CompanyGraph
Datadog Inc.
DDOG · United States
Installs monitoring software across company cloud infrastructure to collect and display everything happening in one place.
Datadog installs a proprietary agent on every server, container, and cloud function a customer runs, and all monitoring data — metrics, logs, and traces — flows through that agent into a single pipeline, so there is no way to get data into the platform without it. Because the agent sits on every host, ripping it out means touching each one individually and rewriting every dashboard, alert, and incident-response rule from scratch — a job that gets bigger the larger the customer's infrastructure is, which is what keeps customers from leaving. The same breadth that makes Datadog hard to leave also makes it hard to maintain: the platform holds live connections to over 750 cloud services across AWS, Azure, and Google Cloud, and every time one of those providers changes how its API works, the affected connector goes dark for all customers running it at once until Datadog's engineers rebuild it. Adding more cloud infrastructure handles the data processing side cheaply, but fixing broken integrations requires engineers who know each provider's quirks in detail, and that kind of specialist knowledge cannot be hired in bulk or automated away.
How does this company make money?
Datadog charges a monthly or annual fee based on how much a customer uses the platform. Customers pay per monitored host, per container, per custom metric tracked, per volume of logs sent in, and per synthetic test run. Customers who want to keep data for longer periods or access more advanced features pay higher rates. All of these charges scale up as the customer's infrastructure grows.
What makes this company hard to replace?
Switching away from Datadog means touching every single server, container, and cloud function that has a Datadog agent installed and reconfiguring each one to send data somewhere else — a task that grows larger the bigger the customer's infrastructure is. Every dashboard a team has built uses Datadog's specific query language and metric naming, which do not transfer to another platform. Every incident response runbook, automated alert, and webhook that engineering teams rely on during outages is wired directly into Datadog's API endpoints and would have to be rebuilt from scratch.
What limits this company?
Datadog currently maintains over 750 live connections to cloud services, databases, and applications. Every time a cloud provider like AWS or Azure releases a new service or retires an old API, someone at Datadog has to rebuild or update that connection. That work requires engineers who deeply understand each provider's specific data formats and security handshakes — and hiring more generic cloud engineers does not clear the backlog. The more services cloud providers release, the longer that maintenance queue grows, regardless of how many customers Datadog already serves.
What does this company depend on?
Datadog cannot run without access to AWS CloudWatch APIs for EC2 and RDS metrics, the Kubernetes API server for container data, Azure Monitor APIs for Azure resource telemetry, Google Cloud Operations APIs for GCP infrastructure monitoring, and OpenTelemetry protocol support for collecting application performance data.
Who depends on this company?
DevOps teams at large companies running infrastructure across AWS, Azure, and GCP at the same time would lose the single view that ties all three together and be forced back to each cloud's native tools separately. Site reliability engineers managing Kubernetes clusters would face dangerous blind spots during outages because their consolidated container and application metrics would no longer exist in one place. Cloud security teams would lose real-time threat detection across their SIEM workflows because unified log collection would stop.
How does this company scale?
Processing telemetry data and rendering dashboards can grow cheaply by adding standard cloud computing resources — that part scales well. What does not scale easily is the engineering work of keeping every integration current. As cloud providers release new services and retire old APIs, the number of connections requiring specialist attention keeps growing, and that work cannot be automated or handed to general-purpose engineers.
What external forces can significantly affect this company?
EU GDPR and data residency rules require that telemetry data stay within specific geographic borders, which conflicts with Datadog's model of aggregating everything into a single global pipeline. When cloud providers like AWS or Azure raise prices for API calls or data transfer, those costs flow directly through to what customers pay to run their monitoring. Federal cybersecurity mandates in the US that require specific log retention periods and incident response capabilities push organizations toward adopting security monitoring tools, which can bring new customers to Datadog but also raise the compliance bar Datadog itself must meet.
Where is this company structurally vulnerable?
If AWS, Azure, or Google Cloud made a sweeping change to a core API — changing how performance data is formatted, shutting down an authentication endpoint, or blocking third-party agents from accessing telemetry — every Datadog customer using that integration would lose visibility at the same moment. If several providers made such changes at the same time, Datadog's integration engineers could not rebuild all the broken connections fast enough, and the breadth of those 750+ integrations, which is normally the company's strongest selling point, would become the source of the problem.