Migration & Modernization
Aligning Strategy and Execution for Successful Cloud Migration and Modernization Journeys
Moving to the cloud is more than just adjusting your IT setup; it’s a game-changer. It’s not about migrating a few workloads and seeing what yields results. We’re talking about a real, thought-out game plan that changes how we do business. However, many companies struggle to balance control and flexibility when adopting cloud or implementing AI and generative AI. Should they centralize everything or let teams handle it independently? Unfortunately, there isn’t a one-size-fits-all answer. Giving each business unit (BU) the freedom to handle its own cloud strategy and execution can create a messy technology landscape and security headaches. But for some, like global retailers needing local market flexibility, that independence might be exactly what they need to stay competitive. The key is to start with your business goals, then work backwards to find the right balance for your organization.
In this blog, we’ll dive into the pros and cons of different approaches for centralizing or decentralizing the two dimensions of transformation, strategy and execution. Our goal is to help you find the optimal approach for your organization.
Strategy and Execution: two dimensions of transformation
Strategy is your game plan for winning. It’s about figuring out where you want to go and how you’re going to get there. Strategy should tie back to your business outcomes. An example of a cloud migration strategy could be as follows: The organization plans to achieve 99.9% system reliability and reduce IT costs by 40% within 24 months by migrating its on-premises infrastructure to the cloud. But even the best strategy is just a bunch of fancy ideas if you don’t actually make it happen. That’s where execution steps in, taking that strategy and putting it into action. An example of a company’s execution plan could be as follows: Migrate the customer database to Amazon Relational Database Service (Amazon RDS) this week, then refactor the inventory application into microservices and deploy it on Kubernetes next week.
Both are needed. For something as game-changing as cloud adoption, you need to be great at both. Strategy without execution is merely a vision, while execution without strategy is akin to aimless activity; you’re expending energy, but you’re not making meaningful progress toward specific goals.
Overemphasis on strategy risks paralysis. While a solid strategy is essential, many organizations spend too much time perfecting it and not enough time testing or adapting it in real-world conditions. A rigid strategy can also fail to account for dynamic market conditions or emerging technologies.
Strategy and execution aren’t linear. Strategy first, execution second sounds good on paper, but in reality, they often occur in parallel or interact iteratively. Execution can provide critical feedback that reshapes strategy, especially in transformative efforts involving emerging technologies like AI or cloud computing.
Organizational structure isn’t set in stone. The key is to recognize that what works today might not work tomorrow. Maybe you started with a tight, centralized cloud strategy to ensure everyone’s on the same page. As your teams develop cloud expertise, you can gradually provide more flexibility, enabling local innovation while maintaining appropriate governance.
Do not ignore cultural fit. Being successful at transformation needs to consider broader dimensions beyond just strategy and execution. It requires people and cultural alignment. Even a brilliant strategy can falter if employees resist change, lack the right skills, or don’t feel incentivized to execute.
Strategy and Execution Matrix
Let’s explore the pros and cons of centralized and decentralized strategy and execution models (Figure 1). Each has unique strengths and trade-offs, and the choice depends on the organization’s goals, structure, and culture. Boxing these dimensions can seem to oversimplify the complex nature of transformation. However, we offer this to you as a starting point. In our experience, most organizations actually adopt a hybrid approach that evolves over time. As organizations scale, they would eventually want teams to be self-sufficient and agile, with automated guardrails that mitigate risk.
Figure 1: Organizational Models for Strategy and Execution
Centralized Strategy and Centralized Execution Model
Think of this model as a cloud command center, where a single team, like a Cloud Center of Excellence, runs the show. They manage everything: driving the roadmap, governance, migrations, provider and tool selections, costs, and training. It’s great for early cloud adopters, regulated industries, or uniform operations. A good example of this approach is how Dow Jones and Crayon used this model during the initial stages of cloud adoption. But it can feel rigid for businesses needing speed or flexibility, often evolving into other models as organizations scale and diversify.
Centralized Strategy and Decentralized Execution Model
In this model, a central team defines the strategy, establishing standards, security policies, preferred cloud providers, and big-picture goals, while individual BUs are fully accountable for execution. This approach offers the consistency of a centralized strategy with the agility of decentralized execution, making it ideal for diverse operations. For example, leadership in a large conglomerate we worked with stipulated that the company must migrate from data centers to the cloud within 5 years. A central team provisioned cloud accounts, standardized tools, monitored overall spending, approved proper use of cloud services, and managed cloud vendor contracts. The BUs were responsible for the actual migration of workloads to the cloud. The challenge? Coordination. If teams don’t align with central guidelines, it can lead to inconsistencies or duplicated efforts.
Decentralized Strategy and Centralized Execution Model
This model strikes a balance by mixing decentralized strategy with centralized execution. Individual BUs create cloud strategies that suit their specific needs, but a central team steps in to handle the heavy lifting. It’s a good fit when BUs don’t have their own IT teams or for functions where pooling resources helps cut costs. For example, one company centralized initial cloud migrations to ensure consistency and compliance before handing operations back to BUs. Another company centralized tasks like incident response, operating system management, and database administration. The trade-off? It can cause delays as the central team juggles competing priorities and risks misalignment with broader organizational goals.
Decentralized Strategy and Decentralized Execution Model
In this hands-off approach, teams or business units have complete freedom to define and execute their own cloud strategies. It’s a good fit for large conglomerates with subsidiaries in diverse markets, holding companies, or those grown through M&A, where flexibility and autonomy are key. Take a manufacturing giant, for example, where each plant decides its priorities. One might modernize production systems, while another focuses on logistics. The upside? Teams can innovate and move fast. The downside? Without some central oversight, you might face inefficiencies, redundancies, or compliance risks, especially when subsidiaries share a parent brand. For example, two subsidiaries might handle customer data differently, one storing European data on US servers without GDPR compliance, while another follows proper protocols. This inconsistency could expose the parent company to regulatory fines. It’s all about balancing independence with just enough coordination to avoid major pitfalls.
Hybrid Strategy and Execution Model
A hybrid approach balances tight governance with team agility, letting a central team set policies and build self-service platforms with built-in guardrails. This frees BUs to innovate and execute without worrying about IT complexities. For instance, a central team might provide automated workflows for ML model training, testing, deployment, and monitoring from concept to production for consumption by all teams. Capital One’s hybrid data governance framework centralizes policies within a common data platform, simplifying workflows for teams. This model offers value but comes with challenges. Central platform teams can unintentionally stifle team autonomy with excessive controls, causing friction. Their value is indirect, as they save developer time, reduce risks, and accelerate innovation. However, this indirect nature often leads to underfunding and limited effectiveness of these initiatives. Striking the right balance is key for success.
Final Thoughts
Transformation is inherently complex and disruptive. It demands the right balance of strategy and execution based on your organization’s goals and market. You have to be ready to adjust your approach as things change. Centralization gives control but can slow things down, while decentralization offers flexibility but risks misalignment. If you’re new to cloud adoption, a centralized model may make sense. For more diverse operations or as you mature, a centralized strategy with decentralized execution works well. A hybrid model often works best, centralizing core aspects like data, security, and governance while allowing flexibility in execution. Note that what works for one company won’t necessarily fit yours, and that’s okay. The key is aligning your cloud adoption model with your business needs. The AWS Prescriptive Guidance on Cloud Operating Model provides guidance on organizational structures and how they will change as your organization matures.
Which of these approaches do you think fits your organization best?
Additional Reading
5 ways the best companies close the strategy execution gap