As we enter the busiest shopping and travel season of the year, customer service agents are bracing themselves for the inevitable increase in calls and messages. And tense situations can get even worse when the personalities of the caller and agent clashCase in point, I prefer to keep my conversations with customer service agents as short as possible, but every now and then, I get the chatty type of agent who tries to sympathize with me, when all I want is information.
But what if callers were matched with agents who are most likely to communicate well with them? Chicago-based behavioral analytics firm Mattersight saw an opportunity to do exactly that by using algorithms to match callers with service representatives.
Founded in 2006, Mattersight was formerly known as the behavioral analytics service line within call center consulting and tech firm eLoyalty. TeleTech (which is also the parent company of 1to1 Media) acquired the integrated contact solutions unit of eLoyalty in 2011 and the behavioral and analytics unit was spun off as an independent firm and renamed Mattersight Corporation.
Mattersight has collected millions of customer and employee call center interactions and analyzed the data points to find patterns in those interactions, which it stores in its Behavioral Analytics platform. The company developed three products that leverage the platform's data: Predictive Behavioral Routing, Predictive Customer Analytics and Employee Performance Management.
The Predictive Behavioral Routing application pairs callers with a client's agents based on communication styles, personality mapping, and behavioral characteristics. By analyzing a client's data about customer interactions with call center agents and comparing the data with its Behavioral Analytics platform, Mattersight groups the agents under four categories: "emotions," "opinions," "reactions," and "thoughts."
The categories refer to the types of callers agents successfully interact with. For example, agents in the emotions category communicate well with callers who respond positively to sensitive comments while agents in the thoughts category perform better with callers who prefer responses that are more fact-driven.
If a new customer calls the service line, the customer is routed to an agent who successfully interacts with all four caller types. Mattersight works with clients from various industries such as healthcare, cable TV, and consumer software.
It is possible for companies to deploy Mattersight's solution without informing agents about the technology, according to Chief People Officer Melissa Moore. "Not many companies choose to tell their agents that they're being placed in these categories," said Moore during a presentation at Frost and Sullivan's Customer Contact 2014 West conference last month.
Given that service calls are frequently recorded and analyzed, agents should not be surprised if companies look for ways to optimize their performance even without informing them, commented Forrester Research Analyst Kate Leggett, who had reviewed Mattersight's technology.
"Agents should assume that everything they do at work is being tracked from phone calls to keystrokes to help the company find more ways to recommend its products and services," she said. "And Mattersight's technology is an interesting way to affect the outcome of calls."
Mattersight commissioned a report from Forrester on the potential ROI of its Behavioral Analytics platform. As part of the review, the consultants interviewed four companies that use the Behavioral Analytics platform and Predictive Behavioral Routing application.
Based on the interviews, Forrester created a composite organization that represented a multinational firm with approximately 1,500 call center agents. Within three years of using the Predictive Behavioral Routing application, the composite firm saved approximately $4 million in reduced callbacks and gained $9.7 million in upsell opportunities. Several of the interviewees also mentioned that they saw higher customer satisfaction survey results after using Mattersight to improve call center interactions.
But Mattersight's caller and agent matching system would not be useful for all call centers, Leggett noted. The technology "works well in large call centers with one thousand to five thousand agents where you can collect a lot of data," she said. "If you only have 100 agents, it's not worth the investment."
Update: Since its briefing with Forrester, Mattersight has updated its price structure to ensure the solution is accessible to call centers of all sizes, according to a company representative.