India’s Leading Domain & Web Hosting Provider

  • Chat
    Live Chat 24 X 7 online support
  • Contact NTSPL Hosting support
    Call Now +91 - 8260003333

Cloud Cost Assessment: Audit Your Live Cloud Bill for Waste and Gaps

Quick answer

A cloud cost assessment is a forensic audit of an organisation's existing cloud bill that quantifies waste, measures discount coverage and shows how much spend cannot be traced to an owner. NTSPL runs it over four to six weeks on AWS, Azure and Google Cloud estates spending ₹10 lakh or more a month, and hands finance and engineering a ranked, costed savings register.

Delivered on

AWS: 5 native toolsMicrosoft Azure: 5 native toolsGoogle Cloud: 5 native tools

In scope

Bill reconciliationWaste inventoryRate coverageAllocation gapsArchitecture cost driversUnit economics
Cloud Cost Optimization & FinOps · Cloud Cost Assessment

What a Cloud Bill Audit Reveals That Dashboards Hide

Most Indian companies that moved to cloud between 2019 and 2023 now pay a bill nobody fully understands. Spend grows 4-8% a month, the CFO sees one invoice line per provider, and engineering sees dashboards that do not reconcile to that invoice. Typically 30-45% of resources carry no owner tag, commitment coverage sits below 40%, and a quarter of compute runs at under 10% average CPU. Without a baseline, every optimisation debate becomes opinion, and savings claimed by one team are quietly eaten by another team's growth.

NTSPL starts from the billing data, not from consoles: we ingest 13 months of AWS CUR 2.0, Azure Cost Management exports or the Google Cloud BigQuery billing export, normalise them to the FOCUS specification and reconcile the totals to the invoice within 0.5%. Every finding is priced in INR per month, tied to a named resource set, and scored for effort and risk, so the register can be executed rather than admired.

Signals Your Cloud Spend Needs an Independent Audit

Signals you need Cloud Cost Assessment

  • Monthly cloud spend has grown more than 50% year on year while revenue or active users grew less than 20%.
  • Finance cannot attribute more than 30% of the bill to a product, business unit or customer.
  • Savings Plan, reservation or CUD coverage is unknown, or below 50% of steady-state compute.
  • A board, investor or PE due-diligence team has asked for unit economics such as cloud cost per order or per active user.
  • You run two or more providers through different resellers and no one has a consolidated monthly view.

When to choose something else

  • You are still on-premises and deciding whether to move: you need a migration TCO and business case (cloud-assessment family), not a bill audit.
  • Spend is under ₹3 lakh a month: native advisors (Trusted Advisor, Azure Advisor, Google Cloud Recommender) will surface most of the value without a paid engagement.
  • The audit was done in the last six months and the register is unexecuted: fund execution through Rightsizing or Continuous FinOps Governance instead of re-auditing.

Seven Lenses Applied to Your Billing Data

01

Bill reconciliation

Billing exports are reconciled to provider invoices, reseller credits, marketplace charges and GST lines so every later number ties back to what finance actually paid.

02

Waste inventory

Idle, orphaned and oversized resources are listed with monthly cost, last-activity date and probable owner, using 30-90 days of utilisation metrics.

03

Rate coverage

Savings Plans, reservations and CUDs are measured for coverage and utilisation per service and region, including commitments expiring within 120 days.

04

Allocation gaps

Tag and label completeness is measured by spend, not by resource count, and shared costs (networking, support, security tooling) are identified for split rules.

05

Architecture cost drivers

Data transfer, NAT, cross-zone traffic, log ingestion and premium storage are traced to the design choices causing them.

06

Unit economics

Spend is joined to one or two business drivers (orders, MAU, API calls, policies issued) to produce a cost-per-unit baseline and trend.

Interactive framework · Cloud Bill Forensics Pipeline

Billing-Data-to-Savings-Register Analysis Pipeline

Cloud Bill Forensics Pipeline

Raw billing exports from each provider are landed, normalised to FOCUS and reconciled to invoices before any analysis runs. Utilisation metrics and inventory are then joined to the cost lines to produce a priced, owner-mapped savings register and a unit-cost baseline.

