Call Center Management Featured Article
August 05, 2009
Accurate Forecasting in the Call Center: 5 Percent Can Make All the Difference
In this current economy, call center managers are under intense pressure to schedule agents as accurately as possible: Schedule too many agents for any particular shift and the manager will have staffers sitting around idly, doing nothing, wasting precious company dollars.
Not only does this lead to unnecessarily high staffing costs, but it also leads to agent boredom, which in turn affects job satisfaction. Schedule not enough agents and call hold times and abandonment rates will increase, service levels will erode, and customer satisfaction will take a dive. What’s more you’ll end up with longer calls, leading to higher telephone network usage and costs, as well as increased agent stress and burnout, leading to higher turnover.
Not only does this lead to unnecessarily high staffing costs, but it also leads to agent boredom, which in turn affects job satisfaction. Schedule not enough agents and call hold times and abandonment rates will increase, service levels will erode, and customer satisfaction will take a dive. What’s more you’ll end up with longer calls, leading to higher telephone network usage and costs, as well as increased agent stress and burnout, leading to higher turnover.
Therefore, the biggest challenge call center managers face today is accurately forecasting how many agents will be needed for any particular shift. This has actually become more difficult in the past couple of years, due to the fact that patterns in customer behavior have shifted dramatically due to the recession.
For example, a call center manager working for an e-commerce company might have a difficult time forecasting call and contact volume based on past customer contact patterns. Not only is it likely that the overall volume of contacts is way down, compared to previous years, the timing of when those contacts come in may have also changed significantly – in other words, customers may no longer be calling in and making purchases at the same time of year, same days of the week, or even the same hours of the day, as they used to in the past.
Add the fact that customers are using new channels – such as e-mail and Web chat – to place orders and make inquiries, and it’s easy to see how challenging it is to forecast contact volume and schedule the proper number of agents. What’s more, call center managers need to schedule agents based on their unique skill sets: Some agents might be trained to handle Web chat and e-mail, while others are trained to only handle phone calls. Put these factors together and it’s easy to see how getting the right “mix” of agents for each shift can be incredibly challenging.
But just how accurate do call center managers need to be when it comes to forecasting call and contact volume? A close look at this topic reveals that small inaccuracies in forecasting can have a huge impact on the bottom line. In fact, by some estimates, being off by as little as 5 percent can have profoundly negative effect on call center efficiency. As such, getting this area of planning right can be essential to the overall health of the business.
The key is to develop a methodology that takes into account linkages between variables and factors overlooked in past data. But this cannot be achieved using manual systems, such as spreadsheets – it takes a software-based approach, and this is where workforce management solutions come into play. Most workforce management systems include advanced analytics capabilities that allow call center mangers to predict how many agents are needed for any particular shift. This is achieved through integration with the call center ACD: The WFM system is able to extract and analyze historical call and contact volumes in order to calculate a forecast.
But as mentioned earlier, customer contact patterns have changed due to the economic downturn. For this reason, call center managers need to put a stronger emphasis on recent trends and patterns when running such forecasts, as opposed to trends that are years old. Using a WFM system, they can put more weight on recent patterns in contact volume in order to arrive at a more accurate forecast. The patterns of 2006 and 2007 certainly still count – but their value in predicting volume in this new economy is greatly diminished.
In addition, forecasts must also incorporate additional, more advanced approaches — explanatory methods that reveal linkages between variables, as well as “judgmental forecasting” that accounts for factors not reflected in historical patterns. This is also where today’s WFM systems are playing an increasingly important role.
Remember that the goal is to not only ensure that the proper number of agents is scheduled for each shift, in order to meet service levels, but also to ensure that agent time is being efficiently utilized. It’s important to account for the relationships between staffing level, service level (percentage of calls handled within a specified timeframe), average speed of answer, agent occupancy (the percentage of time agents handle calls versus wait for calls) and trunk load. Remember that when service levels go up, agent occupancy goes down. A manager might be doing a great job meeting service levels – but how efficiently is the center being run?
The other side of this is the impact on customers. As demonstrated using the Erlang C model, a difference of just a few agents can have a dramatic impact on hold times. For example, suppose that, as per your intended service level, 65 out of 240 customers will have a hold time of five seconds or more, with 34 agents working. Reduce the number of agents to 30 and you will increase hold times to 10 seconds or more for 80 of those 240 customers.
The bottom line is, small inaccuracies in forecasting can have a huge impact on call center efficiency. Using spreadsheets and manual systems for scheduling agents no longer cuts it -- you need advanced forecasting capabilities -- and that’s why today’s workforce management systems have become such critical call center management tools. WFM solutions such as Monet Software’s WFM Live, when properly integrated with other call center systems, provide the means to accurately forecast the needed number of agents for any shift – typically within the aforementioned 5 percent accuracy range.
Monet Software is currently offering its Web-based WFM Live service on a 30-day trial basis. For more information, click here.
Monet Software is currently offering its Web-based WFM Live service on a 30-day trial basis. For more information, click here.
Follow ITEXPO (News - Alert) on Twitter: twitter.com/itexpo
Patrick Barnard is a contributing writer for TMCnet. To read more of Patrick’s articles, please visit his columnist page.