Why AI workloads expand Datadog's total addressable market
Datadog's business is built on a simple principle: complex distributed systems generate more observable data and require more monitoring than simple monolithic applications. Every time a company moves workloads to the cloud or deploys AI agents, it adds more components to its infrastructure — more services, more APIs, more data pipelines, more failure modes. Each new component is a potential Datadog customer module. The company charges based on usage (hosts monitored, logs ingested, traces sampled), so revenue scales automatically as customers' AI-driven environments grow in complexity.
Datadog raised its 2026 full-year revenue guidance to $4.30-4.34 billion, reflecting robust cloud security adoption and an expanding product footprint into AI observability — specifically monitoring the performance and cost of LLM calls inside production AI applications. That LLM observability module is the company's fastest-growing new product because every enterprise running ChatGPT-style applications needs visibility into token costs, latency, and error rates. It is the same land-and-expand motion that built the core platform applied to the AI layer.
- DDOG charges by usage — more AI workloads mean more hosts, more logs, more traces, and automatically higher revenue without additional sales effort.
- The LLM observability module monitors AI application performance and token costs — the fastest-growing new product in the portfolio.
- Net revenue retention above 120% means existing customers are consistently expanding their Datadog usage year over year.