Focus on quality of service and cost containment leads IT service and support management to a practical question: "Is my staff the right size and structure?" The answer depends on what is appropriate for the situation at hand. Staffing and structure depends on a consistent set of IT environmental factors:

• Quantity and complexity of applications
• Devices
• IT support services
• Service levels
• Skill level of staff
• Level of infrastructure standardization
• Budget and maturity of problem and change processes

Taken together, these factors strengthen or weaken the staff's ability to get the job done. How the factors play out in an environment determines whether the IT staff will lean toward high or low staff count. Established typical support staffing ratios rarely match an enterprise's current state and should not be used when truly benchmarking production performance. What is appropriate depends on what the business wants, what it is willing to incur in costs and what it will pay to change things. In this research, we look at the complexity, shift in role responsibility, and support service demand trends that are affecting service and support staffing.

Complexity Factor

Technology adoption factors have a great impact on support staffing demand. The key is understanding how these environmental factors result in altering the demand for IT service and support. The complexity in support is clearly exhibited in a small subset of growing statistics regarding the average number of applications and hardware devices supported per employee. According to Gartner research, the typical help desk in 1995 supported 20 to 25 applications, whereas in 2001, this number ballooned to an average of more than 200 applications. In the financial services industry, this figure can exceed 500 applications for enterprises. A major contributing factor in this growth is the consolidation of the plethora of help desks within an enterprise to a single point of contact, which is also defined as a consolidated service desk (CSD). Within that consolidation exercise, applications were coalesced into more-concentrated Level 1 support.

Growth in hardware devices has increased aggressively in a five-year period. In 1995, PCs were tracked in relationship to employees. On average, there were still many nontechnology users within the enterprise without a desktop device. By 2001, enterprises were finding that PC counts exceeded employee numbers and that the enterprise hardware consumption had expanded beyond PCs to personal digital assistants, wireless devices, enterprise resource planning bar code readers and so on. Fundamentally, in the past 10 years, technology (hardware and software) has moved from a business perk to a business requirement from the executive suite to the mailroom clerk.

Other methods to increase technical complexity have sprung up in our applications. The number of features and function points has increased dramatically. Thus, as the size of an application (that is, the number of lines of code) grows, the fault level potential increases. Ongoing growth in features also expands the demand for end-user assistance toward "how to" problems. These are just two factors of many to illustrate the point that complexity has increased throughout the enterprise.

Service and Support Demand Factor

As consumption and use of technology increases in an enterprise, the request for service and support increases in parallel. Gartner research finds, correspondingly, that the average call volume for support to the service desk has increased from less than one call to 1.36 calls per employee per month from 1995 to 2001. The typical range of call volume is 1.1 to 1.6 per employee per month. The following shows Gartner’s 2002 CSD performance metrics:

• Average range of first-level CSD contact resolution: 54 percent to 77 percent of calls
• Average range of inbound contacts per analyst: 450 to 530 calls

Gartner statistics reflect production improvement in the area of service metrics, such as first-level contact resolution, from 1999 to 2003. Analyst-handling production of inbound contacts increased through the 1990s and has plateaued in recent years (see "Support Automation Improves CSD Productivity"). Based on the statistics outlined, an enterprise that has 10,000 employees can anticipate, on average, a monthly call rate range and then calculate an approximate range of the number of Level 1 staff required (see Table 1).

Table 1

IT Service Desk Baseline Staffing Ratio

Source: Gartner Research (January 2004)

These numbers represent a Level 1 baseline range; however, they will need to be further groomed by contact complexity, outbound contact activity (see Note 1) and service-level requirements. Growing complexity will create longer interaction times, which mean less contact time availability for Level 1 analysts. Higher-than-average outbound contacts will lower Level 1 analyst availability to inbound contacts. Increasing service-level agreement requirements (such as an increase in the first-contact resolution rate) results in increased time allocated to a given contact, which diminishes Level 1 analyst availability.

These basics statistics have been representative of the volume of the demand associated predominantly with problems. There is a growing volume of service requests that have historically been manual, which were under-measured in the 1990s. Service desk tools and the maturation of problem and change processes have begun to capture this volume of service demand, which can represent up to 10 percent of additional activity handled by service desk personnel. Growth in service request activity and management by the service desk will affect Level 1 staffing ratios and must be accounted for when analyzing complexity and service demand factors.

Automation and Change Roles Factor

From a role and responsibility perspective, IT service and support models in the mid-1990s were most often defined as Level 1, Level 2 and Level 3. Level 1 played a predominant role in addressing generalist problems focused on "how to" and "password reset" issues (see "Deploying Self-Service? Know Thy Problem Types"). Other problems outside the simple and consistent were dispatched to more-technically astute support staff within Level 2. As an example, use of desktop support and remote control tools was an established Level 2 responsibility well into the late 1990s.

As technologies have provided new capabilities to automate via self-service to self-healing, a new level was developed and defined as Level 0 (see "Service Desk Tiering Model Improves Self-Support Success"). Subsequently, delivery of support has changed in less than a decade. Automation tools have moved "how to" and "password reset" problem types from the generalist to Level 0. Problems outside the simple and consistent that were dispatched in the 1990s to Level 2 are now moving to Level 1. This has been facilitated by the trend to empower Level 1 analysts using remote control and patch management technologies to solve problems. The results expand the responsibilities for Level 1 to address new problem areas (for example, updating an application remotely on an end user's desktop because of previous configuration conflicts).

Automation of problem knowledge requires analyst resources in documenting knowledge. This translates to Level 1 and Level 2 support staff resources allocated toward problem analysis, root-cause identification and problem resolution documentation. This creates new roles and responsibilities within the support levels and may affect the staffing ratio. More automation of service and support reduces Level 1 and Level 2 direct analyst resources applied to end-user interaction and focuses resources toward Level 0 delivery. This will create a need to shift the staff ratio between the multiple levels. For some enterprises, it translates to a change in skills mix, where Level 1 generalist capabilities are evolved to Level 2 technical skills. Fundamentally, the greater the focus on knowledge documentation and subsequent automation of that knowledge, the greater the number of new staff roles within Level 1 and Level 2, and new positions in Level 0 where positions previously did not exist. This could result in overall staffing growth where understaffing existed. However, maturation will optimize production and could reduce overall staffing head count.

Key Issues We Discussed

What strategies and best practices promote effective IT service and support?

How will IT service management technologies and standards evolve?