Everything you need to know about reducing your Google Cloud bill and how PrecisionTech delivers FinOps services
.
1
What are Google Cloud Committed Use Discounts (CUDs) and how much can they save?
Committed Use Discounts (CUDs) are Google Cloud's commitment-based pricing model — you agree to purchase a minimum level of compute or memory resources for 1 or 3 years and receive significant discounts over on-demand rates. Key CUD types: (1) Compute Engine CUDs — committed vCPU and memory in a specific region (e.g., asia-south1 Mumbai). Up to 57% savings for 3-year commitments, up to 37% for 1-year; (2) Resource-based CUDs — flexible across machine types and families within a region (similar to AWS Compute Savings Plans); (3) Cloud SQL CUDs — committed vCPU and memory for managed databases; (4) Cloud Run CUDs — committed CPU and memory for serverless containers. Unlike AWS Reserved Instances, CUDs apply automatically to matching usage — no instance-type lock-in for resource-based CUDs. Best practice: collect 30–60 days of baseline usage data before purchasing, target 70–80% of stable baseline (not 100%), and reassess quarterly. PrecisionTech analyzes your Billing export and Recommender data to recommend optimal CUD amounts — typically saving ₹2–8 lakh/month for mid-size Indian GCP deployments.
2
What are Sustained Use Discounts and do I need to configure them?
Sustained Use Discounts (SUDs) are automatic, no-commitment discounts Google Cloud applies when you run Compute Engine VMs for a significant portion of the billing month. How they work: as a VM runs more hours in a calendar month, Google progressively applies deeper discounts — up to 30% off on-demand pricing for VMs running the full month. SUDs apply automatically — no purchase, no configuration, no upfront payment. They stack with CUDs: if you have a CUD covering part of your usage, SUDs apply to the remaining on-demand portion. Important nuances: (1) SUDs apply per VM instance, not aggregated across your fleet — 10 VMs each running 50% of the month get less discount than 5 VMs running 100%; (2) Preemptible/Spot VMs do not receive SUDs; (3) SUDs vary by machine family — N2 and E2 families have the most generous SUD tiers. For Indian businesses running production workloads 24×7, SUDs alone can reduce compute costs 20–30% with zero effort. PrecisionTech includes SUD analysis in every cost assessment — often identifying workloads that should run continuously (to maximize SUD) vs workloads that should use Spot or scheduled shutdown.
3
How do Spot VMs and Preemptible VMs work on Google Cloud?
Spot VMs (formerly Preemptible VMs) let you use spare Compute Engine capacity at up to 91% off on-demand pricing. Google can reclaim Spot VMs with a 30-second warning when capacity is needed elsewhere. Key capabilities: (1) No maximum runtime — Spot VMs can run indefinitely until preempted (Preemptible VMs had a 24-hour limit; Spot removed this); (2) Spot provisioning models — use Managed Instance Groups (MIGs) with autoscaling and multiple machine types for resilience; (3) GKE Spot nodes — run batch, CI/CD, and fault-tolerant workloads on Spot node pools with automatic rescheduling; (4) Batch workloads — Dataflow, Dataproc, and Cloud Run Jobs support Spot/preemptible workers natively; (5) Live migration fallback — configure MIGs with a mix of Spot + standard VMs so critical capacity always exists. Best practices: use termination handlers (graceful shutdown scripts), checkpoint long-running jobs, diversify across machine types and zones, and never run stateful databases on Spot. PrecisionTech designs Spot architectures for Indian SaaS, fintech, and data platforms — typically achieving 60–85% savings on batch, rendering, ML training, and stateless web tiers.
4
What is the Google Cloud Recommender API and how does it help reduce costs?
The Recommender API is Google Cloud's ML-powered optimization engine that analyzes your resource utilization and generates actionable recommendations across cost, security, performance, and reliability. Cost-related recommenders include: (1) Machine type rightsizing — identifies over-provisioned Compute Engine VMs based on CPU, memory, and network metrics (e.g., downgrade n2-standard-8 averaging 15% CPU to n2-standard-4); (2) Idle resource reclamation — flags unused persistent disks, idle Cloud SQL instances, orphaned static IP addresses, and idle load balancers; (3) CUD recommendations — suggests optimal Committed Use Discount purchases based on historical usage patterns; (4) Cloud SQL rightsizing — recommends smaller database tiers when utilization is low; (5) Idle GKE clusters — identifies underutilized Kubernetes clusters; (6) IAM role optimization — removes unused permissions (security + cost via audit log reduction). Recommendations appear in the GCP Console (Billing → Cost Management → Recommendations) and via the Recommender API for automation. PrecisionTech integrates Recommender findings into monthly FinOps reviews, prioritizing by savings impact — a typical first Recommender sweep for an Indian enterprise identifies ₹1–4 lakh in monthly savings from idle resources and right-sizing alone.
