Most of the value from cloud does not come from cutting IT infrastructure costs. According to McKinsey research, about 90% of the value at stake in the cloud comes from: - Faster time-to-market - Increased innovation - Improved resiliency - Cost savings in business operations (not just IT) A practical way to prioritize is to ask: “Where will speed, agility, and scalability change outcomes for our customers or our business?” High-potential areas typically include: - Customer-facing experiences (e.g., digital self-service, customer service, personalization) - Data-driven products and services - Workloads that need to scale up and down quickly Lower-potential areas are often highly stable, back-office processes that don’t benefit much from agility (for example, some accounts receivable processes). One large brokerage used cloud to build a new application portfolio and saw feature development speed increase by 5x while cutting operating expenses by 90%. That kind of result comes from intentionally targeting business value, not just “lifting and shifting” existing systems. To capture this value: 1. Identify business capabilities where speed and innovation matter most. 2. Align cloud investments (teams, skills, and budget) to those capabilities. 3. Measure outcomes in terms of customer impact, time-to-market, and operational performance, not just IT spend.

Simply moving applications to the cloud while keeping traditional ways of working usually limits the benefits. Many organizations still rely on frequent handoffs, long review cycles, and manual testing. In that setup, the cloud is like a high-performance car used only for short errands. Companies that see better results typically do two things: 1. **Shift to a product-based operating model** Instead of organizing around projects and functional silos, they: - Treat everything as a “product” (for example, e-commerce product displays, purchase confirmation flows, or personalized email journeys). - Assign small, cross-functional teams (engineering, product, design, operations) to own these products end-to-end. - Make a single team accountable for delivering a working outcome, not just a piece of it. McKinsey’s experience with clients shows that successful product-oriented models can improve development and release productivity by about **20–25%**. 2. **Automate the delivery pipeline** To support this model, organizations invest in automating as much of the software lifecycle as possible, including: - Server provisioning - Infrastructure-as-code generation - Testing and deployment This reduces manual work, shortens release cycles, and makes it easier to use cloud capabilities consistently. If your cloud migration hasn’t translated into faster delivery or better quality, it’s a signal that the operating model—not just the technology—needs to be rethought around products, automation, and end-to-end ownership.

Managing cloud for value is an ongoing discipline, not a one-time migration task. Two areas matter in particular: cost optimization and foundational capabilities (reliability, security, and architecture). 1. **Optimize cloud economics continuously** Cloud is easy to consume, which often leads to higher-than-expected bills. To keep costs aligned with value, organizations typically: - Track usage in real time - Forecast demand more accurately - Automate scaling and resource management When implemented properly, these practices can reduce cloud costs by **20–30%**, and sometimes more. One entertainment company reviews every cloud application and system monthly, focuses on the most expensive ones, and optimizes them (for example, by using serverless services). They report savings of about **$3 million per year per engineer** involved in this optimization work. 2. **Invest in foundational capabilities early** Rushing to migrate or build applications without a solid foundation can create delays, technical debt, and security or resiliency gaps. Key foundational elements include: - Standard reference architectures - Automation for provisioning and configuration - Resilient designs (e.g., redundancy, failover patterns) - Security-as-code (embedding security policies directly into configuration scripts) IT resilience alone represents almost **15%** of the total cloud value at stake. With more resilient architectures, cloud can reduce downtime for migrated applications by nearly **60%**. For example, when one payments company moved its data centers to the cloud, it nearly doubled availability and cut transaction times from **12 seconds to 5 seconds**. 3. **Migrate complete services, not just individual apps** Moving isolated applications often leads to fragmented experiences and limited performance gains. A better approach is to migrate or rebuild complete business services end-to-end (for example, mortgage origination or retail payments), including: - Customer-facing interfaces - Integration and ecosystem layers - Core transaction processing - Data stores and analytics (e.g., fraud detection) - Compliance and servicing components This creates a critical mass of mutually reinforcing applications that can fully use cloud capabilities. In practice, organizations that treat cloud as a managed, evolving platform—optimizing costs, strengthening foundations, and migrating complete services—are more likely to see sustained improvements in performance, reliability, and economics.