How Dealer Blue Sky cut monitoring costs by 75% after leaving Datadog
Dealer Blue Sky builds software for automotive dealerships, handling millions of transactions a day across 12 core services on self-managed hardware. The team built on Better Stack from day one rather than migrating later, and now runs logs, metrics and heartbeats in one place for under $350 a month.
- Industry
- Automotive SaaS
- Products used
- Logs, metrics, heartbeats
- Switched from
- Datadog
When observability becomes overkill
Before Dealer Blue Sky, Principal Engineer Steffan Perry spent years managing large-scale observability stacks at a global cybersecurity firm. That experience shaped what not to do.
“Datadog was powerful, but the cost structure was unpredictable. Every new capability was a separate add-on, and forecasting spend was nearly impossible.”
When the company launched, the team wanted none of that overhead. They needed one platform that could scale with their services without multiplying cost or complexity.
Starting strong with Better Stack
Instead of migrating later, Dealer Blue Sky built on Better Stack from day one. Copy and paste installation and clear documentation meant no outside support was needed.
Over the past 18 months the team moved from AWS to self-managed hardware. Today they run 12 core services across multiple physical servers, with 12 to 30 Docker containers per host. The Docker overview makes it easy to spot anomalies, isolate the problem container and resolve incidents in minutes.
Heartbeats that save time
The event-based architecture relies on background jobs that must trigger reliably. In AWS CloudWatch, every alert had to be configured and tested by hand. With heartbeats, the team creates one in the dashboard, maps it in code and stops worrying about whether jobs ran. A simple feature that saves hours every month.
When every second counts
When customers reported slow dashboards, logs pointed to a sluggish customers service. Correlating metrics and logs side by side, the team spotted unusually high CPU on the insights service and traced it to a new bug causing excessive retries, even on successful jobs.
“Having metrics, app logs and database logs in one place made root-cause analysis straightforward. What could have taken hours with fragmented tools took minutes.”
What it costs now
Monitoring stays under $350 a month, against an estimated $1,500 to $1,800 for a comparable Datadog setup. Nothing gets rationed to stay in budget: no trimmed retention, no dropped logs, no re-architected alerts.
“From spinning up a new host to seeing metrics in Better Stack now takes just seconds.”
Logs, metrics, traces, uptime and incidents in one platform.