Every cloud migration business case starts with the same optimistic math: current infrastructure costs minus projected cloud costs equals savings. The spreadsheet looks compelling. Leadership approves.
Then reality arrives.
Eighteen months later, the organization is mid-migration, over budget, and questioning whether cloud was the right decision. Not because cloud doesn’t deliver value—it does—but because the original business case missed entire categories of cost that only become visible during execution.
We’ve guided over 50 enterprise migrations to Google Cloud. The patterns of budget surprise are remarkably consistent. Organizations that understand true migration costs upfront build realistic plans that succeed. Organizations that rely on vendor TCO calculators build plans that fail.
This guide breaks down what enterprise cloud migrations actually cost—not what vendors want you to believe they cost.
The Cost Categories Everyone Budgets
Let’s start with the obvious costs that appear in every migration business case:
Compute and Storage (The Visible Infrastructure)
This is where most cost analysis begins and, unfortunately, where many end. The math seems straightforward:
| Current State | Google Cloud Equivalent |
|---|---|
| 200 physical servers | Compute Engine instances |
| 500TB SAN storage | Cloud Storage + Persistent Disk |
| 3 data centers | 2 GCP regions |
Vendors provide calculators. You input your current specs. Out comes a monthly estimate.
The problem: These calculators assume like-for-like replacement. Enterprises don’t migrate to run identical workloads identically—they migrate to transform. The instance types you need post-migration rarely match your pre-migration estimates.
Typical variance: 20-40% higher than initial calculator estimates, primarily due to:
- Performance requirements that demand larger instance types than expected
- High availability configurations not fully accounted for
- Development and staging environments that expand in cloud
- Logging, monitoring, and security services that add incremental costs
Networking Costs (The Overlooked Multiplier)
Network egress—data leaving Google Cloud—is where many budgets first encounter reality.
| Egress Scenario | Approximate Cost |
|---|---|
| Data to internet | $0.08-0.12/GB |
| Data to other clouds | $0.08-0.12/GB |
| Data between regions | $0.01-0.08/GB |
| Data within region | Free |
For an enterprise moving 100TB monthly to end users or partner systems, egress alone can exceed $100,000 annually.
The hidden pattern: Organizations underestimate egress because they don’t fully understand their current data flows. On-premises, network traffic between systems is invisible—it’s just “the network.” In cloud, every byte crossing boundaries has a price.
Mitigation strategies we recommend:
- Architect for regional data locality where possible
- Use Cloud CDN for content delivery to end users
- Negotiate committed use discounts that include network
- Consider Premium vs. Standard network tiers based on actual requirements
Software Licensing (The Complicated Middle)
Enterprise software licensing in cloud environments is genuinely complex. Microsoft, Oracle, VMware, and others have licensing models that interact with cloud in non-obvious ways.
Common surprises:
Oracle: Licensing Oracle databases on cloud infrastructure requires careful navigation. Oracle’s licensing policies for cloud environments have changed repeatedly, and running Oracle on Google Cloud Bare Metal requires specific license considerations.
Microsoft: Windows Server and SQL Server licensing on Google Cloud can follow bring-your-own-license (BYOL) models or use Google’s included licensing. The optimal approach depends on your existing Microsoft agreements.
VMware: Organizations using VMware Cloud on Google Cloud (GCVE) face different economics than those migrating away from VMware entirely.
Our guidance: Engage licensing specialists before migration planning, not during. The licensing strategy often influences the technical architecture.
The Cost Categories Most Budgets Miss
Here’s where business cases diverge from reality:
Migration Execution Costs
Moving workloads from on-premises to cloud requires effort—significant effort. This isn’t just “lift and shift.”
| Migration Activity | Typical Cost Range |
|---|---|
| Application assessment and planning | $5,000-25,000 per application |
| Refactoring for cloud optimization | $25,000-150,000 per application |
| Data migration and validation | $10,000-100,000 per major database |
| Testing and validation | 20-30% of migration effort |
| Cutover and hypercare | $50,000-200,000 for enterprise migrations |
The uncomfortable math: An enterprise with 100 applications might face $2-5 million in migration execution costs before a single workload runs in production. This is professional services, internal labor reallocation, and contractor costs.
Why organizations underestimate: Migration is treated as a one-time project rather than a transformation program. The assumption is that existing teams can absorb migration work alongside their normal responsibilities. In practice, migration requires dedicated resources.
