Sampleless vs Datadog
Teams switching from Datadog to Sampleless save 50-70% on observability costs while collecting 100% of their data instead of 1-10%. No sampling. No egress fees. No surprises.
The bottom line
Choose Sampleless if you want:
- 100% data for better ML-driven insights
- Predictable flat annual pricing
- Data sovereignty (BYOC architecture)
- No vendor lock-in (OpenTelemetry native)
- Unlimited users included
Datadog may be better if you need:
- 500+ pre-built integrations
- Infrastructure-only monitoring
- Established vendor with large community
- Synthetic monitoring features
Key differences
Pricing Model
Flat annual fee ($100K-$350K)
Based on cloud account footprint, not data volume. No per-GB, per-host, or per-user charges. Unlimited custom metrics included.
Per-host + per-GB + per-metric
Infrastructure starts at $15-23/host/month. APM adds $31/host. Logs cost $0.10/GB ingest + $1.70/million events indexed. Custom metrics $0.05+ each.
Data Collection
100% of all telemetry
Every trace, every metric, every log. Complete data means better ML baselines, more accurate anomaly detection, and evidence that is always there when you need it.
Sampling required at scale
To control egress and storage costs, most teams sample 1-10% of traces. Sampled data limits the accuracy of ML-driven insights and anomaly detection.
Data Location
Your cloud (BYOC)
Telemetry stays in your AWS/GCP/Azure account. No egress costs. Use your existing security controls and compliance certifications.
Datadog infrastructure
All data shipped to Datadog. Transit Gateway and PrivateLink fees apply. Data leaves your security perimeter.
Behavioral Analytics
ALBA dual-score system
Every entity gets an anomaly score (how unusual) and risk score (how concerning). Detects patterns that static thresholds miss.
Rule-based alerting
Threshold-based alerts and anomaly detection available. No entity-level behavioral scoring or risk assessment.
Users
Unlimited, included
Every engineer, SRE, and stakeholder can access the platform. Collaboration is not a cost center.
Per-user APM pricing
APM features require per-user licensing. Teams often limit access to control costs, reducing collaboration.
Vendor Lock-in
OpenTelemetry native, easy to leave
Built on OpenTelemetry and OpenALBA open standards. Your instrumentation works anywhere. If Sampleless is not the right fit, migrate easily.
Proprietary agents and formats
Datadog agents and data formats create switching costs. Migration requires re-instrumentation and workflow changes.
Real cost comparison
See what a typical mid-market deployment actually costs.
Scenario: 500 hosts, 5TB logs/day, 50 engineers, 20 APM services
Datadog (estimated)
Sampleless
Save ~$170,000/year (32%)
While collecting 100% of your data instead of sampling
Estimates based on publicly available Datadog pricing. Actual costs vary based on usage patterns, committed contracts, and negotiated discounts. Your infrastructure costs depend on your cloud provider rates.
Why teams leave Datadog
Datadog is a powerful platform with excellent integrations and a polished UI. But its pricing model creates a fundamental conflict: the more you observe, the more you pay. This leads to predictable problems.
Bill shock is real
One customer was famously billed $65 million in a single quarter. While extreme, unexpected overages are common. Custom metrics, log indexing, and auto-scaling infrastructure can cause bills to spike without warning.
Sampling destroys value
To control costs, most teams sample 90-99% of traces. But anomalies hide in the discarded data. When incidents happen, the evidence needed to diagnose them has often been thrown away. ML baselines trained on sampled data produce worse models.
The custom metrics trap
Custom metrics can account for 50%+ of Datadog bills. Each unique combination of metric name and tags counts separately. A single high-cardinality tag can cause metric counts (and costs) to explode unexpectedly.
Networking costs add up
Shipping terabytes of telemetry to Datadog requires Transit Gateway or PrivateLink. At 10TB/day, networking alone can cost $200-500/day, separate from your Datadog bill.
Feature comparison
| Feature | Sampleless | Datadog |
|---|---|---|
| Distributed tracing | ||
| Log management | ||
| Infrastructure metrics | ||
| Custom metrics | Unlimited | $0.05+/metric |
| APM | ||
| Real user monitoring | Roadmap | |
| Synthetic monitoring | Roadmap | |
| Behavioral analytics | ALBA scoring | Basic |
| 100% data collection | ||
| BYOC deployment | ||
| Cross-cloud federation | Limited | |
| OpenTelemetry native | Supported | |
| Pre-built integrations | 50+ | 500+ |
| Unlimited users | ||
| Flat pricing |
Migration is straightforward
Sampleless is OpenTelemetry-native. If you're already using OTel collectors, migration is a configuration change. If you're using Datadog agents, we provide migration guides and support.
Deploy Sampleless
We deploy the observability stack in your cloud account. Takes 1-2 days.
Run in parallel
Point your OTel collectors at both platforms. Validate data in Sampleless.
Cut over
Once validated, redirect telemetry to Sampleless only. Cancel Datadog.
Send us your Datadog bill
We'll prepare a detailed cost comparison showing exactly how much you can save. If we can't beat your current costs by at least 40%, we'll tell you to stay where you are.