PEPAI Datadog Cost Check

Fixed price: EUR 450 excl. VAT · Money back if we find less than EUR 450/month · Report within 5 business days · Read-only

What it is

An independent, read-only review of two Datadog cost drivers: custom metrics and logs. We run PEPAI’s open-source Rust CLIs on exports you send, read the results by hand, and write a short report: what drives the volume, what it costs at your prices, and what you could change.

What we look at

• Custom metrics (dd-metric-bill 0.1.2): counts billable metrics (name + tag values, host included) from DogStatsD, JSON-lines or Prometheus; shows which tag adds most series; flags unbounded tags; simulates tag allowlists; prices the surplus above your allotment.
• Logs (log-bill): groups a sample into masked patterns (IPs, e-mails, secrets, numbers); shows bytes and share per pattern; extrapolates to a month; prices ingest per GB; simulates drop/sample rules. With log-bill 0.1.3 or later it also estimates indexed-log cost per million events by retention, checked by hand against your retention and contract price.

Both run offline (no API key, no network) and are open source, so you can rerun every number.

Not covered: Datadog RUM sessions, APM, hosts, synthetics. (rum-meter counts Terraform Resources Under Management and is unrelated.)

What we need from you

• A metrics sample (DogStatsD, JSON lines or Prometheus) and a log sample of a known time window.
• Your plan, paid host count, contract prices, and which logs you index at what retention.

No account access or API keys, ever.

What you get

A written report within 5 business days of complete exports: top metrics, tags and log patterns by cost; concrete options with estimated effect, confidence and caveats; the exact commands used. See the example report (fictional data).

Money-back clause

If the estimated monthly savings in our report total less than EUR 450, you get a full refund of the EUR 450. The savings are estimates from your samples and prices, not a guarantee of what you will actually save; whether you implement them is up to you.

What we don’t do

No changes in your account or code. Figures are estimates from a sample, not invoices. No compliance judgement on which logs you may drop.

Honest limits

• A sample is one snapshot; Datadog bills metrics on the monthly hourly average.
• No Metric Name pricing. The allowlist simulation applies to all metrics at once.
• Indexed logs are per million events and depend on retention (list price, billed annually, read 19 Oct 2026 from datadoghq.com/pricing/list: USD 1.06 / 1.27 / 1.70 / 2.50 per 1M for 3 / 7 / 15 / 30 days). We use your retention and contract price and mark the estimate.
• Results are only as good as your samples. Datadog’s own Usage pages already show top custom metrics and logs by index for live accounts; this check is strongest when you want a no-access, offline estimate, or a pre-deploy / migration cost picture.

Why independent

Open-source tools, no platform or licence to sell, no upsell. If there is little to save, the report says so.

Contact

[email protected]