Datadog is a monitoring platform known for its ease of use due to built-in dashboards and a large number of integrations with included functionalities. However, a common Datadog complaint is that the pricing structure is difficult to understand and expensive. Grafana is primarily a visualization and observability platform but when paired with Grafana’s other open source tools (Mimir, Loki, and Tempo), it forms a complete observability stack that handles not only metrics but also logs and traces. Grafana also offers Grafana Cloud, a managed service that includes the observability stack and additional features, making it a strong competitor to the monitoring and observability features of Datadog for a fraction of the price.
Datadog vs Grafana Use Cases
Datadog is an all-in-one platform with a range of integrated services, such as metric visualization, logs, and Application Performance Monitoring (APM). Its seamlessness across services and number of integrations, each with built-in integration-specific functionality makes it a great option for organizations looking for a unified, out-of-the-box observability solution.
Grafana is highly customizable, and while it specializes in metric visualization, it achieves functionality similar to that of Datadog by integrating it with its full observability stack. One of the most popular Grafana observability stacks is made up of other Grafana tools to form the LGTM stack. It includes:
There are other options for tools that can connect with Grafana, such as Elasticsearch for logs or Jaeger for tracing. However, since Grafana Cloud (see next section) fully manages the LGTM stack in its deployment option, we will focus on that option.
Open Source Grafana vs Grafana Cloud vs Grafana Enterprise
Before we dive into the comparison between Datadog and Grafana Cloud, it’s worth clarifying the difference between Grafana’s deployment options. Unlike Datadog, which is exclusively cloud-based and fully managed, Grafana provides flexibility with three deployment options: Open Source Grafana, Grafana Cloud, and Grafana Enterprise.
Open Source Grafana is free, self-managed, and self-hosted, making it ideal for developers or teams looking for a cost-effective solution. However, the draw comes from maintaining the infrastructure and observability stack (mentioned in the previous section), resulting in operational overhead and engineering costs. Additionally, it lacks advanced features like pre-built dashboards and reporting.
Grafana Cloud is a fully managed, cloud-hosted service that handles the observability stack for logs, traces, and metrics and eliminates the need for maintaining or scaling infrastructure. Grafana Cloud has three tiers: Free, Pro, and Advanced, each with its own set of features and security offerings.
Finally, Grafana Enterprise builds on the open source version, offering additional capabilities such as access control. It is self-managed, but there are enterprise offerings for each piece of the LGTM stack with features (SSO/SAML) and enterprise support packages. There is also the added flexibility of choice between self-hosted and cloud for deployment.
For the purposes of this blog, we will focus on Datadog vs Grafana Cloud since Grafana Cloud’s integration of the observability stack and its fully managed hosting model make it the most comparable option to Datadog.
Data: Datadog vs Grafana
Data Formatting and Data Sources
Grafana and Datadog both support data collection from a wide range of data formats, for example, JSON, Prometheus, and CloudWatch logs, but they handle and process data differently. Datadog reformats and stores incoming data in proprietary structures to fully integrate and manage it within the Datadog platform.
Open Source Grafana does not store data, rather it’s a visualization layer that pulls data from 100+ data sources, such as Loki, Tempo, and CloudWatch, leaving the data format unaltered. This reduces vendor lock-in, as no data transformation is required. Grafana Cloud also allows you to ingest data into its hosted backend.
A workaround to remain vendor-neutral with Datadog is by importing data using open standards, such as OpenTelemetry, to standardize data in a consistent format.
Data Collection
Datadog has several options for centralized methods to collect data, such as the Datadog Agent, integrations, and API. It has over 800 pre-built integrations for various services like cloud services, databases, Kubernetes, and more, for extremely simple data collection.
Grafana supports data collection through Grafana Alloy, APIs, and over 100+ integrations for monitoring infrastructure, services, and more.
Datadog vs Grafana Service Cost and Comparison
Pricing is per service for Datadog and Grafana Cloud, so users can pick and choose which areas of the platforms they want to use. For Datadog, there are 30 billable services; for Grafana Cloud, there are 11. We’ll focus on only the relevant/comparable ones. Pricing varies depending on the tier and Datadog country. We’ll focus on the Grafana Cloud Pro tier (which includes the managed LGTM observability stack) and the Datadog US region. All costs are per month.
