A 0–3 score across the 8 moats, anchored to comparable companies. It surfaces where Datadog's defensibility is real, where it's thin, and where the story outruns the structure.
Datadog scores 12 out of 24, with four moats at 2 or higher: Workflow, Distribution, Ecosystem and Scale. Workflow is the one durable 3. Datadog's agent sits across a customer's infrastructure, and 41% of customers ran six or more of its products as of September 30, 2026, with gross revenue retention in the mid to high 90s. Its installed base doubles as a distribution channel: every new product, from session replay to product analytics to experimentation, is sold into accounts that already pay. What it does not have is data that compounds across customers. Each customer's telemetry stays that customer's, the same ceiling Snowflake hits.
Datadog now calls itself the leading AI-powered observability and security platform, and its 2026 launches lead with Bits AI agents that investigate alerts, find root causes and take action. The implied moat is the intelligence. Structurally, autonomous incident agents are the layer every observability vendor and coding agent is building, and Datadog has opened its own data to outside agents through an MCP server. What actually holds customers is less exciting: a dozen products wired into how a company runs production, bought on one contract, from a vendor already inside the account.
Each moat is scored 0–3 for structural durability — not for how useful or popular it is. The tier is set by how many moats reach 2+, not by the total.
| Moat | Meter | Score | One-line take | |
|---|---|---|---|---|
| 01 | Data | 1 | Enormous telemetry, but each customer's stays its own. | |
| 02 | Workflow | 3 | A dozen products wired into how production runs. Removal is a migration. | |
| 03 | Regulatory | 1 | FedRAMP High is a real edge in government, but matchable. | |
| 04 | Distribution | 2 | Owned land-and-expand reach. The installed base is the channel. | |
| 05 | Ecosystem | 2 | 1,000+ integrations and a partner marketplace. Real, not yet canonical. | |
| 06 | Network Effects | 1 | More integrations help everyone, but users don't pull each other in. | |
| 07 | Physical / Infra. | 0 | Massive ingestion, on rented cloud. | |
| 08 | Scale | 2 | $4.5B revenue and a 28% margin. Bundling gets cheaper with size. | |
| Total | 12 / 24 |
The tier is set by counting how many moats score 2 or higher — not by the total out of 24. One strong moat isn't a stack.
| Exposed | 0 moats at 2+ | No defenses; open to attack from every direction. |
| Partial | 1–2 moats at 2+ | Some defense, but below the stack threshold. |
| Stacked | 3 moats at 2+ | Stack threshold met; defensibility begins here. |
| Fortified← This company | 4+ moats at 2+ | Multiple moats compound; the strongest state. |
Datadog has 4 moats at 2+, which lands it in Fortified.
Scores are anchored against comparable companies in the same category — not isolated opinion.
Datadog processes trillions of data points an hour, and that volume is real infrastructure. But the data belongs to each customer and sharpens that customer's dashboards and alerts, not the product for everyone else. That makes it a switching-cost argument, not a compounding Data moat. Snowflake sits at the same 1 for the same reason; Gong scores 3 because its corpus trains models that improve for every customer. A 2 or 3 would need cross-customer learning that visibly makes detection or diagnosis better for the next account.
Datadog's agent runs across hosts, containers and services, and teams build monitors, dashboards, on-call rotations and incident response on top of it. As of September 30, 2026, 61% of customers used four or more products, 41% used six or more and 16% used ten or more, and gross revenue retention holds in the mid to high 90s. Leaving means re-instrumenting hosts and services and rebuilding monitors, dashboards, alert routing, on-call schedules and incident runbooks, the same multi-quarter, high-risk migration that puts Snowflake and Salesforce at a 3. This is Datadog's deepest moat.
Datadog for Government earned FedRAMP High certification in 2026, and Datadog describes itself as the only FedRAMP High observability platform with integrated incident response. That opens federal accounts smaller rivals can't serve yet. It is still a certification well-funded competitors can pursue, not a license they structurally can't obtain. Salesforce scores the same 1 for Government Cloud; Stripe's money-transmitter licenses are what a 2 looks like.
Datadog lands with developers, expands with an enterprise sales team, and sells through AWS, Azure and Google Cloud marketplaces and a partner network. About 48% of the Fortune 500 were customers by the end of 2025, and new products are sold into accounts that already pay. A team already paying for Datadog can add product analytics, session replay and experiments from a vendor it already uses, which is exactly how it entered those categories. That is owned reach in the HubSpot sense, a 2. It isn't a 3 because Microsoft, AWS and Google bundle their own monitoring into the clouds Datadog runs on.
Technology vendors maintain integrations with Datadog because their customers run it, and its Marketplace lets partners sell products built on the platform alongside resellers, system integrators and managed service providers. That is a real ecosystem forming around Datadog, the Notion and Clay pattern. It stops short of Snowflake's 3, where partners book over $100M a half through the marketplace, and short of Salesforce's AppExchange, where a labor market exists around the platform.
There is a weak indirect effect: the more customers run Datadog, the more vendors maintain integrations, which makes Datadog easier for the next customer to adopt. That is closer to Ecosystem than a network effect. One company's usage doesn't improve the product for another. Snowflake reaches a 2 through its data-sharing marketplace; Datadog has no equivalent cross-customer exchange, so it sits at the HubSpot and Attio level of 1.
Datadog's ingestion and storage run at a scale few vendors could match, with regional sites and a government environment. But it runs on public cloud that competitors can also rent. Every observability vendor, from Grafana Labs to Dynatrace, sits at 0 here. This isn't a weakness; the moat doesn't fit a software model.
Datadog guides to about $4.5 billion in 2026 revenue, up 35% year over year in the third quarter, with a 28% non-GAAP operating margin and infrastructure monitoring alone past $2 billion in ARR. That size lets it add products like session replay and experimentation at a marginal cost no point vendor can match. That is demonstrated leverage, the ZoomInfo and Salesforce 2. It isn't a 3 because the hyperscalers it runs on are larger and can cross-subsidize monitoring.
Not every moat fits every business. Three buckets narrow the field — they don't name which available moat to pursue. That's the Playbook's job.
Datadog has four real moats, and the deepest one is the workflow its marketing talks about least. What it has to decide is sequencing. The Defensibility Stress Test pressure-tests each of these scores, runs Datadog head-to-head against 3 to 5 players across observability, cloud-native monitoring and product analytics, and identifies which moats are carrying the stack. From there, the Stacking Moats Playbook takes over: matching the stack to a durable archetype and naming the missing leg, if any.
Plus a head-to-head competitive analysis against 3 to 5 players in Datadog's field.
The rigorous follow-on to this scorecard, run on your own company with the full methodology visible.
Thanks — your scorecard and a short email series are on the way, and I'll be in touch within one business day. If you want the rigorous version, the discovery call on the left is the fastest way to start.