Proficiency Scoring in a Contact Center:
An Alternative Method of AHT Analytics

By Marshall G. Lee, TTEC

Do you ever have an issue calculating the actual impact of specific agents?  Do your agents feel pressure to get callers off the phone to move on to the next caller?  Is workforce tasked with staffing to a set average handle time (AHT) regardless of an agent’s natural pace?

The role of a contact center agent is arguably one of the highest stress positions in the service industry.  The pressures we expect the modern agent to tolerate may include daily monitoring of customer interactions, rigid schedule adherence, evaluation on multiple other scorecard metrics, and an ever expanding role.  All of this while balancing between customer service, sales, and issue resolution. Between schedule adherence and quality scores, agents must perform at an expected level and be where they are needed when they are needed. Meanwhile the workforce management staff must account for and plan on what traffic agents can and should be able to handle, then offset the impact of shrinkage. 

One shared issue that agents and workforce have in common in many contact centers is an average handle-time (AHT) goal. Agents must hit a prescribed maximum time and workforce management analyzes reasoning for the base AHT being met or not. Workforce then evaluates what that AHT should be. AHT is measured to ensure agent efficiency, efficacy of a staffing model, and to provide a consistent customer experience, through average speed of answer (ASA) and time spent on the call.

An AHT metric or KPI (Key Performance Indicator) can create a conflict between quality and efficiency, while causing added pressure for the agent.  As we all know, not all agents function at the same pace, which can make staffing and forecasting a challenge. This is especially true when trying to account for the actual impact of one particular agent. A slower agent is also not always a bad one. This makes the role of a workforce professional more challenging. 

There is an alternative that can reduce the pressure on agents, and make WFM staff more effective. Proficiency scoring is a method for calculating and ranking agent efficiency without having a specific concrete AHT threshold each agent must meet or beat. In lieu of a static AHT ceiling, a target AHT range with an upper and lower limit are established. Proficiency is a simple multiplier ratio that allows workforce and operations to gauge employee efficiency against the entire floor, and the floor to a team goal.  

A proficiency score works on a base 10 scale with 1.0 as the mean.  For example if the average handle time of the entire contact center is 10 minutes then 10 minutes is 1.0.  This means that if an agent has a 7 minute AHT they are a 0.7 proficiency, and an agent with a 13 minute AHT is a 1.3 proficiency.  So the 2 agents combined have a 10 minute AHT.   When the base 1.0 is set for the floor and each agent’s proficiency score is calculated then the score is ranged and employee impact quantified. 

While a simple scale, this is a powerful tool.  When agent proficiency is known, you then identify a bell curve and an actual individual agent impact.  Proficiency scoring used in place of AHT as a performance metric allows a range above and below an acceptable mean. This range becomes the threshold by which agents are scored versus a predetermined AHT. In our contact center, we use proficiency as an indicator of agent performance, with the actual KPIs that agents are accountable for being hold time and after call work.  If proficiency is beyond the ideal range, they are evaluated to see if there is an issue and what
actions need to be taken.

Proficiency demonstrates actual single agent impact and becomes simple to quantify.  Agent A is a 0.6 so they handle 1.67 calls to the team average.  Agent B is a 2.0 so they handle .5 calls to the team average.  Losing Agent A is like losing just over 1.5 FTE.  The analytics become easy to demonstrate, and FTE utilization and staff loss impact actual. On most ACD platforms, this can be worked out on the agent and skill/queue levels.

Ideally an agent should fall within a 30% range or 3 standard deviations of the mean AHT. This allows for balanced staffing and customer focus.  Agents who fall in the 0.7-1.3 range are considered to be good performers, 0.6 and 1.4 borderline performers, and outliers at the 0.5 range and lower or 1.5 and higher become agents to focus on. This simplifies the coaching of a 1.5+ agent.  What is causing the slower pace can then be identified and compared to the team. With this data and comparison, an explanation the impact of the higher proficiency score becomes obvious to the agent. In short, you can easily demonstrate that they take 1.5 calls to the team’s 1 call.

For workforce management, proficiency can do wonders in simplifying what can be intense analytics and application of agent AHT.  Workforce can more readily quantify the loss of a particular staff member. Scheduling becomes more effective—you can determine what agents to send to break or lunch or staff at the same time; a 1.3 and a 0.7 being off phone at the same time has less impact on consistency than two 0.7 agents at the same time or two 1.3 agents at the same time. When simulated at the agent and skill/queue level in your workforce staffing model, proficiency becomes a key component in forecasting service level. Imagine being able to tell an operations manager the exact impact of a particular agent being absent or leaving, then compensating for the loss based on actual agent efficacy in the staffing model.

When it is time to hire or replace termed agents, proficiency comes into play again.  When trending or simulating for new hires, you can apply a proficiency score to simulate the effect of new hires at different time slots. We do this by applying a 1.6 score to simulate the second full week on the floor.  In our experience,
6 weeks is generally what it takes for a new hire to come into the floor target range and 9 weeks to settle to their personal optimum proficiency. To simulate the floor at the new capacity, simulator profiles are given a 1.0 to simulate the 9 week mark.

The benefits of this method do not stop at scheduling and staffing. From a training, coaching, or process perspective, it becomes another analytics tool. If an agent is a 0.6,  is this due to going too fast on calls with not enough customer focus, are they so advanced they should be training or mentored into a new role, or have they discovered a process that is more efficient?  If an agent is a 1.5, is this because they are spending too much time on hold and ACW, too much talk time, or is this an agent with higher quality and more first call resolution? Matching proficiency against other performance metrics allows you to correlate desirable behaviors with a targeted AHT range.

One downside to this process is that without an anchor, the mean AHT may continually rise or fall. To keep this in check, a goal 1.0 proficiency mean can be set for a team or the entire contact center.  When we began using proficiency, we took a long-range trended AHT and compared that to our staffing budget and requirements.  The ideal mean AHT for staffing in each area became the goal 1.0 for the contact centers supported through the workforce team.  When a team’s mean AHT deviates more than 6% from their goal, Operations begins to look for causes on a global level and then looks for agent outliers.

Having been in contact centers that are AHT-driven and those that are not, the difference is palpable.  Contact centers are like factories; only instead of widgets, contact centers make calls. In the AHT-focused production environment, the agent focus is on number of calls produced. In an environment with no AHT focus, the production line may lag at the expense of efficiency and service level. In the proficiency environment, you balance the two. The line keeps moving at a steady pace; however each contact is addressed at the natural pace of the agent. They can resolve the issue in one contact using the strengths and talents they possess at the pace they naturally meet. This is how staffing an assembly line varies from staffing a call center floor.

As for my experience, when we began tracking proficiency at the agent and skill level level, occupancy went down, morale went up, absences down, and quality scores and first call resolution both tracked up as well. The impact on staffing and scheduling was that we could accurately simulate the impact of individual staff members and compensate accordingly. In the contact center, AHT was rarely if ever discussed — only when there was a real proficiency issue.  There were no reader boards with calls waiting posted: the current caller was the primary focus. We coached on hold and ACW, while educating on proficiency impact. With proficiency scoring, we were able to evaluate AHT without making it a primary focus.

At the end of the day, proficiency does not replace AHT,
in part because it is another way of looking at the same metric. This is another tool and method to view and measure agent impact and efficiency.  If nothing else, it can allow your agents to perform to their individual strengths while allowing leadership to coach to a reasonable mean, and workforce to forecast the actual “Power of One.”

SWPP Board Member Marshall G. Lee is Contact Command Center Workforce at TTEC.