Datadog

Datadog's proprietary agent tightly couples applications to its ecosystem. While it accepts OpenTelemetry, advanced APM features still require the proprietary agent. Migration away requires complete re-instrumentation, and rebuilding dashboards, alerts, and data pipelines from scratch.

Datadog's log indexing charges create a perverse incentive: teams must choose between comprehensive logging (high cost) and reduced cost (limited visibility). Indexing only 20% of logs to cut costs means 80% of data is invisible during incidents precisely when full visibility is needed most. This forces budget-constrained teams to strategically under-log, increasing incident resolution times.

Teams report persistent 1-hour delays in real-time data updates from Datadog, lasting 3–4 months. In high-ingest pipelines, bursty microservice deployments can trigger metadata spikes that inflate queue sizes by 400% within minutes, causing missed alerting windows and degraded user experience without proper traffic shaping and rate-limiting.

Datadog's modular, per-dimension pricing model (per-host, per-GB logs, per-million-events, per-session) makes billing unpredictable and difficult to forecast. Teams experience bills 35% higher than estimates, and costs spiral as infrastructure scales, creating an ongoing operational burden to manage expenses.

When the Datadog agent is not configured for proxy usage, it cannot communicate with the Datadog cloud service, resulting in missed or delayed data collection and inability to access external resources.

Organizations managing 300+ customers with multiple instances/apps in Datadog face difficulties controlling access, enforcing privacy settings, and splitting usage/costs per customer. Lack of granular access control and cost customization makes multi-tenant deployments operationally complex and costly to manage.

Datadog makes it too easy to create monitors without guardrails. Teams quickly accumulate hundreds of alerts (300+ monitors reported) with no built-in alert quality scoring or deduplication. Reaching a healthy signal-to-noise ratio requires significant manual tuning over months.

Datadog's data observability is infrastructure-focused, detecting pipeline failures and schema changes but lacking business-aware context to understand data content. This is inadequate for data-centric industries like FinTech and Healthcare where data quality is critical.

Datadog's extensive feature set and integration options overwhelm first-time users. Setting up custom metrics and alerts requires deep product knowledge. Developers must navigate complex documentation to configure APM, trace collection, and integrations (e.g., environment variables for ddtrace, RabbitMQ compatibility), leading to mistakes and configuration headaches.

In complex distributed systems, identifying the root cause of performance issues requires correlating data across network latency, database queries, and third-party services. Without comprehensive monitoring and correlation tools, developers may spend hours or days troubleshooting issues that could be quickly resolved. Finding the right metric among massive data volumes is like 'looking for a needle in a haystack.'

27% of reported ingestion failures stem from agent API mismatches. Comprehensive integration testing requires container orchestration (Kubernetes, Docker Swarm) with multiple plugin versions, but many teams lack resources for this. 21% higher incident rates occur post-major infrastructure shifts without dedicated integration audits, requiring cross-functional response teams and continuous validation.

Without proactive monitoring of storage growth per topic/service and auto-scaling thresholds, sudden workload spikes cause partition bottlenecks and data loss. Schema evolution and versioning practices are critical; integrating schema evolution tools decreases downtime risk by 60% vs. ad hoc migrations, but many teams lack this infrastructure.

When hostnames are dynamically assigned and change frequently, Datadog struggles to accurately track and differentiate between metrics and logs. Multiple services on a single host compound this problem.

Datadog automatically ingests and charges for recognized GenAI attributes in OpenTelemetry spans by default. To avoid these charges, engineers must manually configure the OpenTelemetry Collector or Datadog Agent to drop/mask GenAI-specific attributes using transform processors—there is no simple UI toggle. This configuration trap is non-obvious and adds complexity.

Datadog's query syntax, dashboard creation, and monitor configuration assume deep familiarity with metrics and distributed systems. Non-engineers (product managers, support teams) struggle with log exploration and dashboard building despite Notebooks and saved views, whereas competitors invest more in accessibility.

As a closed SaaS platform, Datadog offers minimal flexibility for custom telemetry processing or monitoring unsupported technologies. Teams must rely on Datadog's roadmap for new features, with no ability to modify platform internals.

Datadog's graphical user interface suffers from slow load times when drilling deep into subjects and lacks caching optimization. Dashboards feel cluttered and overwhelming for new users; navigation is non-intuitive. Default dashboards don't help teams ramp up faster, and session replay features are clunky. Minor issues like unit display and search syntax cumberousness add friction.

Custom dashboards feature has limited customization options, inflexible layout system, and cannot be shared with non-Sentry users without screenshots. Notably lags behind competitors like Datadog and Grafana in capability and polish.

Datadog agent installation and configuration is not straightforward, requiring understanding of agent architecture. Agents consume measurable CPU and memory overhead on hosts/pods, which is problematic in resource-constrained environments.

During outages, Datadog provided frequent updates (hourly or more) but many were copy-pastes of previous messages offering no new information (e.g., 14 consecutive updates using the same phrase about delayed data ingestion). These updates technically satisfied demand for frequent communication but provided no practical value to customers trying to understand issue status and impact.