5
How does GCP billing export to BigQuery work and why is it essential for FinOps?
Billing export to BigQuery delivers your complete Google Cloud billing data — every line item, every hour — into a BigQuery dataset for SQL analysis, custom dashboards, and chargeback reporting. Setup: (1) Enable billing export in Cloud Console → Billing → Billing export → BigQuery export; (2) Choose Standard usage cost (detailed line items) and/or Pricing export (list prices and contract prices); (3) Data lands in a BigQuery table updated daily (typically by 7 AM IST); (4) Query with SQL — e.g., SELECT service.description, SUM(cost) FROM billing.gcp_billing_export WHERE invoice.month = '202602' GROUP BY 1 ORDER BY 2 DESC. Why it matters: (1) Label-based allocation — allocate costs to teams, projects, and cost centres using resource labels; (2) Custom dashboards — build Looker Studio or Looker dashboards from BigQuery for finance and engineering; (3) Chargeback/showback — generate monthly per-department cost reports; (4) CUD analysis — calculate effective CUD coverage and identify on-demand leakage; (5) Anomaly detection — schedule queries to alert on spend spikes. PrecisionTech sets up billing export, BigQuery views, and Looker Studio FinOps dashboards as the foundation of every GCP cost optimization engagement.
6
What does a FinOps monthly review process look like on Google Cloud?
A structured FinOps monthly review keeps GCP spend visible, accountable, and continuously optimized. PrecisionTech's monthly FinOps cadence: (1) Week 1 — Data collection — pull Billing export from BigQuery, review Budget alerts triggered last month, export Recommender recommendations, and check CUD utilization reports; (2) Week 1 — Analysis — compare actual vs budget by project, label, and service; identify top 10 cost drivers; flag anomalies (e.g., Cloud Storage growth, unexpected egress, new GPU usage); (3) Week 2 — Engineering review — present findings to engineering leads: right-sizing candidates, idle resources to terminate, label compliance gaps, and architecture recommendations; (4) Week 2 — Finance review — present chargeback/showback reports, CUD ROI, forecast vs actual, and commitment renewal timeline; (5) Week 3 — Implementation — execute approved changes (VM rightsizing, disk cleanup, lifecycle policy updates) during maintenance windows; (6) Week 4 — Governance — update budgets, review Organization Policy constraints, assess new Recommender recommendations, and document savings achieved. Deliverables each month: executive summary, savings tracker, optimization backlog, and updated Looker Studio dashboard. This continuous loop prevents cost regression — the #1 failure mode after a one-time optimization project.
7
How do I identify and clean up idle GCP resources?
Idle resource cleanup is the fastest path to GCP savings — often 5–15% of total spend with zero architectural change. Common idle resources PrecisionTech finds in Indian GCP environments: (1) Stopped but billed VMs — stopped Compute Engine instances still incur persistent disk and static IP charges; (2) Unused persistent disks — disks detached from deleted VMs, often 100–500 GB each at ₹850–4,250/month; (3) Idle Cloud SQL instances — databases with zero connections for 7+ days, costing ₹15,000–80,000/month; (4) Orphaned static IP addresses — reserved IPs not attached to any resource (~₹300/month each); (5) Idle load balancers — forwarding rules with no healthy backends (~₹1,400/month each); (6) Unused GKE clusters — clusters with minimal pod activity but full control plane + node costs; (7) Old snapshots and images — accumulated backup snapshots from decommissioned workloads; (8) Empty Cloud Storage buckets — buckets with lifecycle gaps accumulating objects. Detection tools: Recommender API (idle resource recommendations), Billing export queries (resources with zero usage but positive cost), Cloud Asset Inventory, and custom Cloud Functions for automated scanning. PrecisionTech runs a comprehensive idle resource audit as day-one activity in every engagement — typically recovering ₹1–5 lakh/month immediately.
8
Does Google Cloud support INR billing and what are the GST implications in India?
