4.9万
We wanted something similar to Datadog, so that it's easy for engineers to ramp up to this new system with no steep learning curve. With ClickHouse, we see a compression ratio of 10x and a 200x increase in indexed logs. These logs are easily accessible to all engineers... And all of this came at a 0% increase in cost.
- Comparisons
- Observability
ClickStack vs Datadog
Retain everything. Sample nothing. Investigate without limits.
Modern observability shouldn't make you choose between cost and visibility. ClickStack combines ClickHouse performance with an open, OpenTelemetry-native platform, that lets you retain all telemetry, investigate without limits, and power AI agents.
- Retain everything. Keep months of logs, traces, and metrics without sacrificing performance or ballooning costs.
- Investigate without limits. Run thousands of human and AI-powered investigations with predictable resource-based pricing, not per-host, per-investigation charges or api rate limits.
- Built on open standards. OpenTelemetry-native with SQL, open APIs, and the freedom to use any model, agent, or workflow.
Why teams move from Datadog
As telemetry volumes and AI workloads grow, organizations need a platform that offers more predictable economics, greater openness, and the flexibility to build for the future.
Keep every signal
Retain high-cardinality telemetry for longer without sampling or rollups, preserving the context engineers and agents need to diagnose issues.
Simple resource-based pricing
Pay only for the storage and compute you use. No host licensing, feature-based SKUs, or unpredictable charges as telemetry volumes grow.
Built on open standards
OpenTelemetry-native, with SQL, open APIs, and the freedom to integrate with your existing tools, data platform, and workflows.
Query anything with SQL
Explore telemetry with full SQL and advanced analytics, joining it with business data without proprietary query languages or lock-in.
Built for AI, not just dashboards
Power thousands of investigations with predictable resource-based pricing, and build with any model, agent, or MCP server.
100x cost savings
Reduce observability costs by up to 100x compared with existing deployments, while still retaining all your telemetry for longer.
An open stack for modern observability
ClickStack is an open foundation for observability, giving teams the freedom to build with the tools, models, and workflows that best fit their business.
Open by design, it lets you choose the best technology for every layer of your stack, from telemetry collection and dashboards to AI agents, notebooks, and custom workflows, without vendor lock-in.
Collaborate
AI Notebooks
Query
SQL & Lucene
Visualize
Grafana
Data Lake
S3, Iceberg, Delta
Connect
MCP servers
Integrate
External tools
Alerts
incident.io, Slack, PagerDuty
OpenTelemetry
Logs, Traces, Metrics
AI Agents
Claude, Codex, OpenCode
ClickStack
Economics at scale
ClickStack fundamentally changes the economics of observability. Instead of constantly managing spend through shorter retention, sampling, or product SKUs, simple resource-based pricing lets you retain complete telemetry and investigate without compromise.
Cost per TB logs per month
DatadogClickStack
Volume157 TB / month
Lower costs. Better observability.
Datadog · monthly (est.)
~$604k
~$3.85 / GB
ClickStack · monthly (est.)
—
Annual savings
—
saved annually vs. Datadog
— more cost efficient
at 157 TB / month
Assuming one month retention for ClickStack and Datadog. ClickStack estimate based on ClickHouse Cloud on AWS in us-east-1.
See our full pricing ->->
Retain everything
Keep months or years of logs, traces, and metrics instead of choosing what to discard
Sample nothing
Investigate production using complete telemetry instead of approximations and roll-ups.
Query without API limits
Run thousands of human and AI-powered investigations using predictable resource-based pricing instead of worrying about every query, feature, or SKU
Grow without surprises
As telemetry volumes, engineering teams, and AI workloads increase, your pricing model stays simple and predictable
Built for AI-powered investigations
AI is changing how engineers investigate production systems. Instead of manually navigating dashboards and correlating telemetry, they increasingly use agents to gather evidence, test hypotheses, and identify root causes with complete operational context.
ClickStack provides an open foundation for these workflows, combining complete telemetry, purpose-built tools, and open standards so teams can build and evolve AI investigations without vendor lock-in.
claude-code — ~/incident-1247
Observability data is business data
- Investigate with SQL.
Explore telemetry using familiar SQL and advanced analytical functions, without proprietary query-language constraints. - Correlate operational and business signals.
Join logs, traces, metrics, and business data to quantify incident impact and prove observability's value. - Use a broader data ecosystem.
Analyse observability data with BI tools, Python notebooks, and data science frameworks.
SELECT
failures.ContainerId as ContainerId,
failures.Timestamp AS FailureTime,
recoveries.Timestamp AS RecoveryTime,
dateDiff(second, failures.Timestamp, recoveries.Timestamp) AS RecoverySeconds
FROM
events AS failures
ASOF LEFT JOIN
kube_events AS recoveries
ON failures.ContainerId = recoveries.ContainerId
AND failures.Timestamp < recoveries.Timestamp
WHERE
failures.Message LIKE 'exited abnormally%'
AND recoveries.EventType = 'ContainerReady'
ORDER BY
failures.ContainerId,
failures.Timestamp;Migrations made simple
Moving from Datadog doesn’t have to be all or nothing.
