Five Steps to Managing IT Service Performance

Setting the Stage
Over the last several years, IT organizations have been rapidly deploying new applications and building the infrastructure to support them, installing 3-5 servers a week in some cases. Staffing levels haven't kept pace. As a result, proper sizing of IT requirements to support business demands is disregarded in favor of over-provisioning; IT simply buys lots of capacity with hopes of avoiding performance issues.

While businesses have benefited from the quick actions of the IT and application teams, the long-term costs may be too high. IT organizations are now feeling the pains of the unchecked growth. Infrastructure management is complex and difficult. Space is limited, power and cooling costs are high and remote disaster recovery site costs have skyrocketed.

Escalating costs and declining service quality are impacting business operations and margins.

So now what?
Clearly, it's time to regroup. Survey the environment, identify problem areas and gaps, and find better ways to manage existing IT infrastructure. Optimize the planning and delivery of IT services.

These five steps of a process called IT Service Optimization are your guide:

1. Understand business objectives
2. Prioritize services and assess risk levels
3. Establish service levels
4. Plan and provision services
5. Manage service performance

Understand Business Objectives
The goal here is to enable IT to align with business priorities and determine where time is best spent. Priority should be given to business units that have the biggest impact on the company's success. Business units with high potential for substantial performance improvement are also good candidates.

IT organizations and business units must establish a dialog, and the onus is on IT to speak in business terms. These discussions should:

• Identify key business metrics
• Establish long-term goals
• Uncover contradictions among business unit requirements
• Determine cyclic events

Prioritize Services and Assess Risk Levels
Prioritizing IT services helps IT focus its attention and resources where it is most needed in order to generate business value. Business-critical services should receive more attention and planning to minimize risks and ensure consistent delivery of those services.

You can start by building an inventory of applications and services and identifying their relationships and dependencies. Application and service profiles aid in understanding and tracking associated SLAs, infrastructure used, location of those resources and key performance indicators for the service or application. These profiles allow you to compare business unit requirements with the current situation to uncover gaps.

Together with business unit managers, IT can categorize profiles into high, medium and low priority and record agreed upon priorities in the profiles themselves.

Establish Service Levels
The service catalog contains definitions for every service provided for the business by the IT organization. The catalog will identify each service, describe the functions provided and the terms of service (e.g., hours of operation, customer support options), including any service tiers, and the unit costs of the service. The service owner should be identified along with any pertinent support contact information and any required escalation time frames.

Some set of standards or measures needs to be put in place so the extent of each service's success can be measured and the value to the business quantified. The metrics should be attainable, meaningful and measurable.

Forrester analyst group found that 43 percent of North American IT shops didn't have a fully implemented sourcing strategy or process, which meant millions of dollars were being spent with little or no strategic alignment.

Plan and Provision Services
Applications must not only fulfill functional requirements, but they must also perform well and scale to meet future business requirements.

The provisioning process, introducing services and applications into the production environment, should include steps for determining the optimal configuration for systems that will host the new applications and take data center architectural policies into account.

When rolling out new applications, a load-testing tool to simulate transactions coming from end users or other systems can be used to benchmark applications on test systems to be certain they can support required service levels.

New services and applications require a process with more rigor to ensure a successful implementation. The provisioning process should have the following steps:

• Requirements definition
• Prototyping the new services or application
• Costing
• Management approval
• Service Catalog addition
• Provisioning
• Post Implementation Review

Manage Service Performance
Performance management, capacity planning and management reporting are three important aspects of managing service performance.

An effective performance management process finds application and service bottlenecks prior to their impacting business processes, buying time to fix them before they become problematic.

From a testing perspective, time and cost considerations usually make it prohibitive to conduct performance-related tests on the actual hardware using production-level workloads. However smaller, representative loads can be applied to a scaled-down set of test servers and software while performance analysis software takes a baseline reading of performance. Analytical modeling can then be used to rapidly predict how various configurations will perform under a scaled-up production-level workload, all without the need to purchase the actual configurations under consideration.

Capacity planning is important as it monitors infrastructure utilization progress and ensures that any corrective actions take place before capacity is needed and business processes disrupted by less than optimal service performance. A recent Forrester report states that correcting capacity and performance issues costs a lot less before deployment than after.

In order to judge IT's success in delivering services at agreed upon levels, it is necessary to monitor and report those results.

Management reporting must be put in place in order to help quantify results and measure the effectiveness of services. Automating the report process leaves more time for the capacity planning staff to address more productive tasks.

Performance issues can never be completely avoided, so it's ideal to perform continuous service review and improvement. Also business plans, forecasts and technologies change, requiring continuous analysis and adjustment.

Following these five steps will help IT accurately plan for business growth and support requirement goals with little risk. For more detailed information on the ITSO process, please access the white paper, IT Service Optimization: Service Inventory.