Yes. Google Cloud supports INR billing for Indian customers. When your billing account currency is set to INR, all invoices are generated in Indian Rupees with GST applied. GST framework: (1) Billing entity — Google Cloud India Private Limited (GCIPL) is the invoicing entity for India, registered under GST; (2) GST rate — 18% GST applies to cloud computing services (classified as OIDAR — Online Information and Database Access or Retrieval Services under SAC 998315); (3) Input tax credit (ITC) — GST-registered businesses can claim ITC on GCP invoices against their output GST liability, effectively reducing the net cost of cloud services by 18%; (4) Tax invoices — GCP provides GST-compliant tax invoices with GSTIN, HSN/SAC codes, place of supply, and reverse charge indicators where applicable; (5) TDS considerations — certain government and large enterprise customers may need to deduct TDS on OIDAR payments under Section 194-O or applicable provisions — consult your CA; (6) INR vs USD pricing — INR list prices are set by GCIPL and may include a currency adjustment factor vs global USD pricing. PrecisionTech helps Indian businesses configure INR billing, structure billing accounts for proper GST compliance, implement label-based cost allocation aligned with your chart of accounts, and ensure finance teams receive GST-compliant documentation for ITC claims.
9
How do Google Cloud Billing budgets and alerts work?
Cloud Billing budgets let you set custom spend thresholds and receive automated alerts — preventing bill surprises before the invoice arrives. Budget types and features: (1) Amount budgets — set a monthly, quarterly, or annual spend limit (e.g., ₹15 lakh/month for production) with alerts at 50%, 80%, 90%, and 100% of budget; (2) Filter scopes — create budgets scoped to specific projects, services, labels, or credit types (e.g., a budget only for Cloud SQL in the production project); (3) Alert channels — notifications via email, Pub/Sub (for programmatic response), and Slack/Teams via webhook integration; (4) Forecasted spend alerts — GCP predicts end-of-month spend based on current run rate and alerts if you're trending over budget mid-month; (5) Programmatic actions — Pub/Sub-triggered Cloud Functions can auto-disable billing, send Slack alerts, or create Jira tickets when thresholds are breached. Best practices: create budgets at multiple levels — organization-wide, per-environment (prod/staging/dev), and per-team (using label filters). Set the 80% alert to trigger investigation, not panic. PrecisionTech configures multi-tier budget alerts for every GCP billing account as part of our FinOps practice — including Pub/Sub automation that notifies our managed services team within minutes of an anomaly.
10
How does GCP labeling strategy work for cost allocation?
A labeling strategy assigns key-value metadata to GCP resources, enabling cost allocation by business dimension in Billing export and Cost Management reports. Recommended mandatory labels: (1) environment — production, staging, development, sandbox; (2) team or department — engineering, data, marketing, platform; (3) application or project — the product or service the resource supports; (4) cost-centre — maps to your accounting system for chargeback; (5) owner — email of the responsible engineer (for idle resource follow-ups). Implementation on GCP: (a) Define labels in Organization Policy or a central tagging standard document; (b) Enforce using Organization Policy constraints — e.g., constraints/gcp.resourceLabels to require specific labels on resource creation; (c) Remediate untagged resources using Recommender (label management recommendations), Cloud Asset Inventory scans, and automated Cloud Functions; (d) Enable label-based billing reports in Billing → Cost Management → Reports, grouped by label key; (e) Query labeled costs in BigQuery billing export with labels.value WHERE labels.key = 'team'. Without consistent labeling, 20–40% of GCP spend is unallocatable. PrecisionTech designs and enforces labeling strategies as the foundation of every GCP FinOps engagement.
11
How do I optimize Google Kubernetes Engine (GKE) costs?
GKE cost optimization addresses the control plane, node pools, networking, and workload efficiency layers. Key strategies: (1) Autopilot vs Standard — GKE Autopilot charges per pod resource request (no idle node waste); Standard mode requires active node pool rightsizing. Choose Autopilot for variable workloads, Standard with careful node management for predictable high-density workloads; (2) Node pool rightsizing — use Vertical Pod Autoscaler (VPA) and Recommender to right-size node machine types; enable cluster autoscaler with min/max bounds per pool; (3) Spot node pools — run fault-tolerant workloads (batch, CI/CD, stateless APIs) on Spot nodes — up to 91% savings; use taints/tolerations to schedule only eligible pods on Spot; (4) Bin packing — set appropriate pod resource requests/limits to maximize node utilization; avoid over-requested CPU/memory that blocks scheduling; (5) Committed Use on nodes — apply CUDs to stable baseline node pools running 24×7; (6) Network cost reduction — use internal load balancers, minimize cross-zone traffic, and enable GKE Dataplane V2 for efficient networking; (7) Cluster consolidation — merge underutilized clusters; delete idle namespaces and orphaned LoadBalancer services (~₹1,400/month each); (8) Off-hours scaling — scale dev/staging node pools to zero outside business hours using scheduled scaling or Keda. PrecisionTech optimizes GKE environments for Indian SaaS and fintech clients — typical savings: 30–50% on Kubernetes spend within 60 days.
