Leaders at companies like American Express see cloud adoption as inevitable because of the economies of scale and the engineering and innovation capacity that major cloud providers bring. These providers can invest heavily in capabilities such as AI, machine learning, and advanced analytics, which would take internal teams years to build on their own.

The real question is no longer if organizations will move to the cloud, but how and how fast. CIOs and IT teams should focus on:

• Cloud model choices: Deciding between private, public, or hybrid cloud, and which specific mix best fits their business and regulatory needs.
• Business outcomes: Measuring success based on overall business impact, not just a narrow comparison of cloud versus on-premises costs.
• Migration strategy: Planning how to modernize applications so they can take advantage of cloud-native capabilities, rather than simply lifting and shifting legacy systems.

In short, cloud is treated as a new way to deliver technology services that can reshape how organizations operate, compete, and serve customers.

American Express uses a hybrid cloud architecture that lets teams build applications once, inside secure containers, and then deploy them on private cloud, public cloud, or both. This approach gives the company flexibility in where workloads run, while maintaining consistent security and deployment practices.

Cloud-native development is central to this strategy. Applications are:

• Highly available: Designed to keep serving customers across a wide range of scenarios.
• Scalable: Built to scale out and in, so they can handle changing transaction volumes efficiently.
• Reusable and portable: Based on reusable components and best practices in logging, configuration management, port binding, and dependency mapping, which support portability across environments.

Running the same application on multiple clouds also increases resiliency and supports the “always on” experience that more than 100 million card members expect. Over time, American Express expects the economies of scale from this cloud strategy to create meaningful economic advantages compared with traditional data centers.

For leaders planning a cloud move, the emphasis should be on maximizing business outcomes, not just “moving to cloud” as a goal in itself. Key considerations include:

1. Application strategy, not just lift-and-shift

• Avoid pure lift-and-shift of legacy applications; it often delivers limited benefit.
• Prioritize moving applications that are cloud-native or can be modernized, fully or partially, to get better returns.
• Distinguish between true technical debt (out-of-date apps that need refreshing) and current apps that simply aren’t cloud-ready yet.
• Use application refresh cycles to gradually modernize and migrate your portfolio.

2. Cost perspective and “bubble costs”

• A common misperception is that “the cloud is cheaper” or “the cloud is more expensive.” In reality, cloud is a different delivery model.
• Focus on the total cost of the technology stack and the value it delivers, not a simple cloud vs. on-premises price comparison.
• Plan for temporary ‘bubble costs’ while you run parts of applications on both cloud and traditional infrastructure during migration.

3. Talent, mindset, and the role of infrastructure teams

• Invest in training and upskilling engineers on cloud tools and practices.
• Equally important, invest in evangelizing the right mindset so teams understand how to design and operate for cloud.
• Do not assume “if you build it, they will come.” Infrastructure teams need to act as trusted partners, consultants, and advisors to software engineering teams.
• Infrastructure leaders should see themselves as curators of the developer experience, ensuring cloud capabilities are understood and adopted in the right way.

By combining thoughtful application modernization, realistic cost planning, and strong support for engineering teams, CIOs can reimagine how their organizations use cloud to drive business results.