Here are some suggested areas to investigate if service levels are not meeting goals:

o Is there sufficient staff to meet overall demand? Lunch, sick time, vacation, or other absences need to be accounted for, or lower service levels will result.

o Is there enough staff available at the right times? Most support organizations find that their incident arrival varies over time during the day, week, and possibly the month. Use call arrival histories to determine when peak and low-volume times occur. Variations may be caused by business events rather than calendar event.

o How much wait time is too long? There’s no clear answer, but the above questions should help to identify some of the trade-offs. Wait times for phone support of one minute are common, but more or less may be more appropriate to a specific company. This is where measuring satisfaction can help drive the help desk’s business model. Also, if for staffing or other reasons longer wait times are a necessity, choosing a model of support that is less impacted by delays (for example, callbacks or electronic) may be more appropriate.

o Are staff members keeping to the schedule? Particularly with small support groups, being slightly late or leaving a shift slightly early can have significantly negative effects on service levels. Of course, staff who are not on their shifts will have a strongly negative impact on service levels. Especially in small groups, service backlog can take a great deal of time to clear, causing all service requests during that time to be delayed.

o Are there staff members working on other projects, preventing them from meeting their support delivery goals? All too commonly, support staff become involved in projects or other tasks that are quite valuable but impact help-desk delivery. Prioritization of support vs. projects is vital. If someone scheduled to provide support is not there, resources to replace them should be allocated, or service levels will very likely drop.

o Is your historical incident-arrival model accurate? It is important to use past support-request history as a guide for planning help-desk support. However, if you are not taking increasing demand, new product shipment, and seasonalities (such as budgeting periods) into account, service levels may flounder. It is also important to realize that past histories may be a good guide for estimating future demand, but this method is not infallible. An understanding of the issues surrounding changes is also vital.

o Do you have new, inexperienced staff? New, inexperienced staff typically require more time to resolve issues and thus resolve fewer issues during their shifts. They may also require the assistance of other help-desk staff.

o Are there new support burdens you need to address? New versions of products, new software, or changes in computers can all introduce additional demand, either short or long term.