12
How does Cloud Storage tiering reduce costs on Google Cloud?
Cloud Storage tiering matches data to the most cost-effective storage class based on access frequency — reducing storage costs by 60–95% for infrequently accessed data. Storage classes: (1) Standard — hot data accessed frequently (~₹1.70/GB/month in asia-south1); (2) Nearline — data accessed less than once per month (~₹0.85/GB/month, 30-day minimum); (3) Coldline — data accessed less than once per quarter (~₹0.42/GB/month, 90-day minimum); (4) Archive — long-term archival, accessed less than once per year (~₹0.17/GB/month, 365-day minimum). Automation options: (a) Lifecycle management rules — auto-transition objects between classes (e.g., logs: Standard → Nearline after 30 days → Coldline after 90 days → Archive after 365 days → Delete after 7 years); (b) Autoclass — fully automatic tiering based on access patterns (no retrieval fees, ideal for unpredictable access); (c) Object versioning cleanup — lifecycle rules to delete old versions and reduce storage bloat. Cost impact: a company with 50 TB of logs and backups on Standard storage (~₹85,000/month) can reduce to ~₹8,500/month with proper tiering — saving approximately ₹9 lakh per year. PrecisionTech audits every Cloud Storage bucket during cost optimization engagements and implements lifecycle policies + Autoclass tailored to each data type.
13
How do Google Cloud cost management tools compare to AWS?
Both platforms offer robust FinOps tooling, but differ in architecture and maturity: Cost visibility — GCP Billing Reports provide service/project/label breakdowns with 18-month history; AWS Cost Explorer offers 13-month history with ML forecasting. GCP's billing export to BigQuery provides comparable granularity to AWS CUR. Commitment models — GCP offers CUDs (resource-based, flexible) plus automatic Sustained Use Discounts (no commitment needed); AWS offers Savings Plans and Reserved Instances (no automatic sustained discount). GCP's SUDs are a unique advantage for workloads without commitment appetite. Recommendations — GCP Recommender API covers cost, security, and performance; AWS Compute Optimizer + Trusted Advisor cover similar ground. GCP Recommender integrates natively with Billing. Spot/preemptible — GCP Spot VMs (91% off, 30-second notice) and AWS Spot Instances (90% off, 2-minute notice) are comparable; GCP Spot has no 24-hour limit. FinOps stack — GCP: Billing export → BigQuery → Looker Studio/Looker; AWS: CUR → Athena → QuickSight. Both are powerful; GCP's BigQuery-native approach suits teams already on the Google data stack. India pricing — GCP asia-south1 (Mumbai) and asia-south2 (Delhi) offer competitive INR pricing; AWS ap-south-1/ap-south-2 are comparable. PrecisionTech supports both platforms and recommends the FinOps stack native to whichever cloud you're on — or both for multi-cloud environments.
14
What GCP FinOps services does PrecisionTech provide?
PrecisionTech delivers end-to-end GCP FinOps services for Indian businesses — from one-time assessments to ongoing managed FinOps. Our service portfolio: (1) Cost Optimization Assessment — comprehensive review of your GCP environment in 5–10 business days, delivering a prioritized savings action plan with INR impact estimates; (2) FinOps foundation setup — billing export to BigQuery, Looker Studio dashboards, budget alerts, labeling strategy design and enforcement, CUD procurement; (3) Monthly FinOps managed service — ongoing bill analysis, Recommender implementation, CUD management, right-sizing execution, anomaly response, and executive reporting; (4) Chargeback/showback implementation — label-based cost allocation, BigQuery chargeback queries, monthly department-level reports; (5) Architecture cost optimization — GKE rightsizing, Cloud Storage tiering, Spot VM adoption, serverless migration (Cloud Run), and network cost reduction; (6) FinOps training — enable your internal team to run FinOps independently with playbooks, dashboards, and governance policies. We operate as your outsourced FinOps team or train your team to self-serve. Every engagement is backed by our Google Cloud-certified architects and India-based support team.
