Datadog (DDOG)
Unifies cloud observability, security, and developer workflows on one data platform.
Something off on this page? Send us feedback.
Datadog is a shared control room for companies' software: it watches the machines, records what happened, follows slow or broken requests, and increasingly helps security teams and automated helpers act on problems. Infrastructure and logs still pay nearly all the bills. The next chapter is selling more tools to the same large customers, especially around security, data quality and artificial intelligence, while usage can rise or fall with their computing needs.
Item facts: FY2025 · year ended Dec 31, 2025, from filings, earnings calls and company pages.
Judgment weights, not filed revenue
The band summarizes business focus and direction. ~ marks estimates.
9 in detail · 10 more below
Infrastructure Monitoring
The base layer watches servers and cloud resources. Its annual subscription pace topped $1.6 billion in December 2025; what matters now is whether customers keep adding machines faster than they trim cloud bills.
Competes with Infrastructure Observability (Dynatrace) · Amazon CloudWatch Infrastructure Monitoring (AWS)
In plain English
Every modern app sits on a pile of computing equipment. Datadog watches each piece and warns the people running it when something is overloaded, disconnected or behaving strangely — much like a dashboard full of warning lights for a large fleet.
The customer pays as more servers and other computing resources are watched. That makes this Datadog's largest business: a growing online service usually adds more things to monitor, but a customer that shuts systems down or cuts its cloud bill can reduce what it pays.
Log Management
A searchable diary of what machines did and when. Its annual subscription pace passed $1 billion in December 2025, while cheaper storage choices and customers deleting unneeded records can restrain the bill.
Competes with Elastic Observability Logs (Elastic) · Splunk Cloud Platform / Log Observer Connect (Cisco Splunk)
In plain English
Computers constantly scribble activity notes: a login worked, a payment failed, a service stopped. Log Management gathers those notes so an engineer or security team can search backward and find the moment trouble began.
Customers pay for how much they send, keep and search. Flex Logs lets them store a large pile cheaply and pay more to examine selected parts; another option leaves the pile in the customer's own cloud. More activity creates more records, but careful cleanup and cheaper storage can lower Datadog's take.
Application Performance Monitoring
This follows a request through an application until it succeeds or breaks. Its December 2025 annual subscription pace was roughly $700 million or more; missing records or aggressive data trimming reduce its usefulness.
Competes with Application Observability (Dynatrace) · APM 360 (New Relic)
In plain English
A customer taps “pay,” but the request may pass through dozens of small software services before a receipt appears. This product marks the route like breadcrumbs, showing engineers which step was slow or failed and tying that step back to the code and machines involved.
Engineering teams pay for the machines watched and the request records they send. More complicated applications create more paths worth following. The trade-off is volume: if a customer records too little to save money, the missing path may be the one needed to solve an outage.
Digital Experience Monitoring
Checks whether websites and apps actually work for users. The family's annual subscription pace exceeded $300 million in September 2025; privacy limits and the cost of recording many visits shape growth.
Competes with Digital Experience Monitoring and Synthetic Monitoring (Dynatrace) · Browser Monitoring and Synthetics (New Relic)
In plain English
This is the mystery shopper for a website or app. It watches real visits, can replay what happened before an error, and sends automatic visitors through important steps such as signing in or checking out.
Website and app teams pay according to the number of visits watched, replays kept or automatic tests run. They keep paying because a page can look healthy in the back room while customers are staring at a broken button. Recording too many visits gets expensive, though, and personal information has to be handled carefully.
Cloud Security
Finds risky settings, weak access and suspicious activity across cloud systems. The suite passed a $100 million annual subscription pace in June 2025 and was still accelerating from a small base.
Competes with Wiz Cloud Security Platform (Wiz) · Cortex Cloud / Prisma Cloud (Palo Alto Networks)
In plain English
The guard desk for software running in the cloud. Cloud Security checks who can reach what, looks for doors left open by mistake, scans code for known weaknesses and sorts the flood of activity records for signs of an attack.
Security teams pay to cover more systems, events and developers. Datadog can reuse information already collected for monitoring, which makes adding security easier for an existing customer. The product only works well when customers grant broad enough access, and those permissions must still respect where sensitive data is allowed to live.
Data Observability
Watches business data as it moves and flags missing, late or strange results. Database Monitoring neared a $50 million annual subscription pace in March 2025, so the broader family is promising but small.
