The ¥5 billion, 5-year Azure commitment triggered by Microsoft's 2025 license price revision is not merely cost defense.
It is a turning point to extend the cloud utilization capabilities proven with AWS to Azure,
completing multi-cloud as a true business transformation platform.
Fix the roles of AWS (defense), Azure (offense), and GCP (marketing) to realize consistent multi-cloud operations across the entire group.
Create competitive advantage through 5 execution axes: data integration, AI utilization, citizen development, and global connectivity. Drive as an integrated platform, not isolated initiatives.
Establish a CCoE structure, strategic collaboration with Microsoft, and TCS operational ecosystem to build a reproducible execution framework.
Please support the effort to advance Azure not as an "IT department challenge" but as a "Group management priority," enabling the structure for organization-wide execution.
The direct trigger (MS license price revision) and a structural shift (cloud becoming a management platform) coincide—a unique moment for Suntory.
This shift is neither a tailwind nor a headwind — it is an environmental change where the ability to respond creates competitive differentiation.
| Axis of Change | 📉 Before | 📈 Going Forward |
|---|---|---|
| Role of Cloud | Tool for IT infrastructure efficiency | Management platform that determines management speed and transformation capability |
| Source of Competitive Advantage | AWS was virtually the only choice | Design capability and control to use different clouds based on business value |
| Level of AI Utilization | PoC stage for operational efficiency | Advanced decision-making and redesign of business processes themselves |
| Role of Data | Distributed by BU | Utilized as a strategic asset for decision-making across the global organization |
| Security & Governance | Central control via AWS only | Horizontal, automated governance premised on multi-cloud |
The three clouds show clear differences in maturity: AWS is a complete foundation, Azure is strategically important but incomplete, and GCP is limited to supplementary use.
The current analysis reveals not individual cloud technical issues, but three structural problems at a deeper level. Without resolving these, no strategy will have execution power.
The strategic thinking of treating multi-cloud as a management prerequisite is aligned with market trends, but the decision-making units, investment structures, and design philosophy to drive it are not organized at the same layer. Expectations based on AWS success are running ahead, while the conditions AWS went through—initial standard design, operational structure, partner formation—have not been given to Azure.
Azure is positioned as the core of multiple strategic themes including data integration, AI utilization, and citizen development. However, it remains at the level of "concept," "policy," or "individual implementation," not complete as a "company-facing product" with clear usage, responsibility, and cost. This makes scaling data and AI utilization across the organization structurally impossible.
Design, decision-making, and operational responsibilities are distributed, and the organizational unit for cross-functional governance and advancement is unclear. Azure utilization depends on individual coordination skills and field-level efforts—no reproducibility or scalability. Despite positioning Microsoft as a strategic partner, joint governance, role allocation, and decision-making models are not adequately designed.
Rather than "which cloud is superior," roles are clearly defined by "which cloud can most rationally realize which value." Fixing roles enables consistent multi-cloud operations across the entire group.
Rather than positioning Azure simply as the "second cloud" after AWS, we focus on its ability to design holistically across Global, Cross-functional, and End-to-End dimensions, positioning it as a strategic platform complementary to AWS.
Windows Server and SQL Server licenses can be used more affordably than on AWS. Maximizing synergies with existing Microsoft investments reduces the Total Cost of Ownership (TCO) for cloud migration.
By placing the data lake on the same cloud across the company and connecting Power Platform and Azure within the same region, inter-cloud communication egress costs are minimized.
Secure Reference: Through Microsoft Graph connectors, Azure data can be safely used in Copilot responses while maintaining Entra ID permissions. Deep M365 integration enables simultaneous information utilization and control.
Authentication Integration: Entra ID enables consistent authentication from applications to databases.
Secure RAG Construction: Combining Azure AI Search with Private Link enables data search and generative AI integration within a private network.
Confidentiality settings applied to Azure data are automatically applied to responses generated by Copilot. End-to-end maintenance of data handling rules enables consistent information governance up to the generative AI utilization domain.
The five axes generate maximum value only when operated in conjunction. The key is to design and drive them as an integrated platform, not as isolated initiatives.
From a "One Suntory" perspective, integrate data fragmented across Suntory Japan, SBF, and SGS, establishing a "management data foundation where decisions can be made based on the same numbers" globally. Make Microsoft Fabric / Power BI the core of the data lake, realizing a Single Source of Truth for swift executive decision-making.
Position Copilot (business front-end) and Azure OpenAI (business system-integrated AI) as the company-wide standard AI. Internalize as organizational capability business judgments, document creation, and analysis that previously relied on individuals, evolving AI into a sustained productivity improvement platform embedded in business processes.
Make Power Platform the company-wide standard business digitalization platform, clearly role-allocating IT-led large-scale development and field-led responsive improvements. By securely connecting business systems and field applications, and leveraging the AI platform described above, realize the transformation into "an organization where the field continuously improves autonomously."
Standardize infrastructure, CI/CD, security, and monitoring as a common platform, providing the foundation for "building quickly and operating stably in the same way" across businesses and countries. Build an Internal Developer Platform (IDP) for self-service use, establishing reproducible IT delivery capability across the entire group (Platform Engineering).
