Application Portfolio Management 101 – Method and Benefits

Introduction

As organizations grow, IT departments and even individual employees buy applications to solve urgent problems without giving any (or at least enough) thought to the implications. As a result, applications accumulate and become difficult or impossible to integrate with existing applications or other infrastructure. Similar apps for completing the same tasks are purchased multiple times. Others fall out of favor and are no longer used but are still paid for – never being uninstalled.  Still others are bought and never used at all!  Then there are applications that are inconsistent with technological infrastructure, or its near future.  

Meanwhile, businesses are faced with the challenge of controlling costs and improving technology services for future business growth and profitability.  How does an organization provide clear insight into how many and what applications are in use, their business value, supporting technologies, risks and their place in a life cycle?

The Solution – Application Portfolio Management (APM)

APM is a framework for managing enterprise IT software applications and software-based services. APM provides managers with an inventory of the company's software applications and their associated metrics to examine the overall benefits of each application relative to each application’s cost and business / technology fit and risk.  It is in this context that analogous project portfolio management fundamentals are applied to the optimization enterprise technology assets – see Axelerate’s article “3 Stages of PPM Maturity.”  

APM is one of the single most effective IT management paradigms ever developed. When implemented properly, its impacts on the software inventory, its associated infrastructure and the cost of external organizations supporting it are astonishing.

APM builds on the capabilities of traditional application mining tools (also known as impact analysis tools) by establishing a periodically updated application inventory that is augmented with additional business attributes.  Quantifiable business information such as system owner and supporting labor costs can now be collected.  

APM makes use of numerous data types and system inputs – including: Demographics; Cost; Effort; Usage; Availability; Business Value Metrics, Technology Fit Metrics; and Application Risk Factors. This broader APM function allows management to quantify the application portfolio from perspectives other than a purely technical snapshot and enables business and IT management to decide the lifecycle dispositions of applications: Invest; Maintain; Replace; or Retire.

APM Characteristics

APM augments the infrastructure mining tools with the following capabilities:

• Enterprise-wide Focus – Impact analysis tools have historically been used against a single system, a subsystem or a set of programs.  These views supported project-oriented analyses and were therefore limited in scope.  APM takes an enterprise-wide view of applications, and in so doing creates a source of truth about application artifacts and the relationships between them, enabling analysis across the enterprise.
• Continuous Update – Establishing an application source of truth is pointless unless the source of truth is constantly synchronized with changes to production.  When application updates are recorded reliably by linking to source code management systems, the result is an always accurate, and therefore reliable, source of truth about applications.
• Open Repository – Many impact analysis vendors used proprietary database technology, precluding customers from adding information to it. This is one of the chief benefits of APM — adding simple business-oriented information to an established source of application truth to enable better business decisions about IT spending on applications.

APM Process

Conceptually speaking, the APM process is simple.  An application inventory is taken, it is evaluated and a disposition for each application is determined.  The steps are as follows:

1. Take a full application inventory:  This means identifying the application, logging technical and administrative data
2. Score the applications:  Collect business, technical and categorized costs associated with the application.  The collection of certain attributes will be automated; others manually; yet others will be subjectively scored metrics assessed by the technical and business managers.
3. Evaluate the applications: The applications are scored in terms of aggregate collections: Business Value, Technical Value, risk and cost.  The costs are used to compute the primary cost attribute – Application Cost of Operation (ACO).  ACO is decoupled from acquisition costs and is therefore a more appropriate cost metric for determining an applications disposition (the implication here is that we’re interested in what it costs to maintain the application in the future – not obscured by ‘sunk-cost’ purchase information).    
4. Determine application dispositions:  The applications are arranged into their lifecycle disposition ‘buckets’ per figure 1.
5. Plan future application investments:  From the application status determinations in figure 1, future investment strategy (with regard to applications and infrastructure) can be formulated.

APM Metrics

Once the applications have been inventoried and scored, the evaluation must be completed. While there are numerous metrics and approaches to assessing the application portfolio – we provide a typical one here.  TV, BV, ACO and Risk described below offer a reasonable cross-section of attribute groupings.  These cross-sections are described as follows:

• Technical Value (TV): Relates to whether an application meets target technology objectives as determined in enterprise IT architecture standards.  TV generally decreases as application age reaches the life cycle end.  An increase in value may be found with new technical breakthroughs.  
• Business Value (BV): Establishes how well the application supports the business requirements for which it is responsible.  BV decreases as gaps are identified in the application support for business functions.  It can also increase if there are opportunities identified for utilization across other business areas, to eliminate applications that perform same functions, or linkage to increased sales or customer retention.  
• Application Cost of Operation (ACO): Measure of an application’s cost of operation for a year. The cost is defined for non-acquisition, repeating application items, such as maintenance, break/fix or enhancement effort over the course of a year.    
• Risk: An estimation of the probability that the application will not satisfy its TV, BV, and ACO estimations.  Application risk here is not confined to security and may well be distributed within the TV and BV measures.  

Ultimately, these 4 quantities will enable an enterprise to understand the condition of an application in terms of its technological and business status, its recurring operation and support costs as well as a measure of risk to indicate the expected quality of those estimations.  

This set of information is reviewed against set criteria to determine an application's disposition.  Once a disposition is assigned, more specific actions can be employed for each application.  Figure 1 plots BV vs TV in a typical bubble diagram and displays a basic 4 quadrant graphic with these areas identified.  The relative size of the applications circle is proportional to ACO and its color (green, yellow, red) indicates the application’s risk score.

Invest: Warrants additional investments required to move towards an 'Maintain' status.

Maintain: Aligned with functional, technical and architectural standards.

Replace: Investment in prioritized functionality and operational environment. Not targeted to move toward 'maintained' status.

Retire: Targeted for retirement.

Benefits of APM

Application Alignment – Identify which apps support which business capabilities and eliminate those that don’t fulfill a valuable business purpose.

Optimize Software Inventory – Consolidating vendors, getting rid of duplicate applications and ensuring you have the right number of licenses streamlines application effort and costs.

Consolidate Infrastructure – A smaller number of more valuable applications allows you to reduce or reallocate infrastructure resources accordingly.

Minimize Technical Risks – Legacy apps and unsupported licenses can be security and compliance time-bombs. Application Portfolio Management helps to remove them.

Reduce Costs – Eliminating shrink-wrapped apps, and unused licenses could help cut software costs by as much as 15% without affecting business performance.

Eliminate Functional Overlap – Many apps come with a lot of functionality that most organizations will never need or use. The ability to link apps to business capabilities helps you identify which ones you can do without, so you can get rid of them.

Optimize Business Processes – Once you have rationalized your app portfolio, gaps and redundancies that slow down processes will become clear – then you can eliminate them.

Improve Outsourcing – By knowing which applications and services are critical, you can make better decisions about what to outsource before you sign any agreements.

Prioritize Projects – Active Application Portfolio Management helps you prioritize projects and associated applications according to the portfolio investment techniques based on business value they deliver, and eliminate those that aren’t pulling their weight.

Conclusion

APM is one of the single most IT management paradigms ever developed. When implemented properly, its impacts on the software inventory, its associated infrastructure and the cost of external organizations supporting all of it can be astonishing.

I’ll leave you with three examples of APM in action:

APM leads to a cost reduction of $2 million in a single enterprise (Infosys).

At least 10% of IT project costs can be avoided through application rationalization (Oracle).

A considerable amount of infrastructure was eliminated during this APM initiative (Planview).