Competes with Data Observability Platform (Monte Carlo) · Data Observability Platform (Bigeye)
In plain English
Bad data can quietly spoil every decision made from it. This product follows information through databases and scheduled jobs, checks whether tables arrive complete and on time, and shows where a bad result came from — like tracing a spoiled ingredient back through a kitchen.
Data teams pay to watch more databases and data-moving jobs. Datadog added Metaplane's automatic problem spotting and source tracking to its own database tools. Growth depends on connecting to many systems without copying information that a customer is not allowed to move.
AI Observability
Roughly 750 AI-focused customers were using Datadog across their systems by June 2026. Their spending also lands in infrastructure and application monitoring, so this is not a clean standalone business.
Competes with Arize AX / Phoenix (Arize AI) · LangSmith (LangChain)
In plain English
AI software can give a bad answer without crashing. AI Observability records the questions, answers, costs and steps taken by models and automated helpers, then connects a problem back to the surrounding application and machines. Think of it as a flight recorder for software that makes decisions.
Teams pay as they send more of these records and evaluations into Datadog. The product helps them compare versions, control cost and catch odd behavior after launch. Its sales are hard to isolate because the same AI customer also pays Datadog to watch servers and applications.
Bits AI
Software helpers investigate alerts, suggest fixes and can take guarded actions. Datadog sells their use alongside the wider platform, but no separate sales figure yet shows how much customers will pay for them.
Competes with Dynatrace Intelligence and autonomous operations (Dynatrace) · PagerDuty Advance (PagerDuty)
In plain English
Bits is Datadog's set of automated helpers for operations, security, code and testing. One can gather clues after an alert; another can suggest or carry out a repair. They resemble apprentice mechanics who know the whole workshop but still need permission before touching important controls.
Customers pay for how much the helpers are used and for the monitoring activity around them. Slack, incident tools and Datadog's own records give them context. The business case depends on trust: companies must believe the suggestions are useful and that every action can be checked and reversed.
Datadog Platform
At June 30, 2026, customers had committed $3.471 billion of future use, most expected within two years. The catch is that actual usage can change before all of that becomes sales.
Competes with Application Observability (Dynatrace) · Splunk Cloud Platform / Log Observer Connect (Cisco Splunk) · APM 360 (New Relic)
In plain English
One subscription can start tiny and spread. A company may connect one service, then add logs, application paths, user visits and security checks for more teams. Datadog keeps the information in one place, so each added tool becomes more useful beside the others.
Most customers sign annual deals, though some promise several years. They may pay a fixed amount or a bill that changes with use, rather like a utility plan with a commitment and extra charges. That creates repeat business, but customers can still optimize their systems and consume less than Datadog hoped.
Infrastructure MonitoringThe base layer watches servers and cloud resources. Its annual subscription pace topped $1.6 billion in December 2025; what matters now is whether customers keep adding machines faster than they trim cloud bills.The base layer watches servers and cloud resources. Its annual subscription pace topped $1.6 billion in December 2025; what matters now is whether customers keep adding machines faster than they trim cloud bills.
In plain English
Every modern app sits on a pile of computing equipment. Datadog watches each piece and warns the people running it when something is overloaded, disconnected or behaving strangely — much like a dashboard full of warning lights for a large fleet.
The customer pays as more servers and other computing resources are watched. That makes this Datadog's largest business: a growing online service usually adds more things to monitor, but a customer that shuts systems down or cuts its cloud bill can reduce what it pays.
Competes with Infrastructure Observability (Dynatrace) · Amazon CloudWatch Infrastructure Monitoring (AWS)
Log ManagementA searchable diary of what machines did and when. Its annual subscription pace passed $1 billion in December 2025, while cheaper storage choices and customers deleting unneeded records can restrain the bill.A searchable diary of what machines did and when. Its annual subscription pace passed $1 billion in December 2025, while cheaper storage choices and customers deleting unneeded records can restrain the bill.
In plain English
Computers constantly scribble activity notes: a login worked, a payment failed, a service stopped. Log Management gathers those notes so an engineer or security team can search backward and find the moment trouble began.
Customers pay for how much they send, keep and search. Flex Logs lets them store a large pile cheaply and pay more to examine selected parts; another option leaves the pile in the customer's own cloud. More activity creates more records, but careful cleanup and cheaper storage can lower Datadog's take.
