If you’re part of a medium to large IT organization, it’s likely that you’re spending an increasing percentage of your application budget just to keep the lights on. This drain on resources is making it harder for you to evolve applications to support the business’s competitive needs. Meanwhile, you face losing critical skills as baby boom retirement creeps closer, combined with risks associated with older technologies being retired.
Application Portfolio Management should be the first step toward a robust, long-term solution, with an opportunity for rapid payback.
Application Portfolio Management (APM) is the practice of continuously assessing the applications that run your business in terms of their business value, enhancement potential, and cost. Derived from the principles of financial portfolio management, APM is especially useful in midsize to large organizations, where years and even decades of software application development and maintenance have resulted in multiple, sometimes redundant, and sometimes outmoded applications still in use across the enterprise. The goal is to identify all working applications, their strengths and weaknesses, and thereby determine where further investments and divestments regarding these applications should be made for the health of the business.
Although APM has been considered a best practice since the mid 1990s, it is seldom applied and is still one of the most immature enterprise practices. In on our discussions with various industry leaders, we have found that of the organizations doing APM, about 70% are using Microsoft Office productivity tools, such as Excel, and find that they do not scale beyond 50 applications. The other 30% percent use tools that fall into one of three capabilities, depending on the culture and heritage of the organization: traditional project portfolio management (PPM), enterprise architecture (EA), or application mining and analytics. Both PPM and EA provide a top down perspective, which looks broadly across the business functions in play and drills down to discover what applications support those functions, focusing on cost-benefit analysis. Application mining and analytics provide a bottom-up perspective, which looks at each application in use, focusing on technical analysis.
IBM believes that there is a need for a simple APM solution that scales beyond basic productivity tools, while providing an out-of-the-box information model and process. This starting point allows companies to get 80% of the value of APM for 20% of the effort. As companies mature and require a more advanced solution, the entry-level approach can be incrementally upgraded to combine all three capabilities within a single APM solution. This allows the APM efforts to start as a top-down business-driven effort, and as needed complemented by a bottom-up perspective to enable customers to determine the right disposition of all applications in the portfolio.
Years ago, IT budgets allowed for development of newer, innovative applications for running the business more creatively and efficiently. But as those budgets have been slashed to focus solely on application maintenance, IT managers must find ways to invest in new, innovative solutions that fuel business growth and evolve current application assets into a flexible business architecture.
A well-conceived APM strategy can evolve toward enterprise architecture, managed through project portfolio techniques. For mature companies, the best approach recognizes APM as an essential element of “enterprise modernization” -- a gradual move to leverage and enhance time-tested business capabilities with modern hardware and software, without disruption to revenue, supply chain activity, or operational goals. Within this overall modernization framework, an APM strategy can be useful along a broad spectrum of efforts -- from basic application assessment, to business planning, to a fully fledged enterprise architecture where APM constitutes one of many workstreams that keep the architecture highly flexible.
A Detailed Process
While an effective APM solution can vary from company to company, IBM believes the ideal process is powered by a decision and workflow engine, a customizable information model, and is fueled by analytics. APM focuses on incrementally collecting the information and doing the analysis required for effective decision making. You therefore need to collect metrics and attributes related to each of your business applications, and make that information available at time of decision making. This information may include application details; financial performance data; service and technical quality metrics; degree of enterprise architecture alignment; human resource data regarding worker skills and longevity; and overall business alignment, typically sourced via survey.
As businesses go through the APM process, they discover how to best evolve current assets -- from retiring them, consolidating them, or re-vitalizing them. In reality, you may apply more than one modernization strategy to an application. For example, you may find that most of the incidents interrupting smooth operations are related to some critical business logic, and therefore decide to re-architect those modules using a more modern language. Other modules may be rehosted on a newer platform (such as IBM System z) to reduce operational cost.
Suboptimal applications and application performance is a problem that grew slowly over the past twenty to thirty years. While there is no quick fix, doing nothing diminishes the role of IT from “competitive business weapon” to “drag on the business.” When IBM made APM a key asset-based practice, our IT organization managed to cut maintenance costs by 20% and defects by 58%.  We believe a similar approach might enhance your own IT strategy.