Billing ingestion
Normalise and reconcile
Evidence join
Findings and outputs
Interactive reference architecture

Select any component

Cloud Bill Forensics Pipeline: click any of its 10 components to see its role and the design choice NTSPL makes for it, or trace the 7-step primary path.

AWS CUR 2.0 / Data Exports → FOCUS-normalised cost table → Invoice and GST reconciliation → Utilisation metrics join → Resource inventory and tags → Priced savings register → CFO/CTO readout pack

How the Audit Runs on AWS, Azure and Google Cloud Billing

AWS

CUR 2.0 is queried in Athena; Cost Optimization Hub recommendations are used as a cross-check, but NTSPL re-prices them because the hub does not know which workloads are being retired.

AWS Data Exports (CUR 2.0)AWS Cost ExplorerAWS Cost Optimization HubAWS Compute OptimizerAmazon Athena
Microsoft Azure

Exports are taken at billing-account or MCA billing-profile scope so EA/MCA reservation amortisation is visible per subscription, which portal views at subscription scope often hide.

Azure Cost Management exportsAzure AdvisorAzure Resource GraphAzure MonitorPower BI
Google Cloud

The detailed (resource-level) export must be enabled first; NTSPL uses the standard export for history and the detailed one from enablement onward.

Cloud Billing export to BigQueryRecommender / Active AssistFinOps hubCloud Asset InventoryLooker Studio

Judgement Calls That Shape the Savings Number

01

Amortised versus billed cost

Report waste on amortised cost, otherwise an upfront reservation paid in April makes April look wasteful and the rest of the year look efficient. Use billed cost only for cash-flow conversations with finance. NTSPL shows both in the readout, but every savings line and every unit-cost KPI uses amortised cost so the numbers survive a quarter-on-quarter comparison.

02

Look-back window

Use 13 months of billing for trends and seasonality, but 30 days of metrics for idle detection and 60-90 days for rightsizing. Shorter metric windows miss month-end batch jobs and GST filing peaks; longer windows dilute the signal from recent launches. If the estate changed materially in the last 60 days, NTSPL flags those resources as 'too new to judge' rather than guessing.

03

Savings counting rules

Agree the counting rule before the audit: annualised run-rate savings, measured against the trailing 30-day cost of the specific resource, net of any purchase. Exclude savings from workloads already scheduled for decommission. NTSPL caps any single finding's claimed value at the resource's observed cost so registers cannot exceed 100% of a line item, which happens in tool-generated reports surprisingly often.

04

Shared and untagged cost

Do not let untagged spend sit as one bucket. Split it into shared platform cost (networking, security tooling, support plans), which needs an allocation rule, and simply untagged workload cost, which needs an owner. NTSPL traces owners through creation events in CloudTrail, Azure Activity Log or Cloud Audit Logs, which usually resolves 60-75% of untagged spend within the audit.

Four-to-Six-Week Cost Audit Sequence

01 · Week 1

Access and data landing

Read-only roles are created, billing exports enabled or backfilled, and data landed in an in-country analysis dataset.

02 · Week 1-2

Reconciliation and FOCUS mapping

Provider data is normalised and matched to invoices; reseller credits and marketplace charges are separated.

03 · Week 2-3

Waste and rate analysis

Utilisation joins, idle detection, commitment coverage and expiry analysis are run per account, subscription or project.

04 · Week 3-4

Owner validation workshops

Findings above ₹25,000 per month are reviewed with the owning engineering lead to mark each one confirmed, disputed or excluded.

05 · Week 4-6

Register and readout

The validated register, unit-cost baseline and 90-day execution plan are presented to the CFO and CTO together.

