Committed-use discounts: what to actually do about them

Not every finding in a Datadog cost report is a settings change. A recent beta run against a workforce-scheduling SaaS company's org found no acute spike that month — no cost jumps, no forecasted overage, nothing on fire. What it did find was quieter and, over a year, considerably larger: the account was paying roughly $514.39 a month, combined, for eight product families running entirely on-demand, with zero committed-use discount applied to any of them. This is a guide to what that finding actually means, and — more usefully — where the action to fix it actually happens, because it isn't in a config screen.

What "100% on-demand" means, concretely

Datadog, like most usage-based platforms, offers a cheaper per-unit rate if you commit to a volume in advance. If a product family shows $0.00 of committed coverage against real, sustained monthly usage, every unit of that usage is being billed at the uncommitted rate — the full price, with no volume discount at all. This isn't a mistake or a bug; it's simply the default state until someone actively negotiates a commitment. The check itself is a direct comparison: committed coverage from your contract data against actual usage from usage/hourly_usage, product family by product family. Whatever shows $0 committed against real, non-zero usage is a candidate.

The table, ranked by dollar savings — not percentage

Here's the actual action table from this run. Note what it's sorted by: absolute monthly savings, not percent of anything. A family that's 100% on-demand is 100% on-demand whether it costs $4 a month or $200 — the percentage tells you nothing about which row to act on first. The dollar figure does.

Commit to a 12-month plan for	What you give up	What it saves
infra_container_excl_agent	12-month contract, no early cancellation	$198.34/month
apm_host	12-month contract, no early cancellation	$108.00/month
logs_indexed_30day	12-month contract, no early cancellation	$87.41/month
synthetics_browser_checks	12-month contract, no early cancellation	$52.04/month
synthetics_api_tests	12-month contract, no early cancellation	$37.36/month
ingested_spans	12-month contract, no early cancellation	$13.64/month
siem	12-month contract, no early cancellation	$13.20/month
logs_ingested	12-month contract, no early cancellation	$4.40/month

The top three rows alone — infra_container_excl_agent, apm_host, and logs_indexed_30day — account for $393.75 of the $514.39/month total. That's where the conversation should start if you're only going to negotiate one or two commitments this quarter, not the bottom of the list where the dollar amounts round to pocket change.

Where this action actually happens (it isn't a toggle)

This is the part worth being direct about. A retention filter or a sample-rate setting is something you change yourself, in the Datadog UI, in minutes. A committed-use discount is a commercial contract change — in most cases, this is arranged through your Datadog account team or sales rep, not self-served with a single click in your account settings. If your org doesn't already have committed-use pricing configured for these families, the practical next step isn't hunting for a screen to edit. It's bringing this exact table into a conversation with whoever manages your Datadog contract.

That's also why the report drafted a specific support-ticket question rather than asserting exact numbers it can't know from the outside:

"Can you confirm the specific committed-use discount tiers and upfront costs for committing to infra_container_excl_agent, apm_host, logs_indexed_30day, synthetics_browser_checks, synthetics_api_tests, ingested_spans, siem, and logs_ingested?"

Discount tiers, minimum commitment sizes, and any upfront costs are things Datadog's account team can quote accurately for your specific contract — the report can tell you what you're currently paying and what a commitment would save at current usage, but the actual terms are a conversation, not an API response.

The tradeoff on every single row

Every row in that table trades the same thing for the same thing: a 12-month lock-in with no early cancellation, in exchange for a fixed monthly saving. That's not just a cost question — it's a bet on whether this usage holds steady or grows over the next year. Committing to infra_container_excl_agent at its current usage saves $198.34/month if that usage stays roughly where it is or grows; it's a worse deal than staying on-demand if that number is about to drop because a service gets decommissioned or a workload migrates elsewhere. Worth checking against your own roadmap before committing to any row, not just against the current invoice.

What to actually bring to that conversation

• The per-family monthly on-demand figures from the table above — what you're paying today, uncommitted.
• The multi-month on-demand totals behind them (this run: $977.49 for infra_container_excl_agent, $540.00 for apm_host, $426.70 for logs_indexed_30day, and smaller totals down through the list) — proof this is sustained spend, not a one-month blip.
• The exact support-ticket question above, verbatim if useful — it's already phrased as something a Datadog rep can answer directly.

None of this requires touching a Datadog settings screen. It requires having the right numbers in hand before that call, which is what this part of the report is actually for. See our walkthrough of the full report format for how this fits alongside the findings that are self-service fixes, and our guide to deciding if a 12-month-or-longer commitment is actually worth it before you sign anything.