15
What typical cost savings can I expect from GCP cost optimization?
Savings vary by current maturity, but here are typical ranges PrecisionTech delivers on Google Cloud: (1) Quick wins (Week 1–2) — terminate idle resources (VMs, Cloud SQL, disks, static IPs), delete unused snapshots, release orphaned load balancers. Typical savings: 5–15% of total spend; (2) Right-sizing (Week 2–4) — implement Recommender VM and Cloud SQL recommendations, optimize persistent disk types and sizes. Typical savings: 10–20%; (3) Commitments (Month 1–2) — purchase CUDs for stable baseline workloads after 30–60 days of usage data. Typical savings: 20–40% on committed capacity; (4) Architecture (Month 2–6) — GKE optimization, Cloud Storage tiering, Spot VM adoption, Cloud Run migration, network cost reduction. Typical savings: 15–30% additional; (5) FinOps practice (Ongoing) — continuous governance prevents cost regression and captures savings from new GCP pricing and services. Overall, enterprises that have never optimized typically see 25–45% total spend reduction within 90 days. Mature environments see 10–20% further savings. These ranges are based on PrecisionTech's engagements with Indian SaaS, fintech, e-commerce, and enterprise GCP deployments billed in INR.
16
Is PrecisionTech an Authorized Google Cloud Partner with ISO certifications?
Yes. PRECISION e-Technologies Pvt Ltd (PrecisionTech.in) is an Authorized Google Cloud Partner in the Google Cloud Partner Advantage Program, providing GCP cost optimization, FinOps, migration, GKE, BigQuery, security, and managed services across India. Our team holds Google Cloud certifications including Professional Cloud Architect, Professional Data Engineer, Professional DevOps Engineer, and Professional Cloud Security Engineer. Quality and security certifications: (1) ISO 9001:2015 — Quality Management System, ensuring consistent delivery methodology across all engagements; (2) ISO 27001:2022 — Information Security Management System, critical when our team accesses your GCP billing and infrastructure data during cost optimization; (3) CMMI Level 3 — Capability Maturity Model Integration, demonstrating defined and managed processes. With 30+ years serving Indian businesses, PrecisionTech combines Google Cloud expertise with enterprise-grade governance — so your FinOps engagement is delivered by a certified, auditable partner, not a freelance consultant. Contact us for a complimentary GCP cost optimization assessment.
17
What is Google Cloud Cost Management and how do Billing Reports help?
Google Cloud Cost Management is the native FinOps toolkit within the GCP Console — providing visibility, analysis, and optimization without third-party tools. Key components: (1) Billing Reports — interactive charts showing cost trends by service, project, SKU, label, and location over customizable date ranges. Filter, group, and drill down to identify top spend drivers; (2) Cost breakdown — pie/bar charts showing which services consume the most budget (Compute Engine, Cloud Storage, BigQuery, GKE, Cloud SQL typically dominate); (3) Cost forecasting — projected end-of-month spend based on current run rate, with confidence intervals; (4) Commitment analysis — CUD coverage percentage, effective savings rate, and recommendations for additional commitments; (5) Credits and adjustments — track promotional credits, sustained use discounts applied, and committed use discounts; (6) Export capabilities — CSV export for ad-hoc analysis, or automated BigQuery export for programmatic FinOps. Limitations: Billing Reports lack the SQL flexibility of BigQuery export and the ML recommendations of the Recommender API — use all three together. PrecisionTech configures Cost Management dashboards as the first step in every FinOps engagement, giving finance and engineering teams shared visibility into GCP spend from day one.
18
What is chargeback vs showback and how do I implement it on GCP?
Showback shows each business unit their GCP costs for awareness — without actually billing them internally. Chargeback allocates actual cloud costs to each business unit's P&L or budget — making them financially accountable for their cloud consumption. Implementation on GCP: (1) Foundation — implement mandatory cost allocation labels (team, application, cost-centre, environment) on all resources via Organization Policy; (2) Data pipeline — enable billing export to BigQuery, create views that join billing data with label keys, build Looker Studio dashboards grouped by team/project; (3) Shared costs — distribute shared infrastructure (VPC, Cloud NAT, Cloud Logging, Security Command Center, shared GKE clusters) proportionally based on compute consumption or a fixed allocation formula; (4) Reports — monthly automated reports per cost centre showing: direct labeled costs + proportional shared costs + CUD savings attribution + credit adjustments; (5) Governance — set per-team budgets with alerts, track budget vs actual monthly, and review in FinOps meetings. Common challenge: unlabeled resources (typically 20–40% in immature environments). PrecisionTech's approach: enforce 100% label compliance first, then implement showback for 3 months before moving to full chargeback — giving teams time to understand and adjust their consumption patterns.
