Newsletters

Customer Support:   (972) 395-3225

Home

Articles, News, Announcements - click Main News Page
Previous Story       Next Story
    
The Four Operational Reasons Contact Centers Fail - Part 2

by Lawrence McKenzie, Partner, Q-Solutions Group, LLC - January 31, 2017

The 4 Operational Reasons Contact Centers Fail
 
Reason #2 - Agent Scheduling
 
This series of articles address the four primary operational reasons why contact centers fail: Process, Agent Scheduling, Agent Utilization, and Center Management.  The intent of this overview of the major operationally focused failure points is twofold.  First, the operational issues listed above are the fundamentals of the contact center business.  Unless centers consistently exhibit a mastery of these basic factors, excellence in performance will most certainly be unobtainable.   Secondly, there is a disturbing trend in contact centers to apply technology in an attempt to solve fundamental operational issues in lieu of directly engineering basic corrective actions.  By doing this, we generally mask the root problem and merely optimize elements of our operation that remain basically flawed.
 
In the last article, (found here), the topic of Process was discussed.  The effectiveness of contact center processes as realized through one call resolution, branding and core value reinforcement, and customer satisfaction were covered.  In addition, the improvement of key performance indices (KPI) such as Average Handle Time, Increased Sales, and Reduced Agent Cost were discussed.   However, to fully realize the benefits of process enhancement, implementing agent scheduling that accurately reflects the enhanced processes and performance is imperative.
 
Agent Scheduling
 
One of the key failure points in basic contact center operation is inaccurate agent scheduling. Agent scheduling systems vary from home grown or legacy systems to sophisticated scheduling packages.  The most appropriate scheduling system depends upon the center’s specific requirements, size, and budget.  However, to optimally schedule agents in a consistent and manageable manner, all systems need to properly incorporate the following variables:
 
  • Average Handle Time
  •  True Call Demand
  •   Service Level
  •   Manageability
  •   Agent Factors
Failure to properly account for these primary variables will result in poor service levels, misapplying resources, over predicting agent requirements, management issues, and poor agent morale.  The following sections will explore these significant variables in greater detail.
 
Average Handle Time (AHT)
 
For our purposes, AHT is the elapsed time from the agent answering the call until both the call and any associated After Call Work (ACW) are completed.  In our previous article, we reviewed the necessity of implementing an optimal call process in order to ensure that the call ultimately achieves your customers’ as well as your company goals while at the same time optimizing AHT.  To realize the best results from any scheduling system, this process implementation is required for each distinct queue.  After determining the AHT through objective observation, the scheduling system should use the observed AHT as a constant in the scheduling process.  It has been our consistent experience that when agents are being directly observed they attempt to perform at their most efficient, thereby providing a true AHT value.
 
Many scheduling systems utilize system reported (vs. observed) AHT for scheduling purposes.  This creates two major problems.  First, by using the reported AHT, we are accepting the current AHT performance as optimal.  In other words, the current performance of the agents on the calls will adjust the AHT, thus modifying the schedule and ultimately determining the number of required agents.  Secondly, without an objectively derived static AHT, management of call duration becomes arbitrary. 
 
True Call Demand
 
For scheduling systems to accurately match agent resources to call arrival it is imperative that True Call Demand is utilized in the scheduling process.  True Call Demand is basically determined by two factors;: customer call behavior and contact center performance.
 
Customer call behavior, simply put, is the call arrival patterns in discreet time increments.  The call arrival patterns need to be forecasted for growth/decline, seasonal fluctuations, day of week, daily time increments (half hour is suggested), and anomalous days such as holidays.  Generally speaking, unless calling patterns are driven by external factors such as blitz advertising, the call patterns are quite predictable.  Forecasting methodologies range from using statistically sophisticated methods to using running averages.  A concern with complex forecasting algorithms is that when the forecasted call patterns and actual call arrivals do not match, it is extremely difficult to determine the cause and solution given the nature of proprietary forecasting software.  Conversely, using simple average call arrivals to forecast demand has deficiencies as well.  For example, if you are standing with one foot on a block of ice and the other foot in a fire, the average would indicate you are quite comfortable!  The result of forecasting by average call arrivals guarantees that the forecast will be either high or low much of the time! 
 
