Enterprise Software

The key to cost-effective support for specialized applications

Providing your user base with adequate support for highly specialized or home-grown systems presents some challenges. Here are the five components of a solid knowledge transfer plan.

By Chris Ciccolini and Barry Sorkin, SEI Information Technology

As many CIOs know, packaged and proprietary applications are very different animals when it comes to user support—especially in terms of the IT budget. Support for widespread office applications like Microsoft Word and Excel lends itself to standardization, and expertise in these areas is available in the labor pool.

On the other hand, providing your user base with adequate support for highly specialized or homegrown systems presents some challenges, both in terms of cost and quality:
  • Power-user-level expertise for these applications is often held by limited numbers of expensive IT resources.
  • These experts are often in demand for more forward-looking development or implementation projects.
  • Experts in specialized applications are difficult (if not impossible) to find in the labor pool.

The value of knowledge transfer
Despite these challenges, the standard three-tier support model, which is typically used for basic desktop applications, can be employed effectively for sophisticated proprietary applications. The rationale behind this approach is to allocate as many support responsibilities as possible to your less-expensive resources (Level 1 technicians). This frees your senior resources to handle more challenging support issues or to work on strategic IT projects.

The key to a successful implementation that delivers good customer service and cost-efficient support is effective knowledge transfer. When establishing support desks for proprietary applications, many organizations operate in a "document hand-off" mode—technical experts (Level 2 and 3 technicians) hand off some system documentation and let the call center take it from there. As organizations soon learn, this results in a Level 1 desk that is unable to handle many user issues and in Level 2 and 3 technicians who are resolving most issues anyway, but with a delay in service.

To realize the value of the three-tier support model, you must have a constant flow of information from the Level 2 and 3 teams back to the Level 1 call center (see Figure A). This facilitates the kind of service that results in high customer satisfaction by equipping the Level 1 desk to:
  • Resolve a significant number of calls on the first contact.
  • Correctly prioritize the call and escalate it to the appropriate team with sufficient detail to enable the technician to hit the ground running.

Figure A

An effective knowledge transfer plan
A knowledge transfer program that will facilitate this constant flow of information has five key components.

1. Analysis
Identifying the types of calls that the help desk is likely to receive is a critical first step. Developing detailed troubleshooting scripts and conducting training sessions come next, and both require significant effort. These activities should empower the call center to resolve those issues that will have the most impact. In other words, focus training and scripting efforts on the most frequently asked questions and common problems, as well as those issues that are the most time-sensitive.

The net result of this stage should be a list of call types for which the call center will be given troubleshooting scripts and a projection of how the new procedures will affect the first contact resolution rates. Analyze the information and use it to set measurable goals.

2. Collaboration
Once you have identified the support topics that will give the most "bang for the buck," the next step is to gather all teams involved to collaborate on the scripts, training plans, and infrastructure plans. In larger companies, multiple teams may support one application—one team supports the code itself, another supports the database, another supports the servers on which the application resides, and still another supports the WAN through which users access the application. If users call the help desk indicating that they can't log in to the application, the problem may be with any of the above components. Therefore, when writing the troubleshooting scripts, representatives from teams that support all the relevant areas should be involved.

3. Implementation planning
Once the scripts are produced and the scope of the call center's responsibility is defined, devise a plan for transferring support knowledge and responsibility to the call center. This plan should address items such as training staff, equipping the call center with necessary support tools (administrative tools and privileges, reference information, bench systems, etc.), and incorporating documentation into the call center knowledge base. The implementation can begin once these details have been laid out.

4. Evaluation
The last stage of any project should be an evaluation of the project's success. In the Analysis phase, goals were set and projections were made concerning the effect that the newly transferred knowledge would have on the desk's first call resolution rate. Were those objectives met? What remedial action should be taken if performance falls short of projected goals? Is the call center resolving the issues that were targeted in the analysis? Perhaps the scripts need to be refined, or additional retraining is necessary. Using the Analysis information, identify areas for improvement by running metrics reports and auditing support calls.

5. Maintenance
To say that things change in an IT environment is a gross understatement. And while everyone in IT recognizes the constant change, precious few acknowledge the effect that those changes have on a Level 1 call center. IT-released management processes need to address Level 1 scripts, ensuring that they're updated to reflect all production changes and that the call center is notified well in advance of the migration of changes into production.

Clearly, the knowledge transfer process can be a fairly large effort, particularly in environments with numerous specialized applications. However, building a database of repeatable processes that can be performed by lower-level resources not only ensures consistent, quality support, but also maximizes the value of more expensive, highly technical resources and contributes to a cost-effective support environment.


Editor's Picks