Metric Monitoring
24/7 Anomaly Detection That Automatically Diagnoses the Cause
Fig monitors your metrics on a schedule, detects anomalies with statistical scoring, and automatically triggers root cause analysis — before anyone on your team notices.
The Monitoring Pipeline
Five stages from metric definition to delivered diagnosis — fully automated.
Define
Choose the metrics you care about, set the SQL computation behind each one, and configure anomaly thresholds — absolute, percentage, or z-score based.
Schedule
Set cron-based schedules with timezone support. Fig evaluates your metrics at the cadence you choose — hourly, daily, or custom windows.
Detect
Each evaluation produces an anomaly score using statistical methods. When a score crosses your threshold, Fig flags it immediately.
Investigate
Anomalies automatically enqueue a root cause analysis. Fig traces the causal chain through your knowledge graph to find what drove the change.
Report
A structured RCA report is delivered with contributing factors ranked by impact, supporting evidence, and recommended next steps.
Why Monitoring Without Diagnosis Is Just Noise
Traditional Alerts
A dashboard turns red. Your team scrambles to figure out what happened. They open five tabs, run ad-hoc queries, and spend the morning narrowing down the cause.
By the time they have an answer, the impact has already compounded — and the explanation is often incomplete.
Fig Monitoring
Fig detects the anomaly, immediately enqueues root cause analysis, and delivers a structured report — with contributing factors ranked by impact.
Your team wakes up to the diagnosis, not the alarm. They skip the investigation phase entirely and go straight to action.
Alerts tell you something is wrong. Fig tells you why it is wrong — and does it automatically, on the schedule you set.
Real Monitoring Scenarios
See how automated anomaly detection and root cause analysis work together in practice.
Trigger
Collection rate drops 8% on Tuesday
What Fig Finds
Fig's monitors catch the anomaly within the scheduled window. Automatic RCA traces the decline to provider schedule changes that reduced follow-up capacity.
So that your team sees the specific scheduling changes within hours, not weeks — and can act before the drop compounds.
Trigger
Same-store sales decline across 12 locations
What Fig Finds
Auto RCA decomposes the decline and finds two concurrent drivers: foot traffic patterns shifting and seasonal inventory misalignment at specific locations.
So that your team knows exactly where the decline is concentrated and what correlates — not just that sales are down.
Trigger
ROAS drops below your configured threshold
What Fig Finds
RCA traces the performance drop to a bid strategy change that went live the same day a landing page update increased load time by 1.8 seconds.
So that your team sees the two concurrent changes that caused the drop — and doesn't waste a week testing one variable at a time.
Stop Reacting. Start Diagnosing Automatically.
Set up your first monitor in minutes. Fig watches your metrics around the clock and delivers root cause reports — so your team focuses on decisions, not investigation.