### Datadog log ingestion versus indexing costs Ingestion is the cheap step: Datadog accepts the log and bills per gigabyte. Indexing is where logs become searchable, and it's billed per event, on top of ingestion. A team that sends everything and indexes everything pays twice for logs nobody searches. Set exclusion filters on indexes, and keep low-value logs in archives instead of live search. ### How Datadog counts hosts and overages Hosts are counted by what Datadog sees reporting, not by what you provision. Short-lived instances, autoscaling bursts, and containers on nodes can each push the count above what you budgeted. Many plans bill on a high-water-mark style measure rather than an average, so one spike can set the month's number. ### Cap Datadog costs with usage alerts You can alert on estimated usage the way you alert on CPU, because Datadog exposes its own usage data. Create monitors on custom metric counts, indexed log volume, and host counts, each with a threshold set below your budgeted level. When [Activepieces](https://www.activepieces.com) receives the alert and posts to a chat channel or opens a ticket assigned to the owning team, a month-end surprise becomes a same-day fix. The run is the meter there: Activepieces' pricing page lists one credit per flow run whether the flow has two steps or ten, while Zapier's pricing page bills per task and Make's per module. That matters on a bill like this one, because a flow that checks the alert, looks up the owning team, formats the message and files the ticket as separate steps costs the same as one crammed into a single step. The free plan covers 1,000 credits a month and unlimited flows, enough for a usage-alert flow. ## Datadog versus Grafana Cloud, New Relic, Dynatrace pricing On a like-for-like estimate, Datadog usually costs more, because it meters each product separately, while the other three charge mainly on data volume, host size or seats. The table shows each vendor's metered unit, which tells you what to count before you ask for a quote. | | Pricing model | Free tier | Rough 50-host cost | Lock-in versus breadth | |---|---|---|---|---| | Datadog | Per host, per product, plus custom metrics and log volume | Exists but narrow | No single figure; sum host charges for each product you enable | Broad; proprietary agent and query language | | Grafana Cloud | Usage-based on metric series, log and trace volume, plus users on paid tiers | Permanent and the most usable of the four for small teams | Driven by series and gigabytes, not host count | Narrower out of the box; built on open-source parts and OpenTelemetry, so easier to leave | | New Relic | Per gigabyte ingested, plus per full-platform user | Permanent, with a monthly data allowance | Driven by data and seats, not host count | Broad single platform; data is portable but dashboards and queries are not | | Dynatrace | Consumption-based; hosts are charged by memory size and hours | Trial rather than a standing free plan | Calculable from the rate card once you know host memory and features | Broad and highly automated; the most tightly bound to its own agent |  Read the cost column as a warning: only Datadog and Dynatrace let you start from a host count, and even then you must multiply by products. ### Where Datadog costs more or less Enable many products at once and Datadog is dearer, because every host is charged again for each one. For small, metrics-heavy setups, [Grafana](https://grafana.com/docs/grafana-cloud/platform/pricing-and-usage/logs/) Cloud is cheaper, since you pay for series and gigabytes rather than hosts. It turns dearer if your instrumentation produces high-cardinality metrics, because each unique label combination counts as another series. Few people needing full access and a steady data volume make New Relic cheapest. Add many engineers who each need a full seat and the user charge starts to dominate. For large hosts you want fully automated, Dynatrace is the fit. Because memory size sets the charge, a few big machines cost more than the same number of small ones. ### What published prices cannot tell you Nothing on a rate card shows how much data your services emit, how many custom metrics your teams will create, or how many people need full seats. It also hides the cost of leaving.  Moving off a proprietary agent means rewriting dashboards, alerts and queries, and that labour belongs in the estimate as a cost. ## What Activepieces does about this The problem in this article is that Datadog's bill is set by units you don't watch: custom metric series, indexed log events, and hosts counted by Datadog's rules. Activepieces doesn't change Datadog's pricing. It gives you a way to act on the usage data Datadog already exposes, before the month closes. Build a flow that triggers on a Datadog monitor alert, looks up which team owns the service, and posts to that team's channel or opens a ticket. Run it for each of the thresholds above: custom metric count, indexed log volume, and host count. The free plan covers 1,000 credits a month and unlimited flows, and a run costs one credit however many steps it has. The core is open source under the MIT licence, so you can self-host it if alert payloads shouldn't leave your network. Start with one flow for the product you expect to grow fastest, which for most teams is logs. ## Frequently asked questions ### Does Datadog have a free plan? Yes, but it's a limited one: a free tier covers a small number of hosts with short data retention and a restricted feature set. It works for trying the interface, not for running production. Because each product is metered separately, the limits are also per product, so switching on logs or APM sits outside what the free tier covers. Check the current limits on Datadog's pricing page before relying on them, since they change. ### Can you negotiate Datadog's list price? Yes, once you are committing to an annual contract. The list price is the starting point for a sales conversation, not a fixed rate. The size of your commitment, the number of products you take together and your willingness to sign multi-year terms all move the discount. Most buyers forget one term: the overage rate. If your usage runs above the commitment, you pay that rate on every unit above it, so a good discount on the commitment can still leave you exposed. ### Is Datadog billed monthly or annually? Both are offered, and they're priced differently. Most products list a lower per-unit rate for annual billing than for month-to-month, so flexibility costs more per host or per GB. Annual commitments also bring the overage question: usage above what you committed to is typically billed on top. If you are unsure of your volume, a smaller commitment with a clear overage rate can be safer than a large one you won't use. ### How do I estimate my Datadog bill before buying? Estimate by product, then add the two usage-driven items that grow fastest: logs and custom metrics. This is the order that keeps surprises out of the quote: 1. List your hosts by product, since infrastructure monitoring, APM and any other product you switch on are each counted and priced on their own. 2. Estimate log volume in GB per day, split between what you ingest and what you keep searchable. 3. Count your custom metrics, including the tag combinations that multiply each one into many billable series. 4. Multiply each count by the current list price for that product and billing term, then add the lines together. 5. Request a committed-use quote based on those counts, and compare it with your list-price total. 6. Negotiate the overage terms, asking for a cap or a fixed rate on usage above your commitment. Treat the total from step four as your worst-case baseline. The quote in step five shows how much a commitment actually saves against it. Run your trial against a representative slice of traffic first so the counts in steps one to three come from measurement, not guesses. ## Related reading - [What Is Harness Engineering? Building Reliable AI Agents](https://www.activepieces.com/blog/what-is-harness-engineering-building-reliable-ai-agents) - [Support Handoff Automation for Engineering Orgs: 2026 Guide](https://www.activepieces.com/blog/support-handoff-automation-for-engineering-orgs-2026-guide) - [GitLab Pricing in 2026: What Teams Pay](https://www.activepieces.com/blog/gitlab-pricing-in-2026-what-teams-pay) ## References - [Datadog](https://www.datadoghq.com/pricing/list/)Add a single tag such as user ID or container ID and one metric can multiply into thousands of series, so the charge follows your tagging habits, not your metric count.