We primarily use the solution for a variety of purposes, including:

• Watching RUM data for frontend site, using LCP and INP metrics to compare across the old and new architecture to inform rollout decisions.
• Watching APM data for backend services, observing how the backend server reacts (CPU util, memory, requests/second) to make sure the backend can handle the load.
• Using Datadog CCM during our free trial period to get visibility over our AWS spend across accounts and resources and looking at recommendations and acting on those.
• Browsing the service catalog to look at the current state of services that are running and what resources it uses. 

This provides a single place to find monitoring data. Prior to DD, we had some metrics living in New Relic, some in Grafana, and some in Circonus, and it was very confusing to navigate across them. Understanding different query languages is challenging. Here, there's a single UI to get used to, and everything is so sharable.

DD has led to teams making more decisions based on data that they observe about their service metrics and RUM metrics. I've seen decisions get made based on what has been observed in DD, and less based on anecdotal data.

I really enjoyed using CCM since it showed cloud cost data easily next to other metrics, and I could correlate the two.

Across CCM and the rest of Datadog, I like how sharable everything is. It's so easy to share dashboards and links with my teammates so we can quickly get up to speed on debugging/solving an issue.

I also have really enjoyed K8s view of pods and pod health. It's very visual, and as a non-K8s platform owner at my company, I can still observe the overall health of the system. Then I can drill in and have learned things about K8s by exploring that part of the product and talking with the team.

We've had some issues where we had Datadog automatically turned on in AWS regions that we weren't using, which incurred a small but steady cost that amounted to tens of thousands of dollars spent over a few weeks. I wish there was a global setting that lets an admin restrict which regions DD is turned on in as a default setup step.

Sometimes, the APM service dashboard link isn't sharable. I click something in the service catalog, and on that service's APM default view, I try to share a link to that with a teammate, and they reach a blank or error screen. 

I wish there was more organization and detail in the suggestions when I use the query editor. I'm never quite sure when the autofill dropdown shows up if I'm seeing some custom tag or some default property, so I have to know exactly what I'm looking for in order to build a chart. It's hard to navigate and explore using the query autofill suggestions without knowing exactly what tag to look for.

It's been a bit hard to understand how data gets sampled or how many data points a particular dashboard value is using. We've had questions over the RUM metrics that we see and we had to ask for help with how values are calculated, bin sizes, etc to get confidence in our data.

I've used the solution for six months.

I've only been aware of a recent outage that affected the latency of data collection for one of our production tests. Outside of that, the solution seems stable.

The solution seems like it can scale very well and beyond our needs.

Technical support has been stellar. We love working with a team that responds fast, in great detail, and with great empathy. I trust what they say.

Positive

We used New Relic, Grafana, and Circonus. Circonus was flakey, always having downtime and we were always on the phone with them. New Relic and grafana, different metrics lived in either and it was hard for consumers of the data to easily find what they need. And we had licensing issues across the 3 so not everybody could easily access all of them.

I didn't do this portion of the product setup.

In our fast-paced environment, managing and analyzing log data and performance metrics is crucial. That’s where Datadog comes in. We rely on it not just for monitoring but for deeper insights into our systems, and here’s how we make the most of it. 

One of the first things we appreciate about Datadog is its ability to centralize logs from various sources—think applications, servers, and cloud services. This means we can access everything from one dashboard, which saves us a lot of time and hassle. Instead of digging through multiple platforms, we have all our log data in one place, making it much easier to track events and troubleshoot issues.

Before Datadog, we faced the common challenge of fragmented data. Our logs, metrics, and traces were spread across different tools and platforms, making it difficult to get a complete picture of our system’s health. 

With Datadog, we now have a centralized monitoring solution that aggregates everything in one place. This has streamlined our workflow immensely. Whether it’s logs from our servers, metrics from our applications, or traces from user transactions, we can access all this information easily. This unified view has made it simpler for our teams to identify and troubleshoot issues quickly.

In my experience with Datadog, one feature stands out above the rest is the Log Explorer. It has completely transformed the way I interact with our log data and has become an essential part of my daily workflow. 

The user interface is incredibly intuitive. When I first started using it, I was amazed at how easy it was to navigate. The design is clean and straightforward, allowing me to focus on the data rather than getting lost in complicated menus. Whether I’m searching for specific log entries or filtering by certain criteria, everything feels seamless. 

This ease of use allowed me to get up to speed with log management since it's my first time using Datadog.

Interactive tutorials could be a game changer. Instead of just reading about how to use query filters, users could engage with step-by-step guides that walk them through the process. For example, a tutorial could start with a simple query and gradually introduce more complex filtering techniques, allowing users to practice along the way. These tutorials could include pop-up tips and hints that provide additional context or best practices as users work through examples. This hands-on approach not only reinforces learning but also builds confidence in using the tool.

