One line on what each tool does, one line on why it matters for a PM, and the link. Give us your email and this morning’s issue — all five tools — lands in your inbox. No card, no password, nothing to install.
$10/month or $96/year after 7 free days, cancel in one click
Twenty-five engineers is where a Datadog bill stops being a rounding error and starts being a line somebody defends. The meter is still hosts, not people: at Datadog's published $15 per host per month on an annual commitment, our 40–100 host band costs $7,200 to $18,000 a year for Infrastructure Pro alone — and switching APM on across the same band adds $14,880 to $37,200 on top.
~14 min read · prices re-checked 2026-09-06· no sponsorships, no affiliate links
Datadog meters INSTRUMENTED HOSTS, not engineers — it counts your host total every hour and bills the high-water mark of the lower 99% of those hourly readings for the month.
OUR ASSUMPTION, stated so you can reject it: a twenty-five-person B2B SaaS engineering org runs about 40 hosts if it is lean and consolidated, 65 in the typical case, and 100 if it runs several environments, a data plane and a build fleet. Datadog's price list contains no headcount at all; this mapping is ours.
Worth it at twenty-five engineers if you commit annually on Infrastructure Pro, scope APM to the services that page a human, and set a log-indexing budget before the first invoice rather than after it.
at stake in the APM scope decision alone at the top of our host band — 100 hosts traced that never page anybody, over twelve months.
Infrastructure Pro at the published $15 per host per month × 65 hosts × 12.
$31 per APM host per month — the single biggest lever on this page.
Infrastructure Enterprise at $23 per host per month across 100 hosts — $9,600 above Pro.
Saves $3,600 a year at 100 hosts by committing rather than running on-demand, $9,600 by taking Pro instead of Enterprise, and $37,200 by keeping APM off 100 hosts that do not need it — on Datadog's published list prices.
Datadog meters hosts, not engineers. At the published $31 per APM host per month, 100 hosts is $37,200 a year; taking Infrastructure Pro instead of Enterprise across the same 100 hosts is another $9,600, and committing annually rather than running at the $18 on-demand rate another $3,600. Rates read from Datadog's own list pricing page on 6 September 2026; the 40–100 host band is OUR assumption about a twenty-five-person org and the multiplication is ours — read your own 99th-percentile host count in Usage Details and substitute it.
What would flip this. This becomes a straight Buy once you have a rotation paging more than one team and an estate that no single person can hold in their head, and it becomes Do not buy if a competing platform is already bundled into a contract you have signed and cannot exit for a year.
Priced against this org: A twenty-five-person engineering org at a Series B B2B SaaS company: fifteen to thirty services on managed Kubernetes, a data pipeline, a permanent staging environment, GitHub, an incident tool, and a two-team on-call rotation with a written escalation policy.
Every figure above re-read by hand from Datadog — list pricing on 6 September 2026. Our Datadog verdict for a 100–1,000 person engineering team quotes the same published rates and the same $37,200 a year for every 100 hosts left off APM, plus $48,000 in the Pro-versus-Enterprise decision at 500 hosts. Identical rate card, larger estate.
Turning a headcount into a host count is an assumption, not a calculation, and at twenty-five engineers the spread between two real orgs is wider than the spread between two Datadog plans. That is exactly why this page shows a band and refuses to publish a single confident annual figure for "a 25-person team".
| Band | Assumed instrumented hosts | The arithmetic | Why an org this size lands here |
|---|---|---|---|
| Lean — 40 hosts | 40 instrumented hosts | $15 × 40 × 12 = $7,200 / yr on Infrastructure Pro | Production and one shared pre-production environment, aggressive bin-packing, and managed services left uninstrumented. Achievable at this size, and rarer than it should be. |
| Typical — 65 hosts | 65 instrumented hosts | $15 × 65 × 12 = $11,700 / yr on Infrastructure Pro | Two or three environments, a data pipeline with its own nodes, and CI runners that stay warm. Plan the budget on this band unless your own usage page says otherwise. |
| Heavy — 100 hosts | 100 instrumented hosts | $15 × 100 × 12 = $18,000 / yr on Infrastructure Pro | Multi-region, per-team preview environments, or an autoscaling floor set during a launch and never lowered. The high-water-mark rule means the floor is what you pay for, not the average. |
Use your own reading, not ours. Read your real number: Datadog's Usage Details page reports the 99th percentile of distinct infrastructure hosts across every hour of the current month, plus the container high-water mark. That figure is the bill. Put it into the arithmetic above and the page is priced for your org rather than our composite.
