Track the full chain — delivery, authentication, reputation signals, and processing. When something changes, drill down to isolate
the provider, domain, campaign, mailbox, or time window that caused it.
Deliverability
Understand where mail lands and why. Track placement and outcomes across providers, and connect changes back to list, content, and volume.
- Inbox vs spam placement trends, by provider and mailbox
- Acceptance, deferrals, hard / soft bounces, and reason grouping
- Complaint and unsubscribe rates with anomaly detection
Authentication
See the health of SPF / DKIM / DMARC at a glance and catch drift early (broken selectors, expired keys, SPF flattening issues, misalignment).
- Pass / fail and alignment trends over time
- Failure reasons by domain, selector, and source
- Alerting for sudden changes after deployments or DNS edits
SMTP & provider response codes
Turn raw 4xx / 5xx into actionable categories so you can separate throttling from policy blocks, and transient from permanent failures.
- Provider-specific response clustering (e.g. greylisting vs blocking)
- Retry cadence visibility: when/where messages recover
- “New code detected” alerts for emerging issues
Latency & throughput
Measure time-to-inbox across the entire pipeline — not just “sent”. Useful for high-volume bursts, queue spikes, and downstream slowness.
- Queue time, send time, and time-to-provider acceptance
- p50 / p95 / p99 and tail latency tracking
- Throughput and backlog indicators during peak sends
Reputation & list hygiene
Early warning signals for deliverability drift. Correlate reputation-sensitive metrics to segments, acquisition sources, and campaigns.
- Bounce / complaint spikes by list segment and source
- Suppression effectiveness and “repeat offender” detection
- Engagement trend shifts that precede placement changes
Processing & parsing
“Delivered” isn’t the end. Track what happens after delivery: inbound parsing, webhook processing, retries, and downstream confirmations.
- Inbound parsing errors (format changes, malformed payloads)
- Webhook latency, retry storms, and failure reasons
- Delivered-but-not-processed alerts with traceable IDs