My company has recently made Datadog available to it's software engineers and I personally have been using it for almost a year now.

The soluton is used for full stack enterprise performance monitoring for our primarily cloud-based stack on AWS. We have implemented monitoring coverage using RUM for critical apps and websites and utilize APM (integrated with RUM) for full stack traceability.  

We use Datadog as our primary log repository for all apps and platforms, and the advanced log analytics enable accurate log-based monitoring/alerting and investigations. 

Additionally, we some advanced RUM capabilities and metrics to track and optimize client-side user experience. We track SLO's for our critical apps and platforms using Datadog.

We now have full-stack observability, which allows us to better understand application behavior, quickly alert users about issues, and proactively manage application performance.  

We've seen value by implementing observability coordinated across multiple applications, allowing us to track things like customer shopping and orders across multiple applications and services.  

For critical application launches, we've built dashboards that can track user activity and confirm users are able to successfully utilize new features, tracking user activities in real-time in a war-room situation.  

Datadog is our go-to tool for analyzing, understanding, and investigating application performance and behavior.

APM accurately tracks our service performance across our ecosystem. RUM gives us client-side performance and user experience visibility, and the rate of new features implemented in the Digital Experience area recently has been high. Log analytics give us a powerful mechanism for error tracking, research, and analysis.  

Custom metrics that we've created allow us to track KPIs in real-time on dashboards. All of these have proven valuable in our organization.  Additionally, Datadog product support teams are responsive and have provided timely support when needed.

Agent remote configuration should be provided/improved and streamlined, allowing for config changes/upgrades to be performed via the portal instead of at the host.   

Cost tracking via the admin portal is a bit lacking, even though it has gotten better.  I'm looking for usage trends (that drive cost) across time and better visibility or notifications about on-demand charges.  

Network device and performance monitoring could be improved, as we've faced some limitations in this area.  

The Datadog usage-based cost model, while giving us better transparency, is difficult to follow at times and is constantly evolving.  

I've used the solution for three years.

Support has been responsive and helpful.  

Positive

Pricing is straightforward. That said, it's sometimes difficult to estimate usage volumes.

We evaluated Datadog and New Relic in detail and chose Datadog due to their straightforward and competitive pricing model, and their full coverage of monitoring features that we desired, and an easy-to-use UI.  

We use this solution for enterprise monitoring across a large number of applications in multiple environments like production, development, and testing. It helps us track application performance, uptime, and resource usage in real time, providing alerts for issues like downtime or performance bottlenecks. 

Our hybrid environment includes cloud and on-premise infrastructure. The solution is crucial for ensuring reliability, compliance, and high availability across our diverse application landscape.

Datadog has greatly improved our organization by centralizing all monitoring into one platform, allowing us to consolidate data from a wide range of sources. 

From infrastructure metrics and application logs to end-user experience and device monitoring, everything is now collected and displayed in one place. This has simplified our monitoring processes, improved visibility, and allowed for faster issue detection and resolution. 

By streamlining these operations, Datadog has enhanced both efficiency and collaboration across teams.

Synthetic testing is by far the most valuable feature in our organization. It’s highly requested since the setup process is both quick and straightforward, allowing us to simulate user interactions across our applications with minimal effort. 

The ease of configuring tests and interpreting the results makes it accessible even to non-technical team members. This feature provides valuable insights into user experience, helps identify performance bottlenecks, and ensures that our critical workflows are functioning as expected, enhancing reliability and uptime.

One area where the product could be improved is Application Performance Monitoring (APM). While it's a powerful feature, many in our organization find it difficult to fully understand and utilize to its maximum potential. 

The data provided is comprehensive, yet it can sometimes be overwhelming, especially for those who are less familiar with the intricacies of application performance metrics. 

Simplifying the interface, offering clearer guidance, or providing more intuitive visualizations would make it easier for users to extract valuable insights quickly and efficiently.

I've used the solution for four years.

The solution is very stable. Issues happen once or twice a year and are usually solved before we have any real impact on the service.

Scalability has never been a bottleneck for us; we've never felt any issues here.

Support is slow at the beginning, however, they are much better and responsive now.

Positive

Datadog offered the most consolidated approach to our monitoring needs.

This was a migration project, so it was rather complex.

We implemented the solution with our in-house team.

I'd recommend new users look down the road and decide on at least a three-year plan.

We evaluated AppDynamics and Dynatrace.

We have multiple nodes integrated into our Azure infrastructure and our AKS clusters. These nodes are integrated with traces (as APM hosts).