19
What is cost governance on GCP and how do I prevent cloud cost overruns?
Cost governance on Google Cloud is the set of policies, guardrails, and processes that prevent unauthorized or uncontrolled spending. PrecisionTech implements GCP cost governance through: (1) Organization Policy constraints — restrict resource creation by region (e.g., asia-south1 and asia-south2 only for India data residency), machine type (block GPU instances except in approved ML projects), and required labels; (2) IAM permissions boundaries — limit what individual users and service accounts can provision without approval; (3) Billing budgets with Pub/Sub actions — automated Cloud Functions that alert, restrict, or notify when spend exceeds thresholds; (4) Quota management — set project-level quotas on Compute Engine CPUs, GPUs, and persistent disks to cap maximum spend; (5) Approval workflows — require manager approval for resource creation above a cost threshold, implemented via Cloud Functions triggered by Audit Logs or custom internal tooling; (6) Sandbox projects — isolated GCP projects with hard budget limits (e.g., ₹10,000/month) for experimentation; (7) Monthly FinOps reviews — review actual vs budget with engineering leads, track optimization backlog, and prevent cost regression. These layers create defence-in-depth against runaway spend — essential for Indian enterprises where a single misconfigured Dataflow job or forgotten GPU instance can add ₹2–5 lakh to the monthly bill.
20
How does PrecisionTech's GCP Cost Optimization Assessment work?
PrecisionTech's GCP Cost Optimization Assessment is a comprehensive review of your entire Google Cloud environment, delivered in 5–10 business days. The process: (1) Access setup — we request read-only IAM access (roles: Billing Account Viewer, Viewer, Recommender Viewer) to your GCP organization and billing account; (2) Data collection — we analyze Billing Reports, BigQuery billing export (or enable it if not configured), Recommender recommendations, CUD utilization reports, and Cloud Monitoring metrics across all projects; (3) Analysis — we evaluate: idle and underutilized resources, Recommender right-sizing opportunities, CUD coverage gaps, Cloud Storage class optimization, GKE cluster efficiency, Spot VM adoption candidates, network egress costs, labeling compliance, and architectural inefficiencies; (4) Report — we deliver a prioritized action plan with: estimated monthly savings in INR for each recommendation, implementation effort (quick win vs project), risk level, and dependency mapping; (5) Presentation — we walk your engineering and finance teams through the findings with a Q&A session. Typical assessment identifies 25–45% savings. As an Authorized Google Cloud Partner with ISO 9001 and ISO 27001 certifications, we handle your billing data with enterprise-grade security. Contact us for a complimentary assessment.
21
What are common GCP cost mistakes that Indian businesses make?
PrecisionTech commonly identifies these cost mistakes during assessments of Indian GCP deployments: (1) Oversized Compute Engine VMs — provisioning for peak capacity instead of using Managed Instance Groups with autoscaling. An n2-standard-8 running at 12% CPU should be n2-standard-2 with autoscaling; (2) No Committed Use Discounts — running 100% on-demand when 60–70% of workloads are stable and predictable. Missing 30–57% savings from CUDs plus automatic Sustained Use Discounts; (3) Cloud Storage without lifecycle policies — keeping years of logs and backups in Standard class when they should transition to Nearline → Coldline → Archive; (4) Unlabeled resources — 30–50% of resources have no cost allocation labels, making chargeback impossible and hiding departmental spend; (5) Orphaned persistent disks — disks from deleted VMs still billing ₹850–4,250/month each; (6) Dev/staging running 24×7 — development environments that should scale to zero outside business hours (10 hours/day, 5 days/week); (7) Over-provisioned GKE node pools — fixed-size node pools at maximum capacity when cluster autoscaler with appropriate min/max would right-size automatically; (8) No billing export to BigQuery — relying only on Console reports without the SQL-level analysis needed for FinOps; (9) Ignoring Recommender — dozens of cost recommendations sitting unreviewed in the Console. Each of these is a quick win waiting to happen.