Skip to content

Question: Predictive behavioral routing was mentioned in a recent “Ask a DMG Expert” answer. What is that?

Answer:

Predictive behavioral routing (PBR) technology matches customers with agents based on personality, communication style, emotional state, previous interactions, and more, to optimize conversations for both the customer and the organization. This functionality is currently being sold as an add-on to core routing solutions such as automatic call distributors (ACDs). Predictive behavioral routing leverages artificial intelligence (AI)-based algorithms in real time to determine the best pairing of customer and agent for the current need and routes the interaction without noticeable delay. Although there are some differences among the PBR solutions offered in the market, in general, the application captures and analyzes all available information about the customer and the agent, sourced from customer relationship management (CRM) applications or other servicing solutions, internal analytics, performance management applications, etc., to “match” characteristics and behaviors. The solution then routes the customer to an agent with a similar communication approach. A number of AI technologies are leveraged by PBR, including natural language processing/understanding (NLP/NLU) (to understand customer intent and current emotional state), predictive analytics (to determine the optimal customer/agent match), and machine learning (to improve future performance). 

While predictive behavioral routing technology is relatively new to the contact center market, it has proven its ability to reduce average handle time for phone inquiries and digital interactions. However, the benefits from PBR are broader, as these solutions can increase first contact resolution rates, reduce customer effort, decrease customer churn, improve agent engagement and retention, and enhance the customer experience. And the benefits are not limited to routing interactions to human agents. PBR can help determine if it’s appropriate to offer a customer the option of self-service, based on their verbal cues, history, complexity of the current issue, present emotional state, etc.