Datadog's published list rates across our 40–100 host band at twenty-five engineers (read 6 September 2026).
| Plan | Published price | × 25 seats / yr | What it buys, or the trap |
|---|---|---|---|
| Infrastructure Pro, billed annuallyOur pick | $15 / host / mo | $7,200 – $18,000 / yr | What we would buy: metrics, dashboards, alerting and 15-month metric retention across the band, on the committed rate rather than the on-demand one. |
| Infrastructure Pro, on-demand (no commitment) | $18 / host / mo | $8,640 – $21,600 / yr | Twenty per cent more for the same product — $3,600 a year at 100 hosts. Defensible for one quarter of discovery, indefensible as a standing arrangement at this size. |
| Infrastructure Enterprise, billed annually | $23 / host / mo | $11,040 – $27,600 / yr | $8 per host per month above Pro — $96 per host per year, or $9,600 across 100 hosts. Buy it only against a named capability, not because it is the middle column. |
| APM, billed annually, on top of infrastructure | $31 / APM host / mo | $14,880 – $37,200 / yr | Additive, and at this band larger than the infrastructure line itself. $37,200 a year is what APM on 100 hosts costs at list — the single biggest lever on this page. |
| APM Pro and APM Enterprise | $35 and $40 / APM host / mo | $16,800 – $48,000 / yr | APM Pro adds Data Streams Monitoring; APM Enterprise adds the Continuous Profiler. Both are per APM host, so scope discipline matters more than tier choice. |
| Logs: ingest, then index | $0.10 / GB ingested | $3,600 / yr at 100 GB/day, plus indexing | Indexing is the half that scales: $1.70 per million events per month at 15-day retention on an annual commit, $2.55 on-demand, $2.50 at 30-day retention. |
Every per-unit price is a rate Datadog publishes on its own list pricing page, read 6 September 2026; the annual column is that published rate multiplied by the stated seat count's assumed host band and twelve months. The 40–100 host band is our assumption about an org this size, not a Datadog figure. Negotiated rates and committed-use discounts are absent rather than estimated, and no allowance is made for volume Datadog quotes rather than publishes.
Datadog's price list contains hosts, containers, APM hosts, gigabytes and indexed events. It does not contain seats. That is worth stating plainly at this size, because twenty-five engineers is exactly the point where somebody builds a budget spreadsheet with a per-head column and Datadog gets modelled as if it behaved like Jira. It does not: hire five more engineers and the infrastructure bill does not move, while adding one more environment can raise it by twenty per cent.
The mechanic behind the meter is Datadog's high-water-mark plan. Host count is sampled hourly; the invoice is the maximum of the lower 99% of those readings, with the top one per cent of hours discarded so a short spike is forgiven. That rule is unusually reasonable, and it means the number to manage is not your peak, it is your sustained floor — the autoscaling minimum, the always-warm CI fleet, the staging cluster nobody turns off at night.
Get that right and the arithmetic becomes simple: multiply the floor by the published rate, add the meters you deliberately switched on, and compare the result to what an hour of downtime costs a company with twenty-five engineers and paying customers. That is the whole decision.
This is the finding that changes budgets at this size. Infrastructure Pro is $15 per host per month; APM is $31 per APM host per month, published on the same page. Across our typical 65-host band that is $11,700 a year for infrastructure and $24,180 for APM if you instrument everything — and at the top of the band, $18,000 against $37,200. The tool most teams think of as "Datadog plus tracing" is in fact tracing plus a smaller Datadog.
The remedy is not to skip APM. Tracing is the single most useful thing you can point at a twenty-five-person microservice estate at 3am. The remedy is to decide which hosts get it. In practice the services that page a human are a minority of the estate — the API edge, the payment path, the job runner that backs up — and putting APM on those and not on the rest routinely halves the line without costing a single incident.
The tier choice is a smaller lever than the scope choice, and worth stating for completeness: APM is $31, APM Pro is $35 with Data Streams Monitoring, and APM Enterprise is $40 with the Continuous Profiler, all per APM host per month on an annual commitment. Choosing the cheapest tier across the whole estate is worse value than choosing the right tier across a third of it.
The first is the on-demand rate. Datadog publishes $18 per host per month with no commitment against $15 with one — twenty per cent, or $3,600 a year at 100 hosts, for a signature. Teams stay on-demand because nobody wanted to commit before the host count settled, which is reasonable for a quarter and expensive for a year.
The second is the Enterprise column. At $23 against $15 per host per month, Infrastructure Enterprise costs $96 per host per year more, which is $9,600 across 100 hosts. It is frequently bought because it is the tier a quote arrives on, not because anyone identified the capability. Ask for the specific feature in writing; if the answer is a category rather than a feature, you have your answer.
