A helpdesk can be helpful only if it has enough people to serve customers in a timely manner. Otherwise, people will look elsewhere for their support. Sizing a helpdesk staff is very difficult because it changes from situation to situation. Universities often have thousands of students per helpdesk attendant. Corporate helpdesks sometimes have a higher ratio or sometimes a lower ratio. In a commercial computer science research environment, the ratio is often 40:1, and the first-tier SAs have a similar skill level as second-tier SAs at other helpdesks, to meet the more highly technical level of questions. E-commerce sites usually have a separate helpdesk for internal questions and a “customer-facing” helpdesk to help resolve issues reported by paying customers. Depending on the services being offered, the ratio can be 10,000:1 or 1,000,000:1.
Ratios are a no-win situation. Management will always push to have a higher ratio; customers will always demand a lower ratio. You can always increase the ratio by providing less service to the customers, which usually costs the organization more because the customers spend time doing their own SA work inefficiently.
Rather than focus on customer-to-attendant ratios, it is better to focus on call-volume ratios and time-to-call completion. For example, you can monitor the rate at which customers receive busy signals or how long they wait to receive a response to their email or the number of minutes issues take to be resolved—minus, of course, time spent in “customer wait,” as described in Section 16.2.6.
These metrics focus on issues that are more important to the customer. Customer-to-attendant ratios are an indirect measurement of benefit to the customers. In metric-based management, direct metrics are better.
Managing resources based on call volume also presents a more diverse set of potential solutions. Instead of one solution—headcount management— companies can invest in processes that let customers help themselves without need for human intervention. For example, new automation can be created that empowers customers to do tasks that previously required privileged access, online documentation can be provided, new services can be provisioned automatically via web interfaces, and so on.
You need to have appropriate metrics to make decisions about improving processes. Metrics can reveal good candidates for new automation, documentation, or training for both SAs and customers. Metrics can reveal which processes are more effective, which are used heavily, or which are not used at all.