APM Best Practices

An interesting discussion has kicked off on LinkedIn’s “Application Portfolio Management” Group discussion board. The discussion looks at best practices for running an APM initiative. Because APM directly affects strategic decisions around where IT resources should be applied, the topic is an important one.

There are a few interesting branches to pursue that look at APM best practices. Over the next few weeks, I’ll explore each in some detail:

• What’s the aim: You are conducting APM for a reason. It is to make better decisions about the application portfolio and where to focus effort. In this post, I’ll look at the Goal/Question/Metric paradigm and its relevance for APM.

• What’s collected: You’ve decided on the goals for your portfolio management and the questions that you want answered. But what are the data points that will get you there? It can’t simply be a matter of collecting technical metrics if you’re looking to answer business questions. “Lines of code” does not necessarily translate into cost. If it is a business goal of reducing cost or improving resource allocation, then metrics like cost per change or flexibility of a given business process should be collected and presented to a user in a business context. If it is an architect’s goal of reducing instability and performance issues, then metrics like “dependency levels” should be filtered by architectural concept, etc.
  

• Who decides: Different users in the organization will want to answer questions in different ways. How should you approach filtering of data by role or horizontal function to make sure that they have the data needed to make smarter decisions?

• What’s next: Decisions are made to be acted on. After all, if no change results you’ll just measure the same issues next time. How can you tie reporting to functional activities that change your portfolio for the better? And how can APM support ongoing monitoring of the activity and portfolio to see how well you’re doing and where to improve.

New Release of Modernization Workbench

A new release of the Modernization Workbench has just been launched. The new version 3.1 is "3 in 1". It combines best-in-class functionality from Micro Focus' analysis and application portfolio management capabilities into a single platform.

• Single Platform for Managers: Modernization Workbench provides the only business-centric solution for measuring and managing the application portfolio. Its integrated Enterprise View module delivers browser-based dashboards that help prioritise development projects via metrics like application cost, complexity, value, and risk.
• Single Platform for Assessments: Modernization Workbench provides the richest technical information about your applications. Powerful queries, visualizations, and specialized assessment tools, including for platform migrations, are available.
• Expanded Language Coverage: Modernization Workbench provides deep and now even broader language coverage. This new release expands its coverage of Java, JEE, additional job schedulers, and more.
• Mass Change Activities: Modernization Workbench dramatically accelerates the execution of projects that require numerous changes to application code -for instance, to adhere to regulatory requirements like ICD-10.
• Major Relational Database Support: Modernization Workbench 3.1 adds support for Microsoft SQL Server, ensuring that all three major relational databases are supported (adding to existing support for Oracle RDBMS and IBM DB2).

For more information, please see this release here:

Webinar: Application Understanding

You already know that applications automate your core operations. But do you have adequate control over these systems? The answer is often 'no' because of the sheer size and complexity of your application portfolio.

Find out how you can regain control over the applications that run your business at a webinar on May 27. Register here:

http://www.microfocus.com/promotions/wwwcwwmw0509/default.aspx?page=email

Planning for Application Modernization

eWeek just published a strong piece on how to plan Application Modernization initiatives. In the first paragraph industry expert Tim Pacileo identifies the key for any modernization activity: justify it. CIOs face a multitude of competing interests when determining where to allocate resources. The pressures are more intense now that IT has become so tightly interwoven with very visible business processes. As a result, it is critical that the CIO has the right information available to determine which priorities should be selected and why.

This is the role of Application Portfolio Management. It provides a framework to collect measurements and place them into business context. This allows CIOs to quickly determine which projects make sense to act on based on the strategy of overall organization and not just on the narrower needs of IT. Justifying projects is vastly simplified when the interests of IT are tied to the business.

Interestingly, Pacileo suggests using a chargeback model to ensure that costs are properly associated with the correct IT activity. The best approach for this method is via business-centric portfolio management. By placing application portfolios into business context (by operational unit, geography, etc) CIOs can match costs to the exact software (and IT infrastructure) that is incurring the costs.

Pacielo also points out that in the rush to solve problems, IT often will product multiple overlapping applications that drain resources. A business-centric Application Portfolio Management solution is again the ideal approach. When IT compels itself to match business-value, cost, and risk to IT assets these kinds of glaring issues become readily apparent. He cites the example of one company having three overlapping systems. We've seen cases with more than 2 dozen duplicate systems being maintained separately.

Application Portfolio Management is an increasingly rigorous discipline. Executives should turn to this framework when determining where priorities lie and how to justify them.