We also have infrastructure Hosts integrated to see the metrics and the resources of each hosts mainly for Azure VMs and AKS nodes. Additionally, we also have hosts from our VMs in Azure which act as Activemq and we integrate them as messaging queues to show up in the Activemq dashboard.

We have recently added Activemq as containers in the AKS and we are also integrating those as messaging queues to show up in the Activemq dashboard integration 

Logs are great. Having all services with different teams sending the logs to Datadog and having all logs in the same place is very helpful for us to understand what is going on in our app; filtering of the logs a huge help and adding special custom filters is easy, filters are fast. Documentation is better than average, with little room for improvement.

Dashboards are simple, and monitors are very easy to configure and get notified if something is wrong.

With the aggregated logs, we can now see logs from other systems and identify problems in other areas in which we had no visibility before.

Dashboards are the most valuable. We need the observability. We have given the dashboards to a dedicated team to monitor them off working hours and they are reporting whatever they see going red. This helps us since people without any knowledge can understand when there is a problem and when to react and when to inform others by simply looking if the monitor (showing the dashboards) turns up red. 

Traces being connected to each other and seeing that each service is connected through one API call is very helpful for us to understand how the system works.

The monitors need improvement. We need easier root cause analysis when a monitor hits red. When we get the email, it's hard to identify why the trigger has gone red and which pod exactly is to blame in a scenario where the pod is restarting, for example.

Prices are a very difficult thing in Datadog. We have to be very mindful of any changes we make in Datadog, and we are a bit afraid of using new features since, if we change something, we might get charged a lot. For example, if we add a network feature to our nodes, we might get charged a lot simply by changing one flag, even though we are only going to use one small feature for those network nodes. However, due to the fact that we have more than 50 nodes, all of the nodes will be charged for the feature of "Network hosts".

This leads us to not fully utilize the capabilities of Datadog, and it's a shame. Maybe we can have a grace period to test features like a trial and then have datadog stop that for us to avoid paying more by mistake.

I've used the solution for five years.

The solution is stable enough. We found it to be down only a few times, and it's reasonable.

The solution offers very good scalability. When we added more logs and more hosts, we did not notice any degradation in the service.

Support is very good. They answer all of our questions, and with a few emails, we get what we need

Positive

We previously used Elastic. We had to set up everything and maintain it ourselves.

Datadog has very good support and it is not so complicated to set up.

We set up the solution in-house. We integrated everything on our own.

We found the product to be very valuable.

I'd advise others to start small and then integrate more stuff. Be mindful when using Datadog.

We evaluated Splunk and ELK.

Be careful of the costs. Set up only the important things.

We monitor our multiple platforms using Datadog and post alerts to Slack to notify us of server and end-user issues. We also monitor user sessions to help troubleshoot an issue being reported. 

We monitor 3.5 platforms on our Datadog instance, and the team always monitors the trends and Dashboards we set up. We have two instances to span the 3.5 platforms and are currently looking to implement more platform monitoring over time. The user session monitoring is consistent for one of these platforms. 

Datadog has improved our response time and cost-efficiency in bug reporting and server maintenance. We're able to track our servers more fluidly, allowing us to expand our outreach and decrease response time. 

There are many different ways that Datadog is used, and we monitor three and a half platforms on the Datadog environment at this time. By monitoring all of these platforms in one easy-to-use instance, we're able to track the platform with the issue, the issue itself, and its impact on the end user. 

The server monitoring, service monitoring, and user session monitoring are extremely helpful, as they allow us to be alerted ahead of time of issues that users might experience. More often than not, an issue is not only able to be identified, but solved and released before an end user notices an issue. 

We are currently using this as an investigative tool to notice trends, identify issues, and locate areas of our program that we can improve upon that haven't been identified as pain points yet. This is another effective use case. 

I would like to see a longer retention time of user sessions, even if by 24 to 48 hours, or even just having the option to be configurable. By doing this, we're enabled to store user sessions that have remained invisible for a long time, and identify issues that people are working around. 

I would also like to see an improvement in the server's data extraction times, as sometimes it can take up to ten minutes to download a report for a critical issue that is costing us money. Regardless, I am very happy with Datadog and love the uses we have for the program so far.  

I've used the solution for more than four years.

We did not previously use a different solution. 

Our primary use cases include:

• Alert on errors customers encounter in our product. We've set up logs that go to slack to tell us when a certain error threshold is hit.
• Investigate slow page load times. We have pages in our app that are loading slowly and the logs help us figure out which queries are taking the longest time.
• Metrics. We collect metrics on product usage.
• Session replays. We watch session replays to see what a user was doing when a page took a long time to load or hit an error. This is helpful.
  