Dual-Running Costs (The Transition Tax)
During migration, you’re paying for both environments simultaneously.
For an 18-month migration program (typical for large enterprises), expect 12+ months of overlapping infrastructure costs. If your current infrastructure runs $2 million annually, dual-running adds $2+ million to migration costs.
The extended reality: Migrations rarely complete on schedule. Every month of delay extends dual-running costs. We typically model 150% of planned migration duration for financial planning.
Skills and Training Investment
Cloud platforms require different skills than on-premises infrastructure. Your team needs to learn:
- Google Cloud console and CLI operations
- Infrastructure-as-code practices (Terraform, Deployment Manager)
- Cloud-native monitoring and observability
- Security models fundamentally different from perimeter-based approaches
- Cost management and FinOps practices
| Training Investment | Typical Range |
|---|---|
| Formal certification training | $3,000-10,000 per engineer |
| Learning curve productivity loss | 20-30% for 3-6 months |
| External expertise during transition | $200,000-500,000 annually |
The hidden cost: Productivity loss during the learning curve. Teams that managed on-premises infrastructure efficiently will be slower on cloud platforms initially. This translates to delayed projects, longer incident resolution, and increased reliance on external support.
Organizational Change Management
Technology changes. Processes must change with it.
Procurement: Cloud consumption doesn’t fit traditional capex procurement models. Finance teams need new processes for usage-based billing, commitment management, and cost allocation.
Security: Cloud security reviews, compliance validations, and governance frameworks require updates. Security teams need training on cloud-specific threats and controls.
Operations: On-call rotations, incident response procedures, and capacity planning all require modification for cloud operations.
Typical organizational change costs: $200,000-500,000 for enterprise transformations, plus 6-12 months of reduced operational efficiency.
Application Modernization Debt
Migrating applications as-is (lift and shift) is faster but defers optimization. Applications designed for on-premises environments run in cloud—but expensively.
The deferred cost pattern:
- Year 1: Migrate applications as-is, costs higher than expected
- Year 2: Begin optimization efforts to reduce costs
- Year 3: Applications finally running efficiently
Organizations that budget only for migration discover they’ve also committed to multi-year optimization programs.
Our recommendation: Budget modernization alongside migration. The applications that benefit most from cloud—elastic scaling, managed services, serverless—require architectural changes. Plan that work upfront rather than treating it as surprise scope.
The TCO Reality: A Composite View
Let’s build a realistic cost model for a representative enterprise migration:
Scenario: Mid-size enterprise, 150 applications, 2 PB of data, currently spending $5M annually on infrastructure.
Year 1 Costs (Migration Year)
| Category | Conservative Estimate |
|---|---|
| Migration program (professional services + internal) | $2,500,000 |
| Dual-running infrastructure (12 months) | $5,000,000 |
| Google Cloud infrastructure (ramping) | $2,000,000 |
| Training and skills development | $400,000 |
| Licensing transitions | $300,000 |
| Organizational change management | $300,000 |
| Year 1 Total | $10,500,000 |
Year 2 Costs (Optimization Year)
| Category | Conservative Estimate |
|---|---|
| Google Cloud infrastructure (full load) | $4,500,000 |
| Optimization and modernization | $800,000 |
| Ongoing skills development | $150,000 |
| Cloud operations team | $600,000 |
| Year 2 Total | $6,050,000 |
Year 3 Costs (Steady State)
| Category | Conservative Estimate |
|---|---|
| Google Cloud infrastructure (optimized) | $3,800,000 |
| Ongoing modernization | $300,000 |
| Cloud operations team | $600,000 |
| Year 3 Total | $4,700,000 |
The Three-Year View
| Period | Cost | vs. Status Quo ($5M/year) |
|---|---|---|
| Year 1 | $10,500,000 | +$5,500,000 |
| Year 2 | $6,050,000 | +$1,050,000 |
| Year 3 | $4,700,000 | -$300,000 |
| 3-Year Total | $21,250,000 | +$6,250,000 |
The honest conclusion: This migration costs $6.25M more than staying put over three years, measured purely on infrastructure costs.
Where the Business Case Actually Works
If the numbers above seem discouraging, we’re not finished. Cloud migration business cases work—but not on infrastructure cost savings alone.
The Value Drivers That Justify Migration
Speed to market: Organizations on cloud deploy new capabilities in weeks rather than months. A single product launch accelerated by three months can justify years of migration investment.
