- Mean time to detect (MTTD)
First occurrence → human awareness
The single most useful integration metric and the most commonly omitted. Measure from raw telemetry, not from the ticket. Improvement comes from correlating signals across tools and alerting on absence of expected activity, not from adding thresholds.- Mean time to diagnose
Awareness → correct understanding of cause and scope
Collapses when a recurrence is recognised as one. If this is not falling for repeat faults, incident history is not reaching the person responding.- Mean time to escalate
Diagnosis → counterparty acknowledgement
Usually the slowest phase for smaller buyers. Moved by evidence quality in the first message, not by follow-up frequency. Track it per vendor: the variance tells you which relationships to renegotiate.- Mean time to resolve (MTTR)
First occurrence → verified resolution
Include time spent waiting on a vendor. Excluding it produces a flattering number and removes pressure to improve the escalation path.- Recurrence rate
Share of incidents matching a previous signature
If MTTR falls while this holds steady, the team is getting faster at handling the same fault rather than eliminating it. The more honest measure of progress.- Detection source distribution
Share detected by monitoring, by staff, by customers
Customer-detected share is the clearest indicator of monitoring adequacy. Falling monitoring share over time usually signals alert fatigue rather than improving reliability.- Silent failure ratio
Share of incidents that produced no error signal
Identifies integrations where current instrumentation cannot help. A high ratio argues for reconciliation and volume assertions rather than more alerting.- Operational rework hours
Manual effort downstream of integration defects
Usually the largest cost and almost never measured, because it lands in finance, support and operations rather than engineering. Counting the processes that exist to reconcile two systems is a reasonable proxy.
Put this into practice without the manual overhead
Traxivo keeps the inventory, the timeline and the vendor history current as a by-product of handling the signals your tools already produce.
See how Traxivo works Browse use cases