Center performance can significantly affect the ability to accurately forecast true call demand.  For example, if during peak call arrival periods abandonment rates exceed 10 percent, a large portion of the customers will call back at later times.  This will dramatically effect forecasting in two ways – by counting a single caller twice, and by artificially pushing the call demand to time periods later in the day.  This performance based shift in call arrival patterns will ultimately affect the call forecast and agent allocation by scheduling more agents than required and scheduling them to be on the phone during the wrong time period. 
 
In our experience the best scheduling solution for forecasting and utilizing the true demand normalizes call arrival data and uses basic statistical factors that can be easily understood and manipulated by the user.  Adjusting for performance issues is as much art as it is science. However, when done properly the service level will become consistent within acceptable guidelines, the agent schedule will be accurately determined, and operating cost savings can be realized.
 
Service Level
 
Service level, or the speed at which calls are answered, is a key factor in scheduling and agent requirement. Determining the best service level for your center is a balancing act between call abandonment, agent cost, and market acceptability.  Market acceptability, unless working under an SLA, is an important customer satisfaction factor and will be discussed at length in a subsequent article.
 
To best illustrate the abandonment rate verses the agent cost factor let’s consider the following example.  A call center operating with 100 agents and at acceptable performance levels makes the decision to decrease their abandonment rate from their current 5% to 1% in an effort to enhance the customer experience.  This planned improvement would decrease the abandonment rate by 4% and decrease the average speed of answer by approximately 20 seconds.  Based on economies of scale studies, the associated cost to achieve this decrease in abandonment rate would be the need to hire an additional 38 agents or investing approximately $1,000,000 to answer an additional 4% of the calls.
 
Accordingly, most contact centers opt for a five percent abandonment rate.  Although it may be tempting to set your center’s service level to industry average, we suggest fully assessing the benefit cost prior to setting service level goals. 
 
Manageability and Agent Factors
 
The manageability of an agent scheduling system is vital to successful and efficient call center operation and is typically where agent scheduling fails.  If managers and floor supervisors don’t understand or properly manage to the agent schedules, then no matter how accurate the schedules are the power of the system will not be realized.  To avoid management failure the scheduling system should be understandable and simple to use, create schedules that are easy to manage, and be acceptable to the agent population.
 
Choosing a scheduling system that is accurate, cost effective, and simple to use is absolutely critical to assure effective management usage.  Your scheduling system must create schedules that:
 
  • Properly match agent resource to True Call Demand
  • Effectively achieve Service Level goals at the best operating cost
  • Are understandable and simple to manage
  • Improve both supervisor’s and agent’s work environment
To achieve these goals not only does the scheduling system need to meet the requirements discussed above, but the implementation of the system must be done properly.  Floor management will need training on both the use and management of the schedules.  In addition, agent training focusing on their direct benefits is a must.
  
Ultimately, if any of the above topics are not optimized, your center will either be overstaffed, have abandonment rates exceeding goal, or a combination of both!  With all of the above in mind, if you would like to diagnose whether your current agent scheduling requires attention, and to see the potential effect on your bottom line, simply click this link to open a tool that within 5 minutes will provide a personalized bottom line scheduling overview.
 
Lawrence McKenzie is a partner at Q-Solutions Group, LLC and has consulted in hundreds of contact centers worldwide.  Mr. McKenzie may be contacted at 866-960-1270 x701 or at Lawrence.McKenzie@Q-SolutionsGroup.com. For more information visit Q-SolutionsGroup.com
 
 
 
 ___________
For additional articles on our site, please click here

    

 
Return to main news page