Elasticity economics: The ability to scale for demand peaks without provisioning for peak changes the economics of seasonal businesses, marketing campaigns, and product launches.
Innovation access: AI/ML capabilities, managed databases, serverless computing, and global infrastructure enable applications impossible to build on-premises.
Operational resilience: Multi-region availability, automated failover, and infrastructure-as-code reliability exceed what most enterprises achieve on-premises.
Talent acquisition: Engineers increasingly expect cloud-native environments. Organizations on legacy infrastructure face hiring disadvantages.
Quantifying the Value
The enterprises that build successful cloud business cases quantify these benefits:
| Value Driver | Quantification Approach |
|---|---|
| Speed to market | Revenue from earlier launches |
| Elasticity | Avoided infrastructure for peak capacity |
| Innovation | Revenue from new capabilities |
| Resilience | Avoided downtime costs |
| Talent | Reduced recruiting costs, lower turnover |
Example: An e-commerce company spending $5M on infrastructure might save only $500K through cloud migration. But if cloud enables them to handle 3x Black Friday traffic without provisioning year-round capacity, the elasticity value alone could exceed $2M annually.
Building a Realistic Migration Business Case
Based on our experience, here’s how to build a business case that survives contact with reality:
Step 1: Honest Current State Assessment
Don’t use theoretical costs. Gather actual spending:
- Infrastructure invoices (hardware, maintenance, data center)
- Software licensing (including compliance true-ups)
- Personnel costs for infrastructure operations
- Incident costs (downtime, recovery efforts)
Common discovery: Actual current-state costs exceed budget assumptions by 20-40% when all categories are included.
Step 2: Migration Cost Modeling
Use ranges, not point estimates:
| Category | Low | Expected | High |
|---|---|---|---|
| Migration execution | $1.5M | $2.5M | $4M |
| Dual-running | $3M | $5M | $7M |
| Training/change | $400K | $700K | $1.2M |
Plan for the expected case. Reserve budget for the high case.
Step 3: Value Quantification
Work with business stakeholders to quantify:
- What would faster deployment cycles enable?
- What’s the cost of current capacity constraints?
- What innovation is blocked by infrastructure limitations?
- What’s the current cost of downtime?
Conservative quantification is more credible than optimistic assumptions.
Step 4: Timeline Realism
Whatever migration timeline seems reasonable, extend it:
- Assessment and planning: 3-6 months
- Migration execution: 18-36 months for enterprises
- Optimization: 12-24 months post-migration
Rushed migrations fail more often than slow migrations.
Step 5: Executive Alignment
Present the business case honestly:
- Year 1 costs will increase significantly
- Break-even typically occurs in Year 3-4
- The case depends on value creation, not just cost reduction
Executives who understand the investment profile make better decisions than those sold on unrealistic savings projections.
The Committed Use Discount Factor
Google Cloud offers significant discounts for committed usage—1-year and 3-year commitments can reduce compute costs by 20-57%.
The planning challenge: Committing before understanding actual usage leads to either:
- Under-commitment: Missing available discounts
- Over-commitment: Paying for unused capacity
Our recommendation: Run in cloud for 6-12 months before making significant commitments. The data from actual usage patterns enables informed commitment decisions.
When Migration Makes Sense—And When It Doesn’t
Strong Migration Candidates
- Organizations with variable workloads poorly served by fixed infrastructure
- Companies pursuing AI/ML initiatives requiring modern platforms
- Businesses expanding globally needing distributed infrastructure
- Organizations struggling with infrastructure talent acquisition
- Companies where infrastructure constraints limit business growth
Weaker Migration Candidates
- Stable workloads with predictable, constant demand
- Applications with minimal evolution planned
- Organizations without business drivers beyond cost reduction
- Companies with recent significant infrastructure investments
- Regulated environments where cloud approval timelines are years, not months
Moving Forward
Cloud migration is an investment, not a cost-cutting exercise. Organizations that approach it as investment—with realistic costs, quantified benefits, and appropriate timelines—succeed. Organizations seeking quick savings find disappointment.
The numbers in this guide are illustrative. Your organization’s actual costs depend on your specific applications, data volumes, compliance requirements, and transformation ambitions.
Planning a Google Cloud migration?
We help enterprises build realistic migration business cases, design transformation roadmaps, and execute migrations that deliver on their promises. Our assessments examine your specific environment—not generic assumptions.
Schedule a migration assessment to understand what cloud transformation actually looks like for your organization.