Skip to content

How to Design an IT Services Portfolio with Metrics

September 29, 2008

As IT organizations attempt to become more service-oriented and involve process improvement initiatives, questions arise regarding what to measure and how to measure it. We provide a decision framework for designing effective service and process metrics. Performance is the failure to differentiate the measures to share with business customers and the measures that should be used only internally. Getting this distinction right is an essential underpinning to IT organizational credibility.

The first step in building a measurement program is to acknowledge that IT service and process  metrics are fundamentally dependent on the services offered and how they’re bundled. Before an IT organization can identify an appropriate set of metrics, it must articulate a service portfolio. The success or failure of an IT organization achieving service objectives is based on processes. Therefore, IT organizations must explicitly understand which processes are engaged in fulfilling which services. This entails creating a service- o-process map

Email

Telephone

Account Provision

Storage

New User

X

1

Backup

X

X

2

Voice Call

X

X

2

Video Conf

X

X

X

3

TOTAL

2

2

3

2

In general, the average IT organization with application and infrastructure responsibilities might have 15 “true” services that are exposed to the business. Each of these services is specifically designed to deliver some business benefit to the service consumers at a price and quality level acceptable to them. These dimensions of benefit, price and quality are captured in service-level agreements (SLAs), which represent absolute performance requirements and service improvement objectives. Services are executed and service-level targets are achieved through processes.
Unfortunately, a one-to-one relationship between services and their fulfillment processes doesn’t exist. A single service may depend on more than one process. (For example, an IT organization with a full complement of services may execute as many as 400 processes.) Accordingly, each process may affect multiple services. Processes are, in turn, comprised of many steps, business rules and organizational hand-offs, which drive individual process efficiency and effectiveness.
Although every IT organization may offer essentially the same services, the way those services are bundled, positioned, perceived and executed is unique to every business
The, standardized service and process metrics essentially don’t exist, and IT organizations must design their own. The service portfolio and service-to process map provide the foundation for doing so.

From blog

Customer needs and expectations determine service requirements. Those expectations are captured in SLAs. Each service will have a few, critical SLAs. Whether those SLAs are met is based on processes and the outcomes they generate. In this example, the service depends on two processes, each of which has an outcome. The combined results of those two processes drive service results. Process outcomes are determined by the effectiveness of the individual steps or activities that are executed, any one of which could cause necessary process outcomes to be achieved or not.

Layer Description
1) Service-Level Performance All IT metrics, therefore, should be aligned to the services IT organizations offer and the business contribution those services make. The first measurement tier is the SLA. Each service in the portfolio should have at least one and as many as three distinct SLAs in place that have been negotiated with the customer and are representative of the value, price and quality of that service. It is critical that these measures be customer-centric and not IT centric. Generally, if the IT organization has developed a true value-based service portfolio, then the SLAs will become self-evident. If no service portfolio exists, then the SLAs will usually be inappropriately focused on process-oriented metrics, rather than service outcomes.
2) Process Outcome Metric Once the service-to-process map is developed, the IT organization will know what processes drive a particular service outcome. If a service is not meeting its SLAs, then the next step is to examine the outcomes associated with each process involved in the service to pinpoint the “fault.” Every process involved in service delivery must have its outcomes measured. This is the second measurement tier.  It may also be desirable to implement real-time measurement and monitoring against key processes to enable proactive process improvement before service levels dip. The advantage of the service-to-process map is that it will not only tell you what processes are involved in a particular service, but it will also tell you which processes are “core” (those that underpin most or all the services in the portfolio), and which services an individual process supports (preventing you from optimizing a process for one service and breaking it for another). Core processes are good candidates for real-time monitoring.
3)  Process Analytics Once a faulty process has been identified, it must be examined in more detail to discover what is broken. This root-cause analysis capability requires instrumenting the process so that productivity, hand-offs and other issues can be identified and resolved. This is the point of continuous improvement. Once the break, bottleneck or inefficiency is identified and resolved, that instrumentation is no longer required and the IT organization can move on to the next improvement opportunity. These metrics are also for internal IT use only and should not be shared with service customers unless redress affects them in some way.

This three-layer approach to measurement has several implications:
• Metrics that are appropriate for customers are focused on business benefits. Metrics that are appropriate for monitoring and diagnosis are for IT organizations. The two are not interchangeable.
• Metrics and measurement systems are not static. As services, processes and capabilities change, so must SLAs and process outcome metrics.
• Process instrumentation is an ongoing requirement. It requires investing in permanent measurement, quality assurance and performance management capabilities.
• Measure only when necessary. Measurement for the sake of measurement or measurement that is unfocused represents a cost to the business with no commensurate benefit.

Advertisements
No comments yet

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: