How Datadog pricing actually works
Datadog charges across several axes simultaneously: 15 USD per host per month for infrastructure, 1.27 USD per ingested log GB, 0.05 USD per custom metric per month, plus separate APM, RUM, security and synthetic monitoring lines. A modest 20-host cluster with logs and APM easily passes 1,500 USD per month — and bills grow with usage, not with users.
Where the bill surprise comes from
Most teams underestimate two line items: custom metrics and log ingest. A single Kubernetes cluster with default Prometheus scrapes exports thousands of custom metrics. Application logs at debug level easily hit hundreds of GB per month. Both are billed by consumption, not by plan.
Who really needs Datadog
For most teams under 100 hosts focused on uptime, alerts and customer status pages, Datadog is overkill — you pay for capabilities you never use.
- •You run a large fleet (200+ hosts) and need APM with distributed tracing
- •You have a dedicated observability team to tune ingest rules
- •You need full security posture management alongside monitoring
Cheaper alternatives that cover 80 percent of the use case
Tools like AppStatus, Better Stack and Pingdom handle the core monitoring outcomes (uptime, SSL, Kubernetes health, status pages, incident response) at a fraction of the cost. AppStatus specifically focuses on predictable plan-based pricing — so your 1,500 USD Datadog bill drops to roughly 50 to 200 USD depending on plan, with no consumption surprises.
How to migrate without losing visibility
Most teams complete this in one engineer-day. The savings start the next billing cycle.
- •Make a list of every Datadog monitor that has fired in the last 90 days
- •Recreate each as an AppStatus monitor (HTTP, port or heartbeat)
- •Set up notification channels (Slack, Email, SMS) and rules
- •Run both side by side for a week — confirm AppStatus catches everything Datadog catches
- •Cancel the Datadog seats and downgrade to the lowest tier or off entirely