Portfolio Management is Top Priority for 2009

This week Baseline published its top IT trends for 2009. Among the usual contenders for top spot (such as virtualization and Software as a Service) was Project and Portfolio Management. The reason cited for its selection was the need to oversee IT activities and allocate resources toward high-priority tasks – which stands to reason in these challenging economic times.

The linkage between Project Portfolio Management (PPM) and Application Portfolio Management is an interesting one. PPM focuses on helping development managers to monitor and control in-flight development projects. Application Portfolio Management concentrates on the allocation of development and related infrastructure resources toward business priorities.

As discussed previously, Application Portfolio Management is at its most effective when measurements are placed into their appropriate business context. This could include managing applications by the business process they automate or by the geography that manages the applications, or other contexts or combinations of contexts. So, we could uncover complexities within a high-value business system managed in India, and decide to allocate resources to correct the issues.

We can think of PPM’s project entities as being another grouping through which we want to manage APM. This isn’t to say that APM should replicate PPM functionality, but rather that the technologies can effectively collaborate by sharing metadata. Much the same way that APM doesn’t replicate BPM functionality simply because it reuses business process models as a way to structure metrics reports.

Let’s look at how this can be deployed. A CIO of a bank is under pressure to improve responsiveness. Using APM, he looks at change request backlogs as organized by business process and spots an outlier in the customer management business process. He decides to investigate further. He looks at metrics for this process and sees that complexity levels are especially troublesome in the portion that is managed in Brazil. He decides to reallocate resources to attack complexity levels and whittle down the change request backlog.

This is clearly when PPM would become involved. Managers can use PPM to monitor deliverables, outcomes, and other elements that are part of a development project. But in parallel, APM continues to play a critical role. We could now overlay a new grouping onto our software; in this case the grouping would be by project. So, we are now tagging the portion of our application portfolio that is encompassed by the newly introduced project. Then, as we proceed we can collect metrics associated with the grouping of our project. This means that we can use APM to monitor the trend of value, cost, and risk of the software that is being enhanced by our project.

We can even start to combine business contexts, for instance by collecting metrics from software within Project A that is managed by Team A versus software managed by Team B. We start to see within APM very interesting metrics about adherence to SLAs as a result.

Best Practices for Application Portfolio Management: Gartner

An excellent piece of research by Jim Duggan came out from Gartner today. It details how companies should approach Application Portfolio Management – and of course, why you should be interested in APM in the first place. The why is clear: as the economy has slowed, companies must uncover and replicate efficiencies while slashing wasteful spending. The oft-quoted figure of 80% of IT budgets being dedicated to ‘lights-on’ activities is a primary reason why Application Portfolio Management has become such a hot topic.

But how should you discover where to focus rationalization and follow-on modernization activities? The paper relates a number of suggestions. The major thrust is that management should assess which portions of the application portfolio to rationalize based on different perspectives. That is, you should determine the ways in which you manage your business, and then rank your applications based on these views.

For instance, we could start with the most obvious perspective: cost. What are your most expensive applications and do you need to maintain these systems? Do they overlap and can be consolidated? Are they not used by the business? Is there a less expensive architecture?

You can then quickly move to other Application Portfolio Management perspectives. It could be by organization. Are applications that are managed by expensive / low-value providers that could be re-assigned? Or the perspective could be by business process. Are there business processes that could be better managed by an external service provider? Are there applications that support defunct business processes? You can see that executing APM from different perspectives allows you to rationalize based on KPIs that matter to your organization.

There is also a significant side advantage that comes from managing by these perspectives as you focus APM and continue to refresh APM. Once perspectives are in place – and rationalization decisions may have been made – you can focus modernization activities on sub-sets of the portfolio that matter to you. High-cost and low-business value areas? High-risk and frequently-changing applications? Now, resources can be applied to the right area. You may decide to deploy richer code analytics at this stage to get a complete picture about how developers should be concentrated.

Further, these perspectives –especially once placed onto the software – offer a ‘filtration mechanism’ for metrics as you collect them for APM. Looking for cost data on a business process, or risk information about applications managed by a particular organization? The perspectives provide the means to get these business answers. The result is highly business-centered development decisions.

Application Portfolio Management Resource Site

This new site exclusively is dedicated to best-practices for Application Portfolio Management is now online. The site explores what kinds of metrics that you should collect, how they should be combined into measures that have business utility, and what are the key attributes of an APM tool.

Application Portfolio Management is a methodology for identifying and prioritizing development activities. The aim is to locate misalignments between the application portfolio and business requirements, and then allowing managers and other IT professionals to adjust their resources accordingly.

APM is fundamentally about collecting business and technical metrics, combining them into useful KPIs in the right business context, and presenting them to decision-makers. On the site you'll find a wealth of relevant resources.