AWS Cloud Enterprise Strategy Blog

Four Key Practices for Profiling Your Enterprise Application Portfolio and Accelerating Your Migration to Cloud

I am very fortunate to have the opportunity to learn from (and hopefully help) hundreds of executives from the largest companies in the world as they plan for and migrate their IT portfolio to the cloud. They rarely do it alone, so I’ve also been fortunate to work with some of the most cloud-fluent system integrators in the world that are helping their customers migrate to the cloud as well. The Accenture AWS Business Group (AABG) are among the best in this space, which is why I’m delighted to host AABG’s Tony Harris’ thoughts on key mass migration practices.

***

Large-scale enterprise migrations to AWS, better known as mass migrations, continue to become more common. Once enterprises determine they’re “all in for the public cloud,” business and IT executives usually want to step on the accelerator and sustain momentum. Rightly so, because moving potentially thousands of applications to the cloud is a major undertaking that can lead to substantial benefits.

But while speed is critically important, there is value to be gained in strategic migration planning. This process includes profiling the application portfolio against cloud strategy objectives, determining target operational and application architecture landing zones, and applying the appropriate migration approach for each zone (or as my colleague Stephen Orban at AWS describes it: The 6 R’s of Application Migration Strategies).

By using a strategic migration planning approach, enterprises will be much more likely to achieve the outcomes they seek from mass migration to AWS, whether that be significant cost savings, increased flexibility and agility, improved time to market and innovation potential — or some combination thereof.

What’s your motivation to migrate?

We see mass migrations to cloud typically driven by:

  1. Compelling Events — mergers/divestitures, datacenter closures, or investment avoidance around hardware upgrades or software license expirations are all examples of a compelling event that drive an enterprise to migrate a large portion of their IT portfolio to AWS all at once. These event-driven migrations typically have a more straightforward business case centered around total cost of ownership and shorter runways to complete the migration and decommissioning activities. Although the timeframe is compressed, it doesn’t mean “lift and shift” is the only option. There are still opportunities (albeit fewer) to both assess the application portfolio and shape the target operating environment on AWS that can result in a more optimized portfolio.
  2. Business-driven — digital transformation initiatives aimed at leapfrogging competitors, introducing scalability to existing or new revenue generating applications/solutions, and improving developer and operational productivity are examples of business-driven migrations. Business-driven migrations are often tied to enterprise-wide agendas and overarching cloud strategies with specific objectives, and tend to have a more complex business case that provides more time to assess the application portfolio holistically. Thus, companies can use different “lenses” to categorize applications into patterns and target landing zones (more on landing zones below) based on what needs to be achieved. With these objectives top of mind, teams can establish multiple migration factory lanes (i.e., fast, medium and slow) with varying migration methods used within each lane to achieve the objectives related to the migration of those specific applications. Parallel factory lanes allow organizations to achieve the pace required in order to migrate an entire portfolio in a short period of time.

Full speed ahead with business-driven migrations

In my experience, business-driven mass migrations position IT to make an even bigger impact on the business than they do today. This type of migration can also reduce barriers between siloed departments, and introduce new possibilities to optimize application performance, improve market positioning and/or kick-start DevOps/DevSecOps in the cloud environment.

Within the Accenture AWS Business Group (AABG), we suggest four steps for moving to AWS at scale:

1. Set a migration plan that is aligned with your cloud strategy and business objectives

Formulate a high-level mass migration plan based on your organization’s overarching cloud strategy and objectives. Don’t over-analyze at this stage — absolute precision is not necessary. As your organization proceeds toward the mass migration, your business case will mature. In addition, make sure key C-suite executives are engaged in devising, or, at the very least, communicating your migration plan to the broader organization. Proving cloud value quickly to top-level stakeholders will solidify sponsorship and confidence to migrate the portfolio.

2. Define target operational and application landing zones

Review your organization’s application portfolio and desired cloud operating model. An operating model is not just an organizational chart of process framework, but a combination of components that articulate how work gets done. Moving to the cloud allows many of those components to be automated; these changes don’t happen by themselves and require careful thought and planning.

Use identified application patterns and operating model guiding principles to shape target landing zones within the core AWS cloud architecture. These landing zones essentially define the high-level target application architecture patterns and capabilities that you align to each application in your portfolio. Alignment of an application to a landing zone will have varying implications on the migration approach, level of effort, risk/benefits and timeline related to the migration of that application.

Our team has defined a framework for shaping these landing zones, with more detailed layers within each based on application patterns and level of cloud optimization and agility desired by our clients. The framework includes the following landing zone parent categories:

  • Minimum Viable Product (MVP) — Tends to incorporate non-strategic applications that need to be maintained, placed on cold standby, or archived. Applications within this zone are typically rehosted in order to migrate to the cloud at the lowest cost/effort point.
  • Enterprise-Grade — Applications that need to have high availability, regular backups, and more advanced security requirements. There may be requirements and opportunities here to replatform (i.e., tweak the application architecture and operating processes) to enable the migration or to leverage some cloud capabilities.
  • Optimized for Cloud — Applications that can be refactored and even reimagined for cloud. This includes application refactoring for optimization of cost, performance, resiliency, and/or agility — as well as applications that can be rebuilt for cloud using serverless architecture approaches that leverage AWS Lambda and Amazon API Gateway. The level of application refactoring/rearchitecting is dependent on both the existing application technical and functional health, and the level of optimization selected within the framework.

Ideally, an enterprise will establish a standard set of landing zones and have a general idea on the cloud capabilities they want to leverage before detailed migration planning begins. This allows the migration teams to align applications to the right migration method, cost profiles, and value models.

3. Establish cloud center of excellence (CoE)

Build a cloud capability to accelerate the mass migration and enable cloud adoption across all business units. This core cloud team can be composed of resources with skills in cloud architecture, application development, security and other areas — from either internal groups or external managed service providers. Of course, let’s also remember that it’s about more than just the technology! Cloud operating model changes, skills development/training plans and change management are also critical to cloud transformation success, and can be orchestrated by the CoE as well. There is some great material on this topic is some of Stephen’s other blog posts: How to Create a Cloud CoE in Your Enterprise and Common Responsibilities of Your Cloud CoE.

4. Perform application assessment and migration planning

Categorize applications to align with the defined migration objectives and appropriate target landing zones. Ask questions such as: What are the most strategic applications? Which ones support current or future business revenue? Which applications can deliver additional benefits beyond cost savings by leveraging new cloud capabilities?

Use this alignment to evaluate migration complexity, risk, and costs, and then assign a migration method to each. For example, applications identified as MVP may just need a rehost approach to achieve cost savings, whereas those identified for optimization will require investments in rearchitecting for new cloud capability.

Lastly, establish a migration plan that encompasses the different migration factory lanes (fast, medium and slow) with the ability to run them in parallel where dependencies and application bundling allow.

In summary, sustaining enterprise-scale cloud migration momentum is challenging. Careful alignment of your cloud strategy, value objectives and migration plan — paired with organizational readiness to scale cloud capabilities — is the recipe for success.

To get started on strategic mass migration planning and enterprise-scale cloud transformation in your company, contact anthony.e.harris@accenture.com.