Problems spread before you see them
A support surge on Tuesday is in your weekly report on Monday. Seven days of customers experiencing a broken journey — before anyone on the team knows.
Dashboards require someone to look
Dashboards are passive. They don't alert you — they wait for you. Most CX issues are discovered by accident, not by design.
Segment anomalies hide in averages
Overall NPS is stable at 41. But VIP NPS dropped from 72 to 54 this week. That is a crisis — but it is invisible in a portfolio-level metric.
Real-Time KPI Monitoring With Intelligent Anomaly Detection
01
Connect your metrics
NPS, CSAT, CES, sentiment scores, topic volume — all flowing into a unified monitoring layer, updated continuously.
02
Define your thresholds
Set segment-level alert rules: 'Alert me if VIP NPS drops more than 5 points in 7 days' or 'Notify if delivery complaints spike above baseline by 30%'.
03
Anomaly detection fires
Pivony detects statistical anomalies even without predefined thresholds — flagging unusual patterns automatically using baseline modeling.
04
Alert reaches the right team
Alerts route to Slack, email, or ticketing systems. The right person knows about the right anomaly — without checking a dashboard.
What You Can Monitor
NPS by segment in real time
Any segment anomaly — instant alert to the responsible owner
CSAT and CES monitoring
Post-interaction scores tracked continuously, not in batches
Sentiment shift detection
Automated flags when sentiment moves abnormally in any topic area
Topic volume spikes
When a complaint topic suddenly generates 3x normal volume — you know immediately
Comparative performance monitoring
Track metrics vs prior period, vs benchmark, vs target
Cross-segment correlation alerts
When one segment's anomaly predicts another's — early warning system
What to Expect from a KPI Monitoring Platform
Common Questions
How does Pivony define an anomaly?＋
Pivony builds a rolling baseline for each metric and segment using historical data. An anomaly is flagged when a metric moves beyond a statistically significant threshold from its baseline — either in absolute terms (e.g., NPS drops 8 points) or relative terms (e.g., sentiment score moves 2 standard deviations from the 30-day average). Teams can also set manual thresholds.
Can I set different alert thresholds for different segments?＋
Yes. VIP NPS and standard NPS can have different alert sensitivities. A 3-point drop in VIP NPS might trigger an immediate escalation, while a 3-point movement in standard NPS triggers only a monitoring flag. You define the rules per segment and per metric.
Where do alerts get sent?＋
Pivony supports alert delivery to email, Slack, Microsoft Teams, Jira, and webhook-compatible ticketing and CRM systems. The routing logic (who receives which alert for which segment) is configurable.
How is this different from just putting a dashboard on a TV screen?＋
Dashboards are passive — they require someone to look at the right moment. Pivony's anomaly detection is active: it monitors continuously, detects deviations automatically, and pushes the signal to the right person before the issue becomes visible on a dashboard.
What is the typical time between an anomaly occurring and an alert being sent?＋
Under 15 minutes for real-time data feeds. For daily batch sources, the alert fires within the same processing cycle as the data update.
Ready to stop discovering CX problems in last week's report?
Request a demo and see real-time anomaly alerts running on your own metrics.