The great thing about working in contact centers is that data is everywhere: If you want to know how many times...
a customer has called about a delivery issue, you can find it. If you want to know what a call center agent's average time spent on a call is, you can find it. If you want to know a metric for total customer satisfaction with service, you can find it.
But the challenge becomes how to ensure that the call center data can be assembled in a meaningful way. How can the numbers enable decision makers to translate the "science" of the numbers into the "art" of analysis and an action plan?
The goal is for individual data points to become actionable pieces of knowledge that can solve problems. When individual quality-monitoring scores are aggregated and identify the areas where agents score the lowest, they become a valuable piece of information that may identify a training need.
Analytics takes information a step further, enabling the discovery and communication of meaningful patterns of data. Call center data analysis can identify cause and effect in particular events, such as which triggers have prompted customer attrition.
There are three types of analytics that are available to contact centers:
This is the consolidation of all call center data from sources such as the automatic call distributor, a CRM system, quality monitoring forms and workforce management systems. These sources provide a 360-degree view of the internal operation but not an assessment of customer experience.
An example of a simple use of data analytics is to identify the top reasons customers call the contact center. With this type of information, organizations can improve collateral and self-service capabilities to reduce the number of incoming calls.
Speech analytics analyzes the words, voice patterns and speech inflection of customers (and, in some cases, of agents). Speech analytics focuses on the voice of the customer, enabling the segregation of contacts based on specific keywords and verbal cues.
An example of the value of speech analytics is to segregate a sample of calls where the customer says, "I want to cancel" for further evaluation by a customer retention group. By analyzing the dynamics of these calls, contact centers may be able to reduce customer attrition.
Sentiment analytics is relatively new to contact centers. It provides a tool to determine the attitude, judgment or evaluation of a customer. It moves from an objective viewpoint to a subjective one. An example of sentiment analytics is to determine the likelihood of a repurchase decision by a customer as a result of the emotion they emitted during a contact.
The opportunity is to target specific messages to customers that appeal to particular emotions they demonstrated during a call.
The shift from data to information to analytics is a powerful transformation that allows organizations to understand internal results and external customer perception. The challenge for organizations is that conclusions can be subject to interpretation. Leveraging the output of analytics may work best where the contact center can examine trends over time, such as various customers' reactions to a change in delivery process or by analyzing several calls completed by the same agent. In this case, sentiment analysis can bear out some more reliable trends.
These challenges provide limitations to reacting solely to the outcome of various call center data analysis tools. Analytics provides the "science"; there is still a required "art" that adds the "color commentary" to the overall call center data analysis. When you include the "art," there are some rules it's critical that you follow:
- Establish an adequate sample size (it does not have to be a valid sample size) before reacting.
- Establish a baseline and react to changes in trends.
- Establish a process to address the root cause of items.
Contact centers provide an opportunity to report everything that can be measured, and there are various ways for the metrics to be presented: as data points, information or analytics. The key challenge is to understand the value -- and the limitations -- of these metrics and to ensure that a proper dose of art complements the science.
Giving agents the knowledge they need
Characteristics of a modern contact center
Poorly managed automation practices can cost you
Video, live chat represent customer service challenges