Grafana Cloud Pro starts at $19 a month and includes all Grafana Cloud features, 10k metrics, 50 GB logs, 50 GB traces, and more. Datadog pricing for most services is per host, where a host is any OS instance, including physical servers, VMs, containers, and databases. As we’ll get into, per-host pricing can add up quickly since every small OS, such as a single EC2 instance, counts as a host. Datadog charges for hosts based on 99th percentile billing which means you are billed based on your highest sustained usage over time, excluding brief traffic spikes, potentially leading to higher costs if you scale up frequently or have fluctuating workloads.
It’s also worth noting that for larger workloads both Grafana and Datadog offer custom rates.
Metrics and Visualizations
With Datadog, once a host is connected, Datadog automatically generates relevant graphs and dashboards tailored specifically to the host or service. Datadog offers customization, though not to the same level of granularity as Grafana, but the UI is simple and intuitive for technical and non-technical users.
Metrics and visualizations are listed under Datadog’s Infrastructure pricing, which has multiple tiers (this is where the free tier lives). We’ll compare the Datadog Pro tier with Grafana Cloud’s Pro tier. The pricing is a flat $15 per host if you pay upfront or $18 if you pay on-demand, plus no additional charges for collecting data at a lower interval. Relevant metrics for the host are automatically collected and are included in the per-host cost. That cost also includes 100 custom metrics (note—each unique combination of metric name, host, and tag set is counted as a separate custom metric). If you go over 100 custom metrics, the cost is somewhere between $1-5 per 100 custom metrics but you have to talk to a salesperson to see the actual price.
Grafana is often regarded as the gold standard for visualization. The UI is sleek and its dashboards are highly customizable, with tons of visualization options and fine-grained control over layouts, queries, and display styles. This level of customization makes Grafana ideal for advanced users who want tailored solutions, though its interactive query builder can be challenging for less technical users. Also, while Grafana Cloud does include pre-built visualizations, the selection is smaller compared to Datadog’s offerings.
The pricing is a bit tricky to understand at first, as well. Grafana Cloud charges $8 per 1,000 “metrics billable series,” which is calculated based on active series and data points per minute:
- Active series: A series is active when new data points have been received within the last 20 minutes.
- Data points per minute (DPM): Defined as a single measured occurrence of a metric within a time series.
Grafana provides this pricing example:
50,000 active series x (1 DPM / 1 DPM Included) x ($8 / 1000 active series) = $400 / month
For use cases where data needs to be sent more frequently, there is a high resolution plan priced at $16 per 1,000 metrics billable series that includes 4 DPM (every 15 seconds).
50,000 active series x (4 DPM / 4 DPM Included) x ($16 / 1000 active series) = $400 / month
Grafana Cloud also has a feature called Adaptive Metrics that analyzes usage patterns and identifies data that can be aggregated to lower cardinality versions to reduce metric costs.
For visualizations, Grafana Cloud charges $8 per active user. Note–for Enterprise plugins, the cost per active user is bumped up to $55.
Pricing Scenario #1: Metrics and Visualizations for EC2
A startup uses EC2 instances to host its applications and services. It has 20 instances of various sizes for production and development environments. They are looking for a monitoring and visualization solution for their 8 team members to track metrics like CPU utilization, memory usage, and network traffic. On average, they track only 7 unique metrics per instance, and metrics are collected once per minute.
Grafana Cloud Cost
Assuming 1 instance has 1000 active series 1000 active series x 20 instances = 20,000 active series 20,000 active series x (1 DPM / 1 DPM Included) x ($8 / 1000 active series) = $160 for metrics (assume the included 10k metrics are applied elsewhere) $8 per active user x 5 users (Grafana Cloud Pro includes 3 users at no cost) = $40 for users $160 for metrics + $40 for users = $200 total Grafana Cloud
Datadog Cost
The same scenario in Datadog costs more due to each EC2 instance counting as a host. 20 hosts x $18 per on-demand host = $360 total Datadog
TLDR; Grafana Cloud will be cheaper for environments tracking fewer metrics for many hosts since Datadog charges per host, regardless of the number of metrics tracked.
Pricing Scenario #2: Metrics and Visualizations for Custom Metrics
A team of 10 has a project they want to monitor with 5 custom data sources/hosts. On average, there are 15,000 custom metrics/active series per host. They want their metrics loaded every 5 seconds.
