Cloud Economics & Migration: Cost Models, TCO, ROI, Strategies

Key Concepts and Definitions

1. Cloud Cost Models: CAPEX vs OPEX

Cloud economics revolves around two fundamental cost models:

• CAPEX (Capital Expenditure): Traditional on-premises model where you buy hardware upfront (e.g., servers, storage). High initial cost but ownership.
• OPEX (Operational Expenditure): Cloud model where you pay for usage (e.g., AWS EC2 hourly billing). No upfront cost but ongoing expenses.

mindmap
  root((Cloud Cost Models))
    CAPEX
      "Upfront purchase of hardware"
      "High initial investment"
      "Ownership of assets"
      "Depreciation over time"
    OPEX
      "Pay-as-you-go model"
      "No upfront cost"
      "Scalable expenses"
      "Usage-based billing"

Why it matters: Cloud shifts IT budgets from CAPEX to OPEX, improving cash flow and flexibility.

In the Real World

1. Ncell’s Cloud Migration (AWS): Ncell migrated its core network services to AWS to reduce CAPEX by 40%. They used reserved instances for predictable workloads (e.g., customer billing systems) and pay-as-you-go for variable traffic (e.g., data roaming spikes during festivals like Dashain). The Total Cost of Ownership (TCO) analysis showed a 3-year payback period.
2. Daraz’s Elastic Scaling (Alibaba Cloud): During the Dashain sale, Daraz uses auto-scaling to handle 10x traffic. Their TCO includes costs for:
   • Compute: Pay-as-you-go for web servers (₹1.20/hour per instance).
   • Storage: ₹0.20/GB/month for product images.
   • Networking: ₹0.05/GB data transfer. Their ROI is calculated by comparing cloud costs to the revenue generated during peak seasons.
3. Nepse’s Disaster Recovery (Azure): Nepse uses Azure Site Recovery for its trading platform. The cost includes:
   • Replication: ₹500/month for syncing data between primary and secondary regions.
   • Failover Testing: ₹200 per test (quarterly). Their ROI is measured by avoiding downtime during market crashes (e.g., 2015 Nepse hack).

2. Total Cost of Ownership (TCO)

TCO is the total cost of owning and operating a system over its lifetime. For cloud, it includes:

• Direct Costs: Compute, storage, networking, software licenses.
• Indirect Costs: Labor, training, downtime, compliance.
• Hidden Costs: Data egress fees, support contracts, vendor lock-in.

TCO Calculation Example: NTC’s Cloud vs On-Premises

Assume NTC wants to host its 1000-node network monitoring system:

Key Insight: Cloud TCO is often lower for variable workloads but can exceed on-premises for steady, long-term use.

3. Return on Investment (ROI)

ROI measures the profitability of cloud adoption. Formula: $$\text{ROI}=\frac{\text{Net Benefit}}{\text{Total Cost}}\times 100\%$$ Where:

• Net Benefit = Savings (cost reduction) + Revenue (new features, uptime).
• Total Cost = TCO over the investment period.

Worked Example: Pathao’s Cloud ROI

Pathao migrated its driver-matching algorithm to Google Cloud:

• Cost Savings:
  • Reduced server costs from ₹8,000,000/year (on-prem) to ₹4,000,000/year (cloud).
  • Eliminated ₹2,000,000/year in hardware upgrades.
• Revenue Gains:
  • Faster response time → 15% increase in rides during peak hours (₹10,000,000/year).
• Total Benefit: ₹14,000,000/year.
• Total Cost: ₹4,000,000/year (cloud) + ₹1,000,000 (migration).
• ROI: $$\text{ROI}=\frac{₹14,000,000-₹5,000,000}{₹5,000,000}\times 100\%=180\%\text{ (paid back in <1 year)}.$$

4. Cloud Pricing Models

Cloud providers offer multiple pricing models. Compare them for a Khalti transaction processing system:

Visual Comparison:

pie
  title Cloud Pricing Model Adoption (Nepal 2023)
  "Pay-as-you-go" : 45
  "Reserved Instances" : 30
  "Savings Plans" : 15
  "Spot Instances" : 5
  "Flat-Rate" : 5

5. Cloud Migration Strategies

Migrating to the cloud requires careful planning. The 6 R’s of Migration (AWS framework) are widely used:

flowchart TD
  A["Lift and Shift\n(Rehost)"] -->|"Minimal change"| B["Replatform\n(Lift, tinker, shift)"]
  A -->|"No change"| C["Repurchase\n(Buy cloud SaaS)"]
  A -->|"Re-architect"| D["Refactor\n(Re-architect for cloud)"]
  A -->|"Replace"| E["Retire\n(Shut down unused apps)"]
  A -->|"Optimize"| F["Retain\n(Keep on-premises)"]

Strategy Breakdown with Nepali Examples

Worked Example: Kathmandu Traffic Management Assume Kathmandu Metropolitan City wants to migrate its traffic signal system to the cloud:

1. Current State: On-premises servers with ₹500,000/year maintenance.
2. Migration Path:
   • Phase 1 (Rehost): Move to AWS EC2 (₹300,000/year).
   • Phase 2 (Replatform): Use AWS IoT Core for real-time sensor data (₹400,000/year).
   • Phase 3 (Refactor): Implement AI-based traffic optimization (₹600,000/year but reduces congestion by 20% → ₹100M/year in savings).
3. ROI:
   • Year 1: Cost = ₹300,000, Benefit = ₹20M (from reduced accidents).
   • Year 3: Cost = ₹600,000, Benefit = ₹120M.

6. Migration Challenges and Mitigation

Real Picture:

7. Cloud Economics Tools

Exam Tip

1. Memorize Key Formulas:
   • TCO = Direct Costs + Indirect Costs + Hidden Costs.
   • ROI = (Net Benefit / Total Cost) × 100%.
2. Compare Models:
   • Always contrast CAPEX vs OPEX, pay-as-you-go vs reserved instances.
3. Real-World Scenarios:
   • Expect questions on Nepali companies (e.g., "How would NTC calculate TCO for its cloud migration?").
   • Know the 6 R’s of migration and when to use each.
4. Cost Optimization:
   • Discuss right-sizing, spot instances, and reserved capacity in answers.
5. Diagrams:
   • Draw TCO comparison tables and migration strategy flowcharts in exams.

Final Note: Cloud economics is not just about saving money—it’s about aligning costs with business goals. Always tie your answers to scalability, agility, and ROI. For example, when asked about migration strategies, explain how refactoring enables auto-scaling, which directly impacts customer satisfaction (e.g., Daraz during sales).