Datadog alternatives are researched after an invoice, not after an outage. The buyers are engineering teams whose bill jumped following a noisy deployment or a new tag, and who want a meter they can forecast. They move towards OpenTelemetry-native services with one or two published meters, or to open source stacks they can host themselves.
Pricing: Free tier to 5 hosts; per host per month plus per GB of logs and per custom metric, published
Established in: United States
Built for: Organisations that want every observability product from one vendor
Not for: Teams that need the bill to be predictable month to month
01
Why buyers look past Datadog
01
Many meters on one invoice
Hosts, ingested logs, indexed logs, custom metrics, spans, RUM sessions and synthetics are each billed separately. Any one of them can be estimated; the combination is hard to forecast, and our data lists teams that need a predictable monthly bill as the buyer Datadog does not suit.
02
A single tag can multiply the bill
Attach a user id or a full URL as a label and one metric becomes a great many time series, which Datadog bills as custom metrics. Nothing in the console stops an engineer from doing it, and overage after a noisy deployment is the standard complaint in this market.
03
The agent is where the product invests
Datadog accepts OTLP, but our data notes the native agent works better and the strongest features assume it. Teams that want their instrumentation to stay portable see that as a cost deferred to the day they leave, and prefer a service built on OpenTelemetry.
Usage-based per million spans, log records and metric data points, published; 14-day trial
#4
you want hosted OpenTelemetry with the price visible before signing
03
What each one does differently
1
Grafana Cloud
United States · Free tier; Pro has a monthly platform fee plus usage on metrics, logs and traces and per active user, published
Grafana Cloud hosts the open source Loki, Mimir, Tempo and Grafana, which you may also run yourself, and takes OTLP directly with no proprietary agent. It bills on active series and ingested gigabytes, so high-cardinality labels still cost money, and dashboards and alerts are yours to build.
Switch from Datadog if you want broad coverage with an exit that does not mean instrumenting again.
United States · Free tier of 100 GB a month; then per GB ingested plus per core and full-platform user, published
New Relic is the nearest like-for-like platform but bills on two meters, gigabytes ingested and users, where Datadog has six or more. OTLP is a first-class input and there is a usable free allowance. Full-platform seats cost many times a basic one.
Switch from Datadog if you run many small containers and only a few engineers need full access.
Austria · Consumption-based per host hour or memory-GiB hour and per GiB ingested, published rates
Dynatrace matches Datadog's reach for large estates and adds automatic dependency mapping from one agent per host. It does not fix the forecasting problem: pricing is consumption-based across host hours, ingest and retention, and its best features also assume its own agent.
Switch from Datadog if the estate is too large to map by hand and discovery is the priority.
India · Open source self-hosted; cloud from USD 49 a month, priced per GB ingested, published
SigNoz takes OpenTelemetry as its only input and publishes a single price per gigabyte ingested for its cloud, with the same product open source to self-host. Query depth, alerting and the integrations catalogue are shallower than an established vendor like Datadog.
Switch from Datadog if you already instrument with OpenTelemetry and want one legible meter.
United States · Free tier; per event ingested per month, published
Honeycomb charges per event and welcomes the high-cardinality attributes that Datadog bills as custom metrics. It is a querying tool more than a dashboard wall, so a team that relies on prebuilt dashboards and default alerts will find it unfamiliar.
Switch from Datadog if custom metrics drive your bill and your engineers debug by asking new questions.
United States · Per host per month, published; self-hosted edition available
IBM Instana prices per host only and keeps every trace without sampling, so traffic growth does not move the bill the way Datadog's ingest and span meters do. A self-hosted edition exists. The console is dated and buying runs through an IBM process.
Switch from Datadog if you want a bill that stays flat as traffic rises.
Netherlands · Consumption-based on ingest and retention, published rates
Elastic keeps logs, traces and metrics in one searchable index and can be self-hosted, with cost following ingest and retention where Datadog also counts hosts and custom metrics. In exchange you take on shard counts, index lifecycle and retention tiers, which is specialist work.
Switch from Datadog if you already run Elasticsearch and want a single store for all three signals.
Germany · Usage-based per million spans, log records and metric data points, published; 14-day trial
Dash0 accepts OTLP and nothing else, and publishes its price per span and per gigabyte, so the bill can be estimated from your own trace volume. It is the newest product in the category, with a thin integration catalogue against Datadog's wide one, and no self-hosted edition.
Switch from Datadog if you want hosted OpenTelemetry with the price visible before signing.
Datadog is worth keeping when coverage is the requirement: infrastructure, APM, logs, real user monitoring, synthetics and security in one console, with an integration for whatever you already run, so a new service is visible the day it ships. An organisation that wants every observability product from one vendor, and has someone watching tags and retention, will struggle to match that with one alternative.
05
What moving off Datadog involves
01
Instrumentation is the expensive part
Services instrumented with the Datadog Agent have to be instrumented again for most destinations. Doing it once in OpenTelemetry, sent through your own collector, means the next move is a configuration change. Start with one real service and count the afternoons it takes.
02
Measure your volume before asking for quotes
Every alternative bills on a different unit: gigabytes, spans, events, hosts or active series. Take a week of actual telemetry from the current setup, expressed in each of those, to the vendors on your list. A quote built on assumed volume describes a different company.
03
Dashboards and alerts are rebuilt by hand
Monitors and dashboards do not transfer between vendors, and the OpenTelemetry-native tools arrive with fewer prebuilt ones. Use the rebuild to delete alerts nobody acts on and to remove the labels that inflated custom metrics, otherwise the cardinality problem follows you to the new meter.
06
Questions about replacing Datadog
4 answers
What is the cheapest alternative to Datadog?
It depends on the shape of your telemetry. SigNoz and the Grafana stack can be self-hosted as open source, with operations as the cost.
Among hosted services, Grafana Cloud and Honeycomb have free tiers and New Relic a usable free allowance. Price each on a week of real volume.
Is there an open source alternative to Datadog?
Yes. SigNoz is open source and built on OpenTelemetry, and Grafana Cloud is the hosted form of Loki, Mimir, Tempo and Grafana, which you can run yourself. Elastic Observability and Sentry can also be self-hosted. Each makes operations your team's job.
Is New Relic cheaper than Datadog?
New Relic bills per gigabyte ingested plus per user, which favours container fleets that a per-host meter like Datadog's would punish. It gets expensive when many people need full-platform seats, which cost many times a basic seat. Model the seat mix first.
Do I have to re-instrument my applications to leave Datadog?
If you used the Datadog Agent, largely yes. Datadog also accepts OTLP, so services already instrumented with OpenTelemetry can be pointed at SigNoz, Dash0, Honeycomb or Grafana Cloud, which take OTLP natively.