The third is log indexing left unbounded. Ingest at $0.10 per GB is cheap and predictable. Indexing at $1.70 per million events per month at 15-day retention is neither, because event volume grows with traffic and with every new debug statement anybody merges. Set an exclusion filter on the noisy streams, index what you query during incidents, and revisit quarterly. This is the meter that most often turns a defended budget into a surprise.
Buy Infrastructure Pro on an annual commitment for the whole estate, because the correlation story only works if everything is in one place and $15 a host is not the expensive part of this. Put APM on the services that page a human — in a typical twenty-five-person estate that is between five and fifteen hosts, not sixty-five — and review that list every quarter as ownership changes.
Ingest logs broadly and index narrowly, with an explicit monthly budget and an alert on it. Turn container monitoring on only if your pod-to-host ratio is high enough that host-level metrics genuinely hide problems; at $1 per container per month, a fleet of two hundred short-lived containers is $2,400 a year and worth an explicit decision.
Then put fifteen minutes in the calendar each month to read the Usage Details page: the 99th-percentile host count, the container high-water mark and the indexed-event total. Three numbers, once a month. Every over-run we see at this size was visible on that page for weeks before it appeared on an invoice, and no negotiation recovers money the meter has already counted.
Every entry is a real product or a real capability with a link to the vendor’s own page. Where a vendor publishes a price we say so and date it; where they do not, we describe the posture rather than invent a number.
Datadog's own APM billing documentation: APM is billed on the 99th percentile of hosts running an agent that sends traces, hour by hour.
Why it covers the job. Because it is metered per APM host, tracing only the services that wake somebody up is the largest single saving available on this page, and it costs nothing in incident capability.
Cost: Up to $37,200 a year avoided at 100 hosts, on Datadog's published $31 per APM host per month · docs.datadoghq.com
Vendor-neutral tracing and metrics instrumentation with its own published specification and SDKs, which Datadog ingests alongside its own agent.
Why it covers the job. It turns the backend into a procurement decision instead of a rewrite. At twenty-five engineers that optionality is worth more than any single year's discount, because it is what makes a renewal negotiation credible.
Cost: Free and open source; the cost is engineering time, not licence · opentelemetry.io
Managed Prometheus, Loki and Tempo priced on usage, with a published free tier, consuming the same OpenTelemetry data.
Why it covers the job. Batch jobs, internal tools and dev environments rarely need the tool you bought for 3am. Splitting them off keeps the Datadog host count — and therefore the bill — attached to the things that page people.
Cost: Published free tier plus usage-based pricing; no per-host licence · grafana.com
If one of these has happened since 2026-09-06, this page is out of date — tell us and we will change it. Pricing and plan contents rot fastest, so check the vendor’s current page before you budget either way.
Every price on this page is either a figure the vendor publishes on one of these pages or that figure multiplied by an assumed instrumented hosts bandand twelve months. Vendor pricing moves without notice, so each link carries the date we last read it — check the current page before you budget.
Bigger org? Our Datadogverdict for a 100–1,000 person engineering team runs the same published rates at that scale, and the two pages are written to reconcile with each other — if a figure here disagrees with a figure there, one of them is wrong and we want to hear about it.
Written at the Signal Brief desk against the same hard rules the teardown generator follows. The $15 and $23 per infrastructure host per month rates, $18 on-demand, $31 / $35 / $40 per APM host, $1 per container, $0.10 per GB ingested and $1.70 per million indexed events at 15-day retention were all read directly from Datadog's own list pricing page on 6 September 2026; the high-water-mark billing rule and the APM host-counting rule come from Datadog's own billing documentation, read the same day. They reconcile with our 100–1,000 person Datadog verdict, which quotes the same published rates. The 40–100 host band is OUR ASSUMPTION about a twenty-five-person engineering org and is labelled as such wherever it appears; every annual figure is a published rate multiplied by an assumed host count and twelve months. Negotiated rates and committed-use discounts are absent rather than estimated. No sponsorship, no affiliate link, no vendor review of this page.
No sponsorships, no affiliate links and no paid placements anywhere on this site. That is a business-model decision rather than a virtue: a page that names a product and says don’t buy it cannot exist if a vendor is funding it. Corrections to [email protected].
All seat-count verdicts · All evaluations · The do-not-adopt library · A complete sample verdict
You came here to price one tool. Signal Brief is the other half of that job: five new AI tools every weekday morning, one line on what each does, one line on why it matters for a PM, and the link. Start with this morning’s issue, free.
$10/month or $96/year after 7 free days, cancel in one click · a real Adoption Brief, in full · every issue, published free · [email protected]