It's helped us find bugs that customers are experiencing before they're reported to us. Sometimes, customers don't report errors, so being able to catch errors before they're reported helps us investigate before other users find errors

Datadog has helped us investigate slow page loading times and even see the specific queries that are taking a long time to load

Logging lets us see the context around an error. For example, see if a backend service had an error before it surfaced on the frontend.

Dashboards are helpful for reviewing occasionally to get a higher-level overview of what's happening.

The most valuable aspects include: 

• Logging. Being able to view detailed logs helps debug issues.
• Session replays. They are helpful for seeing what a customer was doing before they saw an error or had a slow page load
• Alerting. This is an important part of our on-call process to send alerts to slack when an error threshold is crossed. Alerts/monitors are easy to configure to only alert when we want them to alert.
• Dashboards. It's helpful to pull up dashboards that show our most common errors or page performance. It's a good way to see how the app is performing from a birds-eye-view.

The UI has a lot going on. It should be simpler and have a better way to onboard someone new to using Datadog.

The log querying syntax can be confusing. Usually, I filter by finding a facet in a log and selecting to filter by that facet - but I'm not sure how to write the filter myself

The monitor/alert syntax is also somewhat hard to understand.

Overall, it should be easier to learn how to use the product while you're using the product. Perhaps tooltips or a link to learn more about whatever section you're using.

I've used the solution for two years.

We did not previously use a different solution.

We did not evaluate other options. 

We use the solution for logs, infrastructure metrics, and APM. We have many different teams using it across both product and data engineering.

The solution has improved our observability by giving us rapid log search, a correlation between hosts/logs/APM, and tons of features in one website.

I enjoy the rapid log search. It's such a pleasure to quickly find what you're looking for. The ease of graph building is also nice, and MUCH easier than Prometheus.

It is far too easy to run up huge unexpected costs. The billing model is not flexible enough to handle cases where you temporarily have thousands of nodes. It is not price effective for monitoring big data jobs. We had to switch to open-source Grafana plus Prometheus for those.

It would be cool to have an open telemetry agent that automatically APM instruments everything in the next release.

I've used the solution for three years.

I'd rate the stability ten out of ten.

I'd rate the scalability ten out of ten.

We did not previously use a different solution.

The setup is very straightforward. Users just install the helm chart, and boom, you're done.

We handled the setup in-house.

Be careful about pricing. Make sure you understand the billing model and that there are multiple billing models available. Set up alarms to alert you of cost overruns before they get too bad.

We've never evaluated other solutions.

It's a great product. However, you have to pay for quality.

We use the product for instrumentation, observability, monitoring, and alerting of our system. 

We have multiple environments and a variety of pieces of infrastructure including servers, databases, load balancers, cache, etc. and we need to be able to monitor all of these pieces, while also retaining visibility into how the various pieces interact with each other. 

Tracing data between components and user interactions that trigger these data flows is particularly important for understanding where problems arise and how to resolve them quickly.

It provides a lot of options for integrations and tooling to observe what is happening within the system, making diagnosis and triage easier/faster. 

Each user can set up their own dashboards and share them with other users on the team. We can instrument monitors based on various patterns that we care about, then notify us when an event triggers an alert with platforms such as Slack or PagerDuty. 

Our ability to rapidly become aware of problems focused on the symptoms being observed and entry points into the tool to rapidly identify where to investigate further is important for our team and our users.

The most valuable aspects of the solution include log search to help triage specific problems that we get notified about (whether by alerts we have configured or users that have contacted us), APM traces (to view how user interactions trace through the various layers of our infrastructure and services to be able to reproduce and identify the source of problems), general performance/system dashboards (to regularly monitor for stability or deviation), and alerting (to be automatically informed when a problem occurs). We also use the incident tools for tracking production incidents.

In some ways, the tool has a pretty steep learning curve. Discovering the various capabilities available, then learning how to utilize them for particular use cases can be challenging. Thankfully, there is a good amount of documentation with some good examples (more are always welcome), and support is very helpful. 

While DataDog has started adding more correlation mapping between services and parts of our system, it is still tricky to understand what is the ultimate root cause when multiple views/components spike. Additionally, there are lots of views and insights that are available but hard to find or discover. Some of the best ways to discover is to just click around a lot and get familiar with views that are useful, but that takes time and isn't ideal when in the middle of fighting a fire.

I've used the solution for about four years.

It seems stable.

It seems to scale well. Performance for aggregating or searching is usually very fast.

Technical support is helpful and pretty responsive.

Positive

We did not use a different solution. 

It's hard to say what ROI would be as I have not managed our system without it to compare to.

I don't manage licensing.

We did not evaluate other options. 

It's a great tool with new features and improvements continuously being added. It is not simple to use or set up, however, if you have the right personnel, you can get a lot of value from what DataDog has to offer.