In order to keep IT operational costs low, it’s important to get the most of your service desk support agents. No one wants full-time personnel sitting on their hands during slow periods or any instance in which the contact volume doesn’t justify the position. To ensure that, help desk managers regularly track the best measure of agent productivity: their utilization rate. While low utilization is an obvious concern, anything significantly beyond the industry average of 47.8% per a HDI study can be too much of a good thing. In fact, rates that hover above 75% for all agents not only imply that they are overtasked but could also indicate poor SLA performance such as high abandon rates and longer Average Speed to Answer (ASA).

KPIs may also reveal poor performance in the form of inordinately long average handle times. Often this is because Level 1 service desk agents are spending too much time troubleshooting versus escalating the issue to someone with the relevant access, training, or time required to resolve it. Even if the agent has the skills and access, if their average handle time is well beyond 10 minutes, queue management (i.e. agent availability for the next inbound contact) will be substandard. If, on the other hand, average handle time is within the industry standard 6.5 minutes yet SLAs still show high abandon rates, the correlation is often a sign of insufficient staffing levels. In this scenario, agents are getting off the calls quickly, but the high utilization is due to a backlog of contacts (phone, voicemail, email, web submits, etc.) because there aren’t enough agents logged in to manage the volume.

To study the mathematical correlation between average handle time and utilization rates, they are calculated as follows:

(Average number of contacts handled by an agent in a month) × (Average contact handle time in minutes) divided by (Average number of days worked in a month) × (Number of hours worked in a day) × (60 minutes/hour)

So if an agent handles 700 contacts a month at an industry-standard 6.5-minute average contact handle time the calculation would be:

700 x 6.5 = 6,500 ÷ 21 days x 7.5 hours x 60 minutes per hour = 48% utilization rate

In the above scenario, a Level 1 agent regularly troubleshoots complex and time-consuming issues through resolution without escalating it to Level 2 or 3. Even if SLAs such as Average Speed to Answer or high abandon rates don’t matter for an internal help desk, increased utilization may be borne out like this. Let’s say an agent handles 500 contacts a month, but at a much longer 15-minute average contact handle time the calculation would be as follows:

500 x 15 = 7,500 ÷ 21 days x 7.5 hours x 60 minutes per hour = 79% utilization rate

All other variables being equal, resolution rates will inevitably be higher for a 15 minute average handle time than for a 6.5 minute one, but 79% is well into the danger zone for agent turnover. Plus if these numbers are consistent across the board for all other agents at the service desk (whether due to complex contact types or poor agent training), end users can expect longer wait times to more than double as a direct consequence of those handle time ratios of 15 to 6.5. And it’s not difficult to conceive of how long wait times will increase the abandon rate.

The other obvious drawback to high service desk agent utilization rates on a prolonged basis is agent turnover from being pushed too hard which can lead to eventual burnout. Service desks that operate at high utilization may be maximizing ROI and productivity over the short term, but the long-term costs of high turnover (recruiting and training) will more than cancel out those temporary savings. Ideally, a well-run service desk will strike a balance in the numbers so that agent utilization does not come at the expense of missed SLAs, high turnover, and low end-user satisfaction. Since all of these variables are interdependent, keeping agent utilization at reasonable levels will ensure all of the other numbers add up in the way intended: low turnover, low costs, and high quality of service.