Competes with Elastic Observability Logs (Elastic) · Splunk Cloud Platform / Log Observer Connect (Cisco Splunk)
Application Performance MonitoringThis follows a request through an application until it succeeds or breaks. Its December 2025 annual subscription pace was roughly $700 million or more; missing records or aggressive data trimming reduce its usefulness.This follows a request through an application until it succeeds or breaks. Its December 2025 annual subscription pace was roughly $700 million or more; missing records or aggressive data trimming reduce its usefulness.
In plain English
A customer taps “pay,” but the request may pass through dozens of small software services before a receipt appears. This product marks the route like breadcrumbs, showing engineers which step was slow or failed and tying that step back to the code and machines involved.
Engineering teams pay for the machines watched and the request records they send. More complicated applications create more paths worth following. The trade-off is volume: if a customer records too little to save money, the missing path may be the one needed to solve an outage.
Competes with Application Observability (Dynatrace) · APM 360 (New Relic)
Digital Experience MonitoringChecks whether websites and apps actually work for users. The family's annual subscription pace exceeded $300 million in September 2025; privacy limits and the cost of recording many visits shape growth.Checks whether websites and apps actually work for users. The family's annual subscription pace exceeded $300 million in September 2025; privacy limits and the cost of recording many visits shape growth.
In plain English
This is the mystery shopper for a website or app. It watches real visits, can replay what happened before an error, and sends automatic visitors through important steps such as signing in or checking out.
Website and app teams pay according to the number of visits watched, replays kept or automatic tests run. They keep paying because a page can look healthy in the back room while customers are staring at a broken button. Recording too many visits gets expensive, though, and personal information has to be handled carefully.
Competes with Digital Experience Monitoring and Synthetic Monitoring (Dynatrace) · Browser Monitoring and Synthetics (New Relic)
Cloud SecurityFinds risky settings, weak access and suspicious activity across cloud systems. The suite passed a $100 million annual subscription pace in June 2025 and was still accelerating from a small base.Finds risky settings, weak access and suspicious activity across cloud systems. The suite passed a $100 million annual subscription pace in June 2025 and was still accelerating from a small base.
In plain English
The guard desk for software running in the cloud. Cloud Security checks who can reach what, looks for doors left open by mistake, scans code for known weaknesses and sorts the flood of activity records for signs of an attack.
Security teams pay to cover more systems, events and developers. Datadog can reuse information already collected for monitoring, which makes adding security easier for an existing customer. The product only works well when customers grant broad enough access, and those permissions must still respect where sensitive data is allowed to live.
Competes with Wiz Cloud Security Platform (Wiz) · Cortex Cloud / Prisma Cloud (Palo Alto Networks)
Data ObservabilityWatches business data as it moves and flags missing, late or strange results. Database Monitoring neared a $50 million annual subscription pace in March 2025, so the broader family is promising but small.Watches business data as it moves and flags missing, late or strange results. Database Monitoring neared a $50 million annual subscription pace in March 2025, so the broader family is promising but small.
In plain English
Bad data can quietly spoil every decision made from it. This product follows information through databases and scheduled jobs, checks whether tables arrive complete and on time, and shows where a bad result came from — like tracing a spoiled ingredient back through a kitchen.
Data teams pay to watch more databases and data-moving jobs. Datadog added Metaplane's automatic problem spotting and source tracking to its own database tools. Growth depends on connecting to many systems without copying information that a customer is not allowed to move.
Competes with Data Observability Platform (Monte Carlo) · Data Observability Platform (Bigeye)
AI ObservabilityRoughly 750 AI-focused customers were using Datadog across their systems by June 2026. Their spending also lands in infrastructure and application monitoring, so this is not a clean standalone business.Roughly 750 AI-focused customers were using Datadog across their systems by June 2026. Their spending also lands in infrastructure and application monitoring, so this is not a clean standalone business.
In plain English
AI software can give a bad answer without crashing. AI Observability records the questions, answers, costs and steps taken by models and automated helpers, then connects a problem back to the surrounding application and machines. Think of it as a flight recorder for software that makes decisions.
Teams pay as they send more of these records and evaluations into Datadog. The product helps them compare versions, control cost and catch odd behavior after launch. Its sales are hard to isolate because the same AI customer also pays Datadog to watch servers and applications.
Competes with Arize AX / Phoenix (Arize AI) · LangSmith (LangChain)
Bits AISoftware helpers investigate alerts, suggest fixes and can take guarded actions. Datadog sells their use alongside the wider platform, but no separate sales figure yet shows how much customers will pay for them.Software helpers investigate alerts, suggest fixes and can take guarded actions. Datadog sells their use alongside the wider platform, but no separate sales figure yet shows how much customers will pay for them.
