Go to Logtail and start your log management in 5 minutes.
[/summary] ## Application Performance Monitoring  Datadog’s APM can work in two ways. Either by using a dedicated collector agent like dd-trace for Python applications or via instrumentation with Open Telemetry. Instrumenting Datadog is fairly generic with some issues occurring here and there, which can be resolved by either taking one of their courses, thoroughly reading the documentation, or consulting their GitHub pages. Datadog even provides you with a simple Getting Started wizard for each language and enables you to pre-configure scripts used for invoking the agent and collecting metrics from your services.  AWS uses AWS X-Ray for collecting and analyzing traces from your applications, however X-Ray is directly integrated into CloudWatch UI. CloudWatch offers a really powerful Application for Performance Monitoring. It offers 5 key capabilities: ServiceLens, Resource Health, Synthetic Canaries, RUM, and Evidently. In terms of instrumentation, CloudWatch offers you to test its APM capacities on a sample setup which is really helpful, especially for those, who are new to the AWS ecosystem. Thanks to being a part of the infrastructure CloudWatch offers a powerful APM feature called Evidently, which allows you to deploy and test new features before global launch by serving it just to a portion of your customers and collecting all the important information before full-release. ## Incident Management  Datadog allows you to create monitors and declare incidents from multiple tools within UI. Datadog enables you to declare incidents and set additional information like who’s responsible, attach a link to a video call, set the severity of the incident, or send notifications. You can also draft postmortems using Datadog’s notebooks to store valuable information on the incident. Incident Management is not part of CloudWatch directly but is available from within the ecosystem. CloudWatch offers some fundamental alerting features, but in order to get a full-fledged Incident Detection and Response solution, you need an Enterprise support plan on AWS and an additional subscription fee. [note]Go to Better Stack and start managing your incidents in 2 minutes.
[/note] ## Ease of use and onboarding  Datadog’s UI feels a bit neater and easier to follow than CloudWatch. Tools are available with a click and neatly organized in a left-hand vertical toolbar. It’s fairly easy to use pre-designed dashboards or create your own with custom queries using Datadog’s query language. Datadog’s setup might feel stressful at the beginning, but once you get to know the environment, and understand how the documentation is written and how individual tools operate together, you’re all set. Luckily, all tools offer a Getting Started wizard page containing the steps you need to start and also links to various pages in the documentation. If you feel lost, or want to move your operations to Datadog and feel like you want to do some research and hands-on exercises before you fully commit, you can always the 14-day free trial. If you’re really serious about learning, you can enroll in some courses provided by Datadog’s learning center touching most of the topics, and try setting up Datadog in a virtual environment in a web UI.  To get started with CloudWatch you already need to have some resources up and running on AWS. The UI might feel overwhelming at the start. The AWS Management Console is a starting point for every tool in the ecosystem and it takes some time to figure out. From there you need to get to CloudWatch. CloudWatch UI welcomes you with a Get started with CloudWatch page containing links to individual features of the tool. Subpages like Alarms, Logs, Metrics, or Traces are packed into a side panel. Each feature is available in a toggle heading, enabling you to drill down to a specific tab within the category from anywhere in CloudWatch. The UI feels more practical and technical. Amazon offers a really broad range of onboarding resources, often in a form of “certifications”. You can pick a plethora of various certifications, many of those focused on professional DevOps skills. A lot of resources for these are also free. ## Customer Support  When it comes to specific issues and interaction with the vendor Datadog offers a support tier system on top of their regular pricing. Support is divided into Free, Standard, and Premier. Free support is considerably generous for a free tier. While you are “on your own” you have a plethora of resources available: Online documentation, a Learning Center, Slack Community, and “Foundational Enablement”. If you’re a paying client you’re automatically entitled to the Standard tier. In this tier, Datadog provides you with less than 2 hours of response time for Business-critical issues, email support during business hours, and time-limited chat support. To get to the premium class, you’ll pay an additional 8% on top of your at least $5000 monthly bill and you also need at least 12 month contract with Datadog. For these at least $400 you’ll get 30-minute response times in Business-critical issues, 24/7 Email, Chat, and Phone Support. You also get a designated team of global support engineers, or a priority runbook.  For this comparison, we need to look at AWS support plans. Basic support is available with all AWS Support Plans. In the Free plan, you get 24x7 access to customer service, documentation, whitepapers, and AWS re: post. You also get AWS Trusted Advisor recommendations and Priority. You also get access to the AWS Health Dashboard. The first paid subscription model is the Developer, starting at $29/month, which gives you access to technical support during business hours, AWS Cloud support, associates, and Email support. You can also have one primary contact for case management and a 12 hours response window for cases impairing your systems. To understand the full scope of AWS’ support, let’s skip all the way to Enterprise, starting at $15000. In this subscription model you get a 15 minutes response window in case of critical system outages, AWS support API, access to managed AWS, support for third-party integrations, review of integrations, training online labs, and the already mentioned Incident Detection and Response for an additional fee. ## Datadog vs. AWS CloudWatch, which to choose? First of all, it’s most probably never one or the other, you can often see third-party DevOps tools integrated with CloudWatch as it enables some powerful monitoring and automation features. However, when deciding where to dedicate your resources for monitoring, the following could help you decide. ### Datadog is a perfect go-to solution for those who: - **Need to monitor distributed, multi-cloud solutions with resources hosted outside of AWS.** - Need a powerful and reliable dedicated observability solution enabling you to monitor your resources - Want to mitigate performance and security incidents from one place - Would like to construct your stack from multiple DevOps and observability tools. ### Sticking with CloudWatch is a good solution for those who: - Plan on fully-dedicating their resources and deploying their operation on AWS - Those who can do fine with the features offered by CloudWatch and don’t yet need to opt for a dedicated Observability platform - Solo engineers, freelancers, and smaller start-ups - DevOps engineers who plan to broaden their knowledge in the AWS ecosystem and want to prepare for one of the certification paths. ## Final thoughts Today we’ve taken a look at CloudWatch and Datadog and tried to shed some light on how their different, how they could work together, and possibly help you choose which one would suit you the most. That's it for today, if you got all the way here, thank you for reading my article. If you want to see how Datadog compares with other DevOps tools, check out our comparisons with [New Relic](https://betterstack.com/community/comparisons/datadog-vs-newrelic/), [Dynatrace](https://betterstack.com/community/comparisons/datadog-vs-dynatrace/) or [Sentry](https://betterstack.com/community/comparisons/datadog-vs-sentry/).