Cloud application modernization: PwC

Business and IT leaders understand the need to modernize applications. But identifying the right time, finding the right path and unlocking maximum value from an initiative can prove daunting. All too often, a migration from a mainframe or legacy system to the cloud may fall short. Although an enterprise may see an incremental productivity bump and modest cost savings, more significant gains associated with improved user experience and qualitatively better processes simply don’t materialize. Executives may find themselves asking why.

Typical way

A migration of legacy applications without considering app configurations, dependencies and code tweaks that unlock powerful cloud features like automatic scaling, dynamic storage and advanced networking
An inability to take full advantage of containers, microservices and other powerful cloud features that support advanced technologies like AI
A dependency on large teams and highly specialized coding and IT skills that can lead to higher costs, technical debt and higher administrative overhead
Frequently disjointed applications and workflows that result in diminished user experience and subpar value creation. At worst, initiatives that fail

Better way

A modernization path that can identify all critical information and settings while also accounting for core business value drivers, technical complexity and change management considerations
Rewritten applications and updated settings that can eliminate dependencies and unlock cloud-native features and capabilities
An approach that focuses on both quick and strategic wins that an enterprise can build on. An environment that boosts integration internally as well as with business collaborators
Fast, high-quality releases, lean teams and highly improved business processes that unlock the full value of the cloud

Application modernization isn’t a one-size-fits-all endeavor

The challenges related to app modernization hinge on a straightforward challenge: mainframe computers and legacy code are firmly entrenched in many businesses. Many companies continue to rely on these outdated systems for tasks as diverse as payroll, operations, human resources and general data analytics. Adapting these applications to take advantage of highly agile cloud frameworks with advanced features can be remarkably difficult.

In many cases, it’s tempting to turn to a straightforward replatforming approach (a.k.a. “lift and shift”) to directly transfer software, settings and data from an older system to the cloud. This approach is attractive because it keeps essential apps running and potentially helps an organization incur minimal disruption and keep upfront costs down. But they gain little from being in the cloud.

Then there are companies that attempt to make changes to applications to take advantage of cloud but fall short of their goals. Why? Business and IT groups may not always adequately scope out the value drivers, fully grasp the technical complexities or entirely comprehend the factors that drive change management. Lacking a complete inventory of code, dependencies, workloads and integrations, initiatives often fall short.

Current approach: why replatforming without changes can fail to deliver maximum value

All of this points to an inconvenient truth: Re-platforming alone can deliver limited value, and it can lock an organization into a rigid framework that’s difficult to change. For example, while there are tools to make COBOL code run in the cloud, they may not take full advantage of leading-edge features, such as artificial intelligence (AI) tools. The technical knowledge and time required to connect such legacy code to modern features and application programming interfaces (APIs) can prove daunting.

The resulting problems can play out in unanticipated ways. For instance, as workers with mainframe and COBOL skills retire, an enterprise may find itself saddled with higher costs and technical debt. For instance, a large commercial bank discovered that even after migrating from a mainframe to the cloud, it encountered frustrating limitations. It took months to add new digital services. The reason? It hadn’t modernized systems during the migration. As a result, they were no nimbler than before. Had it been more strategic and taken a modernization path, new releases could have been measured in days.

The lesson: A faster migration isn’t necessarily better — and short-term savings can result in long-term costs. Consider: another large company, after migrating to the cloud, discovered that it had inadvertently created a duplicate payroll system because it had neglected to update its API calls from the legacy system. The controls they had in place did not identify the issue because the monitoring processes were never updated.

To be sure, transferring a monolithic legacy app to the cloud isn’t the same as rewiring it to use highly agile and efficient cloud containers, microservices and serverless frameworks. Smaller, modular pieces of code can be assembled and reassembled to unleash powerful features and highly efficient automation.

What effective app modernization looks like

A better way views an application modernization initiative as an opportunity to gain agility and build a competitive business advantage. By identifying workloads that are ready for modernization and focusing on quick and strategic wins, it’s possible to reduce overhead and technical debt — and embrace progress.