With the Datadog receiver in the OpenTelemetry Collector, telemetry continues flowing to Datadog while the same data is mirrored into ClickStack. This lets teams compare results side by side, validate queries and dashboards, and build AI-powered investigation workflows before confidently transitioning production workloads on their own timeline.
Evaluate side by side
- Validate dashboards
- Migrate workflows
- Run AI investigations
- Compare query performance
- Evaluate cost efficiency
- Migrate when ready
ClickStack compared to Datadog
Modern observability demands more than proprietary workflows. As telemetry volumes grow and AI becomes central to every investigation, organizations need an open platform that retains complete data, scales economically, and can serve ever-increasing query workloads.
Simple resource-based pricing
Native
Months or years at low cost
No sampling required
Full SQL
Bring your own models, agents and MCP servers
Complete logs, metrics, and traces
Open APIs and ecosystem
Petabyte scale with sub-second queries
Multiple pricing dimensions and SKUs
Supported
Cost often drives shorter retention
Often used to control costs
Proprietary query languages
Vendor-centric AI experience
Often optimized for cost
Proprietary platform
Scales, but costs increase unpredictably with volume
Migrate your workload from Datadog today
Stop choosing between cost and visibility. Discover an open observability platform built for AI at scale.
FAQs
We're here to make observability simple, fast, and open. Explore our FAQs to learn more about ClickStack, and if you don't see what you need, we're always happy to chat.
01
ClickStack is an open observability platform built on ClickHouse for storing and analyzing logs, metrics, traces, and events at petabyte scale. Unlike Datadog's proprietary platform, ClickStack is built around open standards including OpenTelemetry, SQL, and open APIs, giving teams the flexibility to integrate with their existing tools, AI models, and workflows.
02
Most teams initially migrate from Datadog as telemetry volumes and observability costs grow. ClickStack's resource-based pricing, efficient storage engine, and open architecture allow organizations to retain more data, eliminate sampling, and run more investigations without the complexity of multiple pricing dimensions or vendor lock-in.
Increasingly, teams are also building open platforms for agentic SRE. As AI agents execute thousands of concurrent investigations, organizations need a platform capable of serving high-concurrency, low-latency query workloads without pricing models that penalize more data, more queries, or more investigations.
03
Savings vary by workload, but many organizations reduce observability costs by 100x to 200x after migrating from Datadog while simultaneously increasing retention and indexing significantly more telemetry.
For example, Clever increased searchable log volume by 200x while keeping costs flat after migrating to ClickHouse Cloud, allowing engineers to search all of their logs instead of just a small indexed subset.
04
Yes. ClickStack includes a Datadog receiver for the OpenTelemetry Collector, allowing you to mirror telemetry into ClickStack while continuing to use Datadog. Teams can validate dashboards, compare investigations, and build AI workflows before transitioning production workloads on their own timeline.
05
Yes. ClickStack is OpenTelemetry-native, supporting logs, metrics, and traces using standard OpenTelemetry collectors, protocols, and semantic conventions. Teams can adopt ClickStack without replacing their existing instrumentation.
06
In most cases, yes. Existing OpenTelemetry instrumentation continues to work without modification. Teams using Datadog agents can migrate incrementally using the Datadog receiver before transitioning fully to standard OpenTelemetry pipelines if they choose.
07
Yes. ClickStack is designed as an open foundation for agentic observability. Connect any model or agent framework through MCP, SQL, and open APIs, allowing teams to build AI workflows around their own systems, operational knowledge, and engineering practices instead of relying on a proprietary AI assistant.
08
ClickStack combines complete telemetry, predictable resource-based pricing, and semantically rich observability tools to give AI agents the context they need to investigate production systems accurately. Unlike platforms where cost often drives sampling or shorter retention, ClickStack makes it practical for agents to reason over complete operational history.
AI Notebooks provide a collaborative workspace where engineers and AI agents work together on investigations. Teams can share investigations created from their preferred AI tools, inspect every query and reasoning step, build on previous investigations, and create a persistent body of operational knowledge that improves over time. We believe the future of observability is collaborative, with humans and AI agents investigating systems together.
09
Yes. ClickStack stores logs, metrics, traces, and events within ClickHouse, allowing teams to correlate signals across telemetry types using SQL and OpenTelemetry. This enables richer investigations and supports both traditional dashboards and AI-powered workflows.
10
Yes. ClickStack is built on open-source technologies including ClickHouse, HyperDX, OpenTelemetry, and Vector. Teams can deploy and customize the stack themselves or use Managed ClickStack.
11
Yes. Managed ClickStack is a fully managed observability platform built on ClickHouse Cloud. It eliminates operational overhead with automated scaling, upgrades, backups, security, and managed OpenTelemetry ingestion, allowing teams to focus on investigations instead of infrastructure.
12
Yes. ClickStack can be deployed on your own infrastructure, in any major cloud, or consumed as Managed ClickStack. This gives organizations complete flexibility over where telemetry is stored and how observability workloads are operated.