What you receive

Reconciled 13-month spend baseline by provider, service, region and owner
Priced savings register (CSV and dashboard) with effort, risk and owner per line
Commitment coverage and expiry report with recommended purchase range
Tagging completeness scorecard by spend with untagged-cost owner tracing
Unit-cost baseline for one or two business drivers
90-day execution plan sequenced by savings per engineering day

Baseline Numbers the Assessment Establishes

KPI Typical target
Identified run-rate savings 18-35% of monthly spend
Spend attributed to an owner after tracing Above 85%
Invoice reconciliation variance Below 0.5%
Findings confirmed by owners 70-85% of register value
Savings realised within 90 days 40-60% of confirmed value

Where Cloud Cost Audits Go Wrong

Risk How NTSPL prevents it
Register inflated by tool recommendations that double-count the same resource (rightsize and delete, or rightsize and reserve). One resource, one primary action: NTSPL de-duplicates by resource ID and applies actions in a fixed order (delete, schedule, rightsize, then commit).
Engineering rejects findings as 'finance not understanding the system'. Owner validation workshops happen before the readout, and disputed items are shown separately with the reason recorded.
Billing data leaves the country or is retained by the consultant. Analysis runs in the client's own account in an Indian region; NTSPL access is revoked and the dataset deleted on a dated schedule.
Commitments bought on audit numbers before waste is removed, locking in oversized capacity. The plan sequences cleanup and rightsizing first; commitment recommendations are given as a range recalculated after 30 days.
Buyer comparison

Cloud cost assessment or a pre-migration TCO business case: which do we need?

If workloads already run in cloud and you are paying a monthly bill, you need a cost assessment: it works on actual billing exports and utilisation, and its output is a list of resources to change. A TCO business case works on server inventories and assumptions, compares a three-to-five-year on-premises refresh against cloud, and its output is a go/no-go decision. Companies mid-migration often need both, but in sequence: audit the 40% already in cloud first, because its real unit costs make the business case for the remaining 60% far more credible than list-price estimates.

Key facts: Cloud Cost Assessment

●The FinOps Foundation's FOCUS specification defines a common schema for cloud billing data, and AWS, Microsoft and Google Cloud each provide FOCUS-aligned cost exports.
●AWS CUR 2.0 is delivered through AWS Data Exports and can include resource-level line items when resource IDs are enabled.
●Google Cloud's detailed billing export to BigQuery includes resource-level cost data only from the date it is enabled; it is not backfilled.
●The FinOps Framework organises practice into three iterative phases: Inform, Optimize and Operate.
Illustrative industry case studies

Cloud Cost Assessment in Practice

2 Cloud Cost Assessment scenarios, from SaaS (unicorn-stage B2B software) and Diversified conglomerate (group shared services), plus related platform case studies.

Cloud Cost Assessment: Frequently Asked Questions

How much does a cloud cost assessment cost and how long does it take?

Most assessments take four to six weeks. NTSPL prices them as a fixed fee scaled to the number of accounts, subscriptions or projects and providers; for a single-provider estate of ₹20-60 lakh monthly spend, the fee is typically recovered within the first month of executing the no-regret findings. A fee linked to realised savings is available for larger estates.

Do you need write access to our cloud accounts?

No. The assessment uses read-only billing, monitoring and inventory roles. Any change, even deleting an obviously orphaned disk, is listed in the register and executed later by your team or under a separate change-managed engagement, so the audit never touches production.

Does the billing data analysis stay in India?

Yes. NTSPL lands billing data in a dataset in your own account in AWS Mumbai or Hyderabad, Azure Central India or Google Cloud Mumbai or Delhi. Billing exports can contain user email addresses in tags and resource names, so this supports your DPDP Act 2023 obligations and keeps the data under your access controls.

Can you audit spend bought through a reseller or CSP partner?

Yes, but reseller invoices often apply margins, credits or bundled support that the provider export does not show. NTSPL reconciles provider-native exports to the reseller invoice and reports the gap separately, which sometimes reveals billing errors or credits that were never passed through.

What savings figure should we expect?

Estates that have never been audited typically show 18-35% identified run-rate savings, of which owners confirm 70-85%. Roughly a third is no-regret cleanup, a third needs testing such as rightsizing, and a third needs commitments or design change. Estates with mature FinOps practices usually show 8-12%.

What happens after the cloud cost assessment report is delivered?

The register is sequenced into execution workstreams: idle cleanup, rightsizing, scheduling, storage and commitments. You can execute internally, use NTSPL for specific workstreams, or move into Continuous FinOps Governance so savings are tracked monthly and new waste is caught before it accumulates.