In plain English
Bits is Datadog's set of automated helpers for operations, security, code and testing. One can gather clues after an alert; another can suggest or carry out a repair. They resemble apprentice mechanics who know the whole workshop but still need permission before touching important controls.
Customers pay for how much the helpers are used and for the monitoring activity around them. Slack, incident tools and Datadog's own records give them context. The business case depends on trust: companies must believe the suggestions are useful and that every action can be checked and reversed.
Competes with Dynatrace Intelligence and autonomous operations (Dynatrace) · PagerDuty Advance (PagerDuty)
Datadog PlatformAt June 30, 2026, customers had committed $3.471 billion of future use, most expected within two years. The catch is that actual usage can change before all of that becomes sales.At June 30, 2026, customers had committed $3.471 billion of future use, most expected within two years. The catch is that actual usage can change before all of that becomes sales.
In plain English
One subscription can start tiny and spread. A company may connect one service, then add logs, application paths, user visits and security checks for more teams. Datadog keeps the information in one place, so each added tool becomes more useful beside the others.
Most customers sign annual deals, though some promise several years. They may pay a fixed amount or a bill that changes with use, rather like a utility plan with a commitment and extra charges. That creates repeat business, but customers can still optimize their systems and consume less than Datadog hoped.
Competes with Application Observability (Dynatrace) · Splunk Cloud Platform / Log Observer Connect (Cisco Splunk) · APM 360 (New Relic)
Named in filings, launches and programs
- Software DeliveryProduct lineTracks whether software tests and releases succeed; it sits among the smaller products outside the six measured families.
- Service ManagementProduct lineCoordinates alerts, on-call staff, incidents and routine workflows, with pricing driven mainly by users and completed tasks.
- Eppo by DatadogBrandAcquired in May 2025; helps product teams test which changes work and switch features on or off.
- Metaplane by DatadogBrandAcquired in April 2025; spots broken data and shows where it came from, strengthening Data Observability.
- Adaptive ML / Datadog AI ResearchProduct · Pre-revenueA June 2026 acquisition adding research expertise for training automated systems through feedback; no product sales were disclosed.
- Datadog for GovernmentService · RampingCleared in May 2026 for sensitive U.S. federal work, opening government systems that require stricter security approval.
- Largest-AI-company renewalCustomer programAn unnamed AI-focused customer renewed a nine-figure deal across 17 products; a coming usage cut was already included in guidance.
- Online-media own-cloud log programCustomer program · RampingAn unnamed media company signed for more than $30 million, including massive log storage kept in its own cloud.
- Large-bank monitoring consolidationCustomer programAn unnamed bank signed a three-year contract worth more than $60 million to consolidate its software monitoring.
- FY2025 Q4 large-contract cohortCustomer program · RampingEighteen contracts topped $10 million of total value in the quarter, including two above $100 million; counterparties remain unnamed.
Software DeliveryProduct line
Tracks whether software tests and releases succeed; it sits among the smaller products outside the six measured families.
Service ManagementProduct line
Coordinates alerts, on-call staff, incidents and routine workflows, with pricing driven mainly by users and completed tasks.
Eppo by DatadogBrand
Acquired in May 2025; helps product teams test which changes work and switch features on or off.
Metaplane by DatadogBrand
Acquired in April 2025; spots broken data and shows where it came from, strengthening Data Observability.
Adaptive ML / Datadog AI ResearchProduct · Pre-revenue
A June 2026 acquisition adding research expertise for training automated systems through feedback; no product sales were disclosed.
Datadog for GovernmentService · Ramping
Cleared in May 2026 for sensitive U.S. federal work, opening government systems that require stricter security approval.
Largest-AI-company renewalCustomer program
An unnamed AI-focused customer renewed a nine-figure deal across 17 products; a coming usage cut was already included in guidance.
Online-media own-cloud log programCustomer program · Ramping
An unnamed media company signed for more than $30 million, including massive log storage kept in its own cloud.
Large-bank monitoring consolidationCustomer program
An unnamed bank signed a three-year contract worth more than $60 million to consolidate its software monitoring.
FY2025 Q4 large-contract cohortCustomer program · Ramping
Eighteen contracts topped $10 million of total value in the quarter, including two above $100 million; counterparties remain unnamed.