PM Observability
(2026 Edition)
Observability pricing is the hardest call in the category: ingest-based models built around cardinality and retention scale unpredictably with traffic, which is why Datadog's fast growth drew pushback while Honeycomb and Grafana now compete on pricing predictability instead. PMs manage this by watching monthly ingest volume, alert-to-investigation conversion, and custom dashboard adoption, while fighting alert noise and adapting to OpenTelemetry's growing pull away from proprietary agents.
By Naman Goyal · Product manager · Builder of PM Streak · Updated July 3, 2026
5 dynamics and 5 metrics for observability PMs.
Build Observability PM Skills — Free →5 Dynamics
Three pillars — metrics, logs, traces — bundled is sticky but expensive
Pricing is the hardest product decision — cardinality, ingest, retention
Alert noise is the enemy — too many alerts = no alerts
Custom dashboards drive retention — power users build worlds
OpenTelemetry reshaping interop — proprietary agents losing ground
5 Metrics
Monthly ingest volume
Active custom dashboards per account
Alert-to-investigation conversion
Query P95 latency
Seat-based adoption within customer orgs
FAQ
Why is observability pricing so controversial?
Because ingest-based pricing scales unpredictably with traffic. Customers get shocked bills when traffic spikes. Datadog grew fast on this model but faced pushback; Honeycomb and Grafana now compete on pricing predictability. In 2026, expect more value-based and hybrid models to emerge.
Keep learning