This approach revolves around three foundational concepts:

Think big, but start small. Successful application modernization balances possibility with practicality. It incorporates goals, features and new and improved services grounded in business and IT reality. There’s no “big bang,” and it doesn’t revolve around any specific cloud provider. Effective app modernization floats free of vendors and the whims and preferences of IT staff. It’s all about approaching app modernization from a pragmatic framework that matches an organization’s budget, capabilities and principles to an ideal cloud architecture.
Take things one step at a time. PwC identifies priorities and helps deliver a map for effective migration through a series of clearly defined steps and phases, with appropriate testing, validation and adjustments along the way. It determines the right choices and outcomes. Metrics and key performance indicators (KPIs) are baked into the migration and modernization process. Each step supports the next. For example, an enterprise may identify skill gaps and address them by hiring or training staff before a new feature or service goes live.
Focus on incremental delivery and wins. All too often, application modernization initiatives disappoint because the horizon for return on investment (ROI) extends 3 to 5 years out. With no obvious, early results — and no big wins for the foreseeable future — senior-level executives are prone to throw in the towel. Instead, it’s important to see progress in weeks and measurable results within months. By identifying opportunities that can help deliver quick wins and ROI, it’s possible to build a more sustainable framework for app modernization.

“App modernization is more than a routine technical endeavor. It’s a holistic journey that touches both business and IT — and extends out to workflows and processes. The right application modernization model can deliver a blueprint for success, and a roadmap for business innovation.”

Ussama BaggiliPwC Application Modernization Engineering Leader

3 critical phases make up application modernization

We think of application modernization as a three-step process that transforms applications into a digital-ready state quickly and at scale. This approach breaks down monolithic mainframe and legacy structures and maps them into key functional areas that convert to the new cloud environment. In many cases, it’s possible to complete a plan within six to 10 weeks — and move rapidly into deployment.

Step 1: Engage in discovery

A deep assessment of an IT environment can yield insights about the state of legacy code, dependencies and technical challenges within the stack, data security, and where the greatest cost and value drivers reside. For instance, does an organization want to achieve greater flexibility and faster time to market? Reduce operational complexity and long-term manageability? Unlock advanced analytics and machine learning, or other forms of AI?

Essentially, the discovery process represents an opportunity to reduce business risks that erode value. Once an enterprise clearly understands user journeys, touchpoints and other factors, it can develop a roadmap for modernization, identify specific business cases and streamline tasks. It can tailor workloads to meet specific strategic goals.

For example, an online retailer found that its support process was too complicated. Customers often had to click through multiple screens to find answers — so many gave up and called the company. However, after modernizing the process, an AI-powered chatbot cut inquiries by 30 percent and boosted sales by 5 percent. This allowed the organization to illuminate a path to progress and better meet its KPI goals.

Step 2: Tackle modernization

Understanding how, when and where changes are needed is vital to building out a customized modern cloud infrastructure. This may include an initial phase that focuses on analyzing, converting, recompiling and integrating legacy code. It might also weigh the pros and cons of plugging in commercial off-the-shelf (COTS) software — with input from key stakeholders and careful attention to the request for proposal (RFP) and selection process. This phase also includes a transcoding process, which updates data formats, settings, APIs, reporting, security and other functions.

Throughout this modernization phase, the focus is on improving technology to help deliver the greatest business value and establish gateways for achieving delivery and wins. This technology framework, along with the resulting processes and workflows it changes, matches the needs of specific business groups. For example, this may lead an enterprise to undergo a full code rewrite or a combination of COTS and proprietary coding in preparation for the new framework going live.

Step 3: Operate and optimize

A crucial final step as an organization transitions legacy software code to the cloud is validating results, making essential tweaks and adjustments, and making sure that the modernized code leads to the desired outcomes. Not surprisingly, there are often opportunities to introduce improvements and enhancements along the way. This includes innovation sprints that lead to new products, services and enhancements.

This final phase of the migration confirms that people, processes, tools, metrics and KPIs are all in close alignment. Done right, an organization eliminates manually intensive tasks, steep skill requirements, technical debt and other factors that can undermine business performance and results. In its place, a company gains the ability to help drive continuous improvement, achieve agility and flexibility and, ultimately, undergo business transformation.

What’s next?

App modernization is a moving target. For example, with the rise of generative AI (GenAI) capabilities, it’s no surprise that organizations are looking for ways to integrate it into modernization strategies. Its ability to analyze input and automate and expedite processes can speed modernization efforts. For example, an enterprise GenAI model (like we use at PwC) can help generate personas, work items and test cases. It can also write first drafts of code or troubleshoot your employees’ code.

A microservices architecture — which can simplify and streamline logins, sales, payments, analytics, third-party integrations and more — is now essential. Modernization can also benefit cybersecurity, regulatory compliance, DevOps and numerous other key areas.

An effective app modernization journey can trim unnecessary steps, tasks, screens and processes — delivering an enhanced user experience. With a streamlined IT framework and more dynamic apps in place, an enterprise is better suited to maximize the benefits and value of true app modernization.