Through a network infrastructure centered on Azure Virtual WAN, escape from on-premises-centric assets and centrally connect and govern Azure, AWS, GCP, and on-premises. Transition from MPLS to SD-WAN to achieve both cost optimization and improved operational productivity, while decommissioning the current global backbone.
| Domain | 📉 Current State (As-Is) | 🎯 Target State (To-Be) |
|---|---|---|
| Technical Standards | Separate standards at SI2 and SGS, person-dependent updates | Integrated Azure standards, semi-automated updates with AI support |
| Data Platform | Fragmented across 3 companies, difficult cross-reference | Virtual integrated data lake, global cross-reference possible |
| AI Utilization | Each individual selects and uses their own AI tools, scattered and not established | Adopt an Azure OpenAI–first approach for server-side GenAI |
| Development Platform | Manual, person-dependent, slow speed | High-speed development via self-service portal and IaC automation |
| Network | MPLS-centric, complex global connectivity | SD-WAN / Virtual WAN consolidated, cost optimized |
| Operations | Manual response, TCS/individual dependent | Automated/semi-automated incident response using AIOps |
| Organization & Advancement | Individual coordination dependent, no cross-functional governance | CCoE-based governance, MS strategic partnership functioning |
The Azure commitment period (5 years) is divided into 3 phases, progressively increasing maturity. The phase concept is "Build the Foundation → Scale Utilization → Realize Transformation."
In response to Structural Challenge ③, establish a governance model that realizes cross-functional strategy. Not "restrictive governance" but "enabling governance"—pre-arrange standards, guardrails, and operational models so field speed is not impeded.
Build an operational ecosystem for Azure that exceeds what was established for AWS
| Role | Responsibility | Personnel / Organization |
|---|---|---|
| IT Project Leader | End-to-end management responsibility for scope, milestones, and outcomes | Tom (SHD) |
| Executive Sponsor | Determination of strategic direction and support for major decisions | Lusman (SHD), Ian (SGS) |
| Enterprise Architecture Lead | Definition of target architecture and standardization approach | Rohan (SHD), Joshua (SBF), TCS Architect Team (TBD) |
| Engineering Lead | Execution of Azure service implementation and construction | Rohan (SHD), Son (SHD), TCS Engineering Team (TBD) |
| Security & Governance Team | Ensuring compliance and security policy adherence | Turner (SHD), Vincent (SBFI) |
| VMO (Vendor Management Office) | MS commitment achievement status management, contract/commercial coordination | Janice (SHD) |
| Azure Operations Team | Operations preparation, runbook creation, BAU transition support | Rohan (SHD), Son (SHD), TCS Operations Team (TBD) |
| Business Unit Representatives | Requirements provision, design validation, implementation support | M Cruz (SBFE), Ram (SBFI), Mitsu (SBFA), TBD (SBFO), Larry (PBV), etc. |
Strategy progress and outcomes are continuously measured and visualized from both quantitative and qualitative perspectives.
| KPI | Target | Measurement Timing |
|---|---|---|
| Azure Commitment Achievement Rate | ¥5 billion cumulative over 5 years (on-plan consumption) | End of each fiscal year |
| Azure Standardization Coverage | Standardization rate for major managed services: 95%+ | End of Phase 1 (March 2027) |
| Azure Engineer Count | 60+ engineers equivalent to Azure Solutions Architect Associate | End of Phase 1 (March 2027) |
| Data Integration Coverage | Data lake integration rate for 3 companies' major business data | End of Phase 2 (March 2029) |
| SD-WAN Migration Rate | Completion rate of migration from MPLS to SD-WAN | End of Phase 3 (March 2031) |
Overview of costs for Azure foundation construction and operations. Costs for individual services based on BU-specific requirements are separate.
| Cost Type | Item | Cost / Conditions | Notes |
|---|---|---|---|
| Initial Cost (Capex) | Initial build cost for each BU's SI2 Azure environment | TBD | Newly established TCS Azure team scheduled to handle implementation |
| Initial Cost (Capex) | Initial setup cost for dedicated Azure connectivity | TBD | Depends on service fees of each BU's telecommunications carrier |
| Monthly Cost (Opex) | Monthly base infrastructure cost for each BU's Azure subscription | USD 2,600/month | Network and security costs for Azure Foundation Infrastructure |
| Monthly Cost (Opex) | TCS Azure Operations Team cost | TBD | Calculated on FTE-based model unlike AWS; cost allocation method to Suntory Global currently undecided |
Key risks are evaluated by likelihood and impact, with specific response policies defined.
Key terms and concepts used in this strategy document.
| Term | Explanation |
|---|---|
| Azure Landing Zone | Common infrastructure, security, network, and monitoring design for Azure. The standard environment forming the "foundation" for all services. |
| CCoE | Cloud Center of Excellence. A cross-functional team responsible for cloud standardization, governance, and architecture support. |
| Entra ID | Microsoft's identity management platform (formerly Azure AD). Core of cross-cloud authentication and authorization. |
| Microsoft Fabric | Microsoft's data integration platform. Integrates data lake, data warehouse, and analytics. |
| Platform Engineering | Construction and operation of an Internal Developer Platform (IDP) enabling developers to use infrastructure self-service. |
| FinOps | Operational framework for cloud cost visualization, optimization, and budget management. |
| AIOps | AI-powered operational automation (incident detection, response, preventive maintenance). |
| SD-WAN | Software-Defined WAN for network management. More flexible and cost-efficient than MPLS. |
| QBR | Quarterly Business Review. Quarterly management/strategy review meeting. |