Grafana Cloud Cost
15,000 active series x 5 data sources = 75,000 active series 75,000 active series x (12 DPM / 4 DPM Included) x ($16 / 1,000 active series) = $3,600 for metrics (Assume the included 10k metrics are applied elsewhere) $8 per active user x 7 users (Grafana Cloud Pro includes 3 users at no cost) = $56 for users $3,600 for metrics + $56 for users = $3,656 total Grafana Cloud
Datadog Cost
5 hosts x $18 per on-demand host = $90 host costs ((15,000 custom metrics / 100 custom metrics) x $5) x 5 hosts = $3,750 for metrics $90 host costs + $3,750 custom metrics = $3,840 total Datadog
TLDR; For situations with a large number of custom metrics, Datadog is often more expensive compared to Grafana Cloud. For cases involving high DPM rates and/or many Grafana Cloud users, Grafana Cloud may be costlier.
Logs
Datadog’s ability to correlate logs with metrics and traces is one of its strong suits, as well as its advanced features like log enrichment and full-text search. It decouples log ingestion from indexing, so users can store all logs and selectively index those needed for analytics and search (saving some money). The ingestion cost is only $0.10 per GB ingested or scanned. However, those logs are only available for a 15-minute processing window which is only useful for specific use cases such as real-time analysis or correlation with metrics and traces. Also, indexing logs is expensive. It is one of the areas where cost differences between Datadog and Grafana Cloud are highly apparent.
The cost is per log event and depends on the retention period. The indexing cost has retention options (when logs are searchable and analyzable) of 3, 7, 15, 30, and 30+ days. The 30-day price is $2.50 per million log events. 7-day is $1.27 per million log events. Note—that’s the cost for annual billing, on-demand is even more expensive.
Grafana Cloud takes a unique approach to logs—rather than indexing the full log line, it indexes the metadata (labels). This results in lower costs, as it requires less storage and processing power compared to traditional log indexing. However, it limits advanced features like full-text search and advanced log enrichment compared to Datadog. Grafana Cloud also has a feature called Adaptive Logs, which similar to Adaptive Metrics, identifies commonly ingested log patterns and creates a set of recommendations for what logs to keep and what to drop based on how frequently those patterns are queried.
Grafana Cloud only has a 30-day retention period or 30+ days (in 30-day increments). The 30-day retention cost is $0.50 per GB of logs ingested. There may be additional charges if you query more than 100X of your ingested logs volume.
Pricing Scenario #3: Logs
25,000 GB of log data is ingested and stored for 1 month.
Grafana Cloud Cost
25,000 GB x $0.50 per GB = $12,500 total Grafana Cloud
Datadog Cost
25,000 GB x $0.10 per GB = $2,500 for ingestion 25,000 GB / 1 KB (assuming average log event size is 1 KB) = 25 billion log events 25 billion log events x ($2.50 / 1 million log events) = $62,500 for indexing $2,500 for ingestion + $62,500 for indexing = $65,000 total Datadog
TLDR; Datadog is expensive for logs indexed. Even if the Datadog retention period was 3 days the Datadog cost would still be over double the Grafana Cloud cost.
APM & Traces
Datadog offers a comprehensive APM solution with support for distributed tracing, error tracking, and real-time performance analysis. Grafana is less focused on APM; however, it’s making strides to extend its APM offering. Grafana Cloud has an APM solution called Application Observability that gives the ability to correlate metrics, logs, and traces, and detect anomalies.
Application Observability in Grafana Cloud is $0.04 per host hour, so running it all month would be $28.8 per host. Datadog is $31 per month and includes tracing.
Grafana Tracing is billed outside of it’s Application Observability so you can still use tracing without the associated APM costs. It’s charged at the same rate as logs, $0.50 per GB of logs ingested at a 30-day retention period or 30+ days (in 30-day increments).
Free Tier
Datadog’s free tier includes “core collection and visualization features,” 5 hosts, and only 1-day metric retention. Grafana Cloud’s free tier is much more generous, with all Grafana Cloud features and 14-day metric retention. Also, 50 GB of Logs, 50 GB of Traces, 2,232 host hours for Application Observability, and more.
Conclusion
Grafana Cloud and Datadog both offer full monitoring platforms, with Grafana’s main strength being in its customizability and Datadog’s in its unified approach and extensive integrations. However, Grafana Cloud is often significantly more affordable, especially in scenarios with many hosts, making it the cost-effective choice for many organizations.
Monitor your Datadog costs.
