Observability Ingest Cost Calculator

Workload preset:

Presets are editable starting points, not measured usage.

Pricing profile:

The rate-card note below states exactly which meters the selected profile applies.

{{ pricingProfileLabel }} {{ pricingProfileSummary }}

{{ stream.label }}:

Display unit

Collector reduction:

Apply one measured blended reduction across the three streams.

Indexed or searchable share:

Use 0 when indexing is bundled into ingest or the selected plan has no event-index meter.

Retention target:

days

Billing days:

days/month

Monthly telemetry growth:

Use recent usage growth or a planned traffic change; negative values model contraction.

%/month

Forecast horizon:

Monthly budget cap:

$/month

Contract discount:

Measure	Value	Costing meaning	Copy
{{ row.label }}	{{ row.value }}	{{ row.detail }}

Scenario	Monthly cost	Monthly savings	Operational move	Copy
{{ row.label }}	{{ money(row.monthly_total_usd) }}	{{ money(row.monthly_savings_usd) }}	{{ row.action }}

A small change in telemetry volume can become a large monthly bill because observability platforms may charge at several points between collection and long-term storage. Logs, traces, metrics, and profiles also grow at different rates, so a single average from last month can hide the stream that is driving the next invoice.

Useful cost planning starts with measured daily volume before paid ingest. Collector filtering, trace sampling, aggregation, and attribute removal can reduce that volume, but the reduction must reflect a real pipeline rather than a hoped-for saving. The remaining data is multiplied by billing days and then evaluated against the selected plan's included allowance and usage rates.

Processed volume: The telemetry received and prepared by a platform before any separately metered written or retained volume.
Billable ingest: The monthly volume left after collector reduction and any included ingest allowance.
Indexed share: The portion of estimated events sent to a separately priced searchable tier. It has no cost effect when the rate is zero.
Hot retention: The searchable period billed by the observability plan. Archive storage and later rehydration may follow different prices and are outside this model.

Cost is not a complete measure of observability value. Dropping noisy records may reduce spend without harming investigations, while aggressive filtering can remove the only evidence of a rare failure. Savings scenarios should therefore be checked against incident coverage, query needs, compliance retention, and the actual rate card before a pipeline change is approved.

How to Use This Tool:

Begin with a representative usage period and the contract meter that most closely matches the account.

Choose a Workload preset or enter measured daily logs, traces, and metrics or profiles in GB or TB. Treat a preset as an editable starting point.
Select a Pricing profile. Use Custom contract when the invoice includes processing, ingest, event-index, retention, fixed, or user charges that differ from the provided public-rate examples.
Enter the measured Collector reduction, searchable share, retention target, and billing days. Set the indexing rate or indexed share to zero when indexing is bundled into ingest.
Add the monthly growth rate and forecast horizon. Use a negative rate only when a sustained contraction is expected, not for a one-time traffic dip.
Review Cost ledger first, then compare Savings plan rows and the forecast. If a budget cap is entered, check the variance against the monthly total.

Interpreting Results:

The monthly total is a planning estimate for the entered meters, not a reproduction of a vendor invoice. Check Billable ingest to see whether an included allowance is still absorbing volume, and use Effective cost per GB only when comparing scenarios built from the same pricing basis.

A collector scenario can lower processing, ingest, indexing, and retention together because less telemetry reaches every downstream meter.
An indexing scenario saves money only when a nonzero per-million-event rate applies.
A shorter retention target saves money only above the included-retention period and when the extended-retention rate is nonzero.
The forecast compounds one growth rate across all three streams. Recalculate separately when logs, traces, and metrics have materially different growth patterns.

Technical Details:

The model converts each daily stream to decimal gigabytes, where 1 TB equals 1,000 GB. Collector reduction is applied before the monthly allowance and all usage meters. Event counts use 1 GB as 1,048,576 KB because average event size is entered in kilobytes.

Formula Core:

Monthly cost is assembled from after-collector volume, billable ingest, estimated indexed events, extended retention, fixed charges, user charges, and a final contract discount.

\begin{array}{lcl} G_{raw} & = & G_{logs} + G_{traces} + G_{metrics} \\ G_{month} & = & G_{raw} \times (1 - r) \times d \\ G_{billable} & = & \max (0, G_{month} - A) \\ E_{indexed} & = & \frac{G_{month} \times 1048576}{k \times 1000000} \times s \\ C & = & (G_{month} p + G_{billable} i + E_{indexed} x + G_{month} q t + F + U) \times (1 - c) \end{array}

Symbols used in the observability cost formula
Symbol	Meaning	Unit or basis
r	Collector reduction	Decimal share from 0 to 0.90
d	Billing days	1 to 31 days per month
A	Included ingest allowance	GB per month
k	Average event size	KB per event
s	Indexed share	Decimal share from 0 to 1
p, i, x	Processing, ingest, and event-index rates	USD per stated meter
q	Extra retention in 30-day periods	max(0, target minus included days) divided by 30
t	Extended-retention rate	USD per GB per 30 days
F, U	Fixed monthly and user costs	USD per month
c	Contract discount	Decimal share from 0 to 0.80

The forecast for month n multiplies every daily stream by (1 + monthly growth)^{n - 1}, then reruns the full cost formula. This preserves fixed charges and included allowances rather than scaling the first month's total directly.

Accuracy Notes:

Public pricing profiles are snapshots and intentionally omit meters the model does not represent. Grafana metrics active-series charges, vendor-specific query charges, archive storage, rehydration, taxes, regional premiums, and commitment mechanics can change the invoice. Replace every example rate with the current order form or billing export before using the result for a budget decision.

Worked Examples:

Mixed-meter contract

Daily volume of 1 TB of logs and 100 GB of traces becomes 550 GB per day after a 50% collector reduction. Across 30 billing days that is 16,500 GB. With 500 GB included, $0.05 per processed GB, $0.40 per billable GB, a $1.50 per million indexed-event meter based on 2 KB events and a 25% indexed share, 30 extra retention days at $0.10 per GB, $200 in fixed and user charges, and a 10% discount, the monthly estimate is $11,087.13. This case shows why reducing collection affects more cost components than changing only the indexed share.

References:

Transforming telemetry, OpenTelemetry documentation, updated May 22, 2025.
New Relic pricing, New Relic.
Grafana Cloud logs, traces, and profiles pricing and usage, Grafana Labs.
How to check Logstash pipeline metrics, Simplified Guide.