Delivering a superior customer experience is more challenging than ever: There are more touch points to manage, more data to analyze, and customer behavior is often unpredictable and irrational, which puts companies on the defensive when it comes to anticipating and understanding current and future customer needs.
Data analytics is transforming contact center outcomes by taking key aspects of customer information and cycling it through an automatic process of data gathering and workflow planning throughout the customer journey. Then, these results can be analyzed and suggestions can be made as to what should be changed or done to optimize for the desired results. While predictive analytics can help businesses understand how their current contact center teams, processes, scripts, and offers will perform in the future, they should not stop there. To improve contact center performance and the customer experience, prescriptive analytics is an equally critical component.
The value of predictive analytics in the contact center
Contact center interaction management solutions have traditionally taken a retroactive approach, examining past interactions to form judgments on the value of customers today. In many contact centers, callers are still routed to the agent who has been available the longest or next in line using traditional routing strategies without regard to the agent or contact center business performance.
Advanced contact centers today are going a step further, using intelligent routing to send callers to the agent best able to handle the call based on the customer needs. As opposed to availability-based routing, analytics enable organizations to define rules in a decision engine and let predictive analytics route calls in a more dynamic fashion using historical performance data. For example, in the call center, historical data would be the outcomes of past customer interactions (first call resolutions, sales, upsells, customer satisfaction) based on customer past behaviors, demographics and needs; as well as the scripts agents used, their training, skills, and performance delivering the desired results with similar customers with similar needs.
Predictive analytics would use this historical data to predict the customer's propensity to have specific needs, to churn, abandon a call, or be upsold or cross-sold and the potential outcomes of the customer interaction. Using analytics to predict behaviors and interaction outcomes allows organizations to eliminate unpredictability, anticipate customer needs, and take the right course of action to meet or exceed them. By doing so, organizations can shift from simply routing customers to the next available agent and instead connect them with the right agent with the best performance, skills, and offers at the right time.
Prescriptive analytics increases value of predictive analytics
However, predictive analytics, on its own, does not fully reduce the business risk of decision making in the contact center. Gartner analyst Jim Hare cautions as much, noting that, "despite the massive amount of available computing power, data, and analytics, many organizational leaders continue to make critical business decisions based on intuition and speed rather than robust analysis." That is where prescriptive analytics come in; it leverages predictive analytics to go one step further and reduce the risk of decisions and optimize the contact center for positive outcomes and business performance.
Prescriptive analytics looks at the predictive analytics results and makes suggestions as to what should be changed or done to optimize for the desired results. To be most effective, this should be an iterative process performing what-if simulations for different strategies and seeing the business outcomes under each. For example, in the contact center this might be simulating different call routing strategies or the use of different offers and scripts to see their impact on business results.
Prescriptive analytics is important because it can be used to determine the best course of action to meet customers' need. It can determine the best way to increase agent performance through training, new scripts, skill development, or by only using them for certain kinds of callers or customers with specified needs (sales, retention or support). Prescriptive analytics can also identify ways to optimize results based on the key performance indicators (KPIs) that drive the overall business, not just KPIs that measure contact center efficiency. This includes things such as profitability, customer satisfaction, and first call resolution.
Predictive and prescriptive should be a package deal
Together, prescriptive and predictive analytics can be used to optimize customer experiences and the overall performance of the contact center, both from operational and business perspectives. For example, when a customer engages with the contact center, predictive analytics predicts the customer need and the outcome of the call (or agent performance) if the customer is connected to any given agent using any given script.
Prescriptive analytics follows by determining the best course of action to optimize the outcome of the customer interaction. With the increasingly competitive global economy, businesses need the contact center to optimize not just for service levels, but for business outcomes such as profitability, customer retention and customer satisfaction. Predictive and prescriptive analytics-driven contact centers allow businesses to accomplish these objectives.
Steve Kaish is vice president of product marketing and alliances at BroadSoft, a provider of unified communication software as a service (UCaaS).