Transera Taps into Big Data Analytics to Optimize Customer Interactions



Transera, customer engagement analytics in the cloud provider, has debuted what it says is the first statistical performance analytics solution for contact centers that provides reliable predictions on agents’ future performance and customers’ needs and propensities based on historical agent and customer activity data. Using months of data, Transera can match the customer with the best available agent for the desired business outcome.

This approach is being used to prioritize customers based upon their needs. Agents are assigned to the customer based upon the agent’s ability to achieve the desired results, including conversions, retentions and customer satisfaction. The agent is selected based upon an algorithmic scoring model that combines individual agent and overall agent population past performance data to predict an agent’s future performance with high fidelity. The algorithmic score takes into account agent churn, new agent ramp time and the unique abilities of each agent.

“Transera has created a practical, scalable solution for unlocking business value from the enormous amounts of contact center agent and customer data in enterprises today,” said Prem Uppaluru, Transera’s president and CEO, in a statement. “With Transera’s Customer Engagement Analytics SaaS offering, performance analytics is now within reach of any contact center, providing rich insights into customer interactions at a realistic price point.”

Transera’s approach to achieving true analytics-driven customer engagement consists of four stages: information; insights; intelligence; and improvement.

  • Information – Data from the various contact center systems are collected and connected in a cloud-based Customer Engagement Repository, which contains a contact center system-aware universal data dictionary and uses Big Data techniques and technologies such as Hadoop and NoSQL.
  • Insights – A customer engagement analyzer is used to segment, profile and visualize the data in the Customer Engagement Repository to identify correlations, trends and patterns that may impact performance.
  • Intelligence – Using predictive models, simulation engines, machine learning, the R analytics language and analytics techniques, the Transera Data Science Team validates insights that may predict customer propensities and agent performance.
  • Improvements – At the end of the process, the Transera data science team recommends how to optimize contact center system and agent behaviors for the desired business outcomes.

“Our analytics-driven approach to agent scoring and determining customer propensity offers a better and more accurate alternative to the many subjective approaches currently in use,” said Kumaran Ponnambalam, director of data science and analytics, Transera, in a satement. “It is because our approach is based on data science and big data analytics best practices that we have been able to increase sales conversion rates by upwards of 20 percent for our customers.”

For each of its clients, the Transera Data Science Team conducts a descriptive and predictive analytics-as-a-service process to perform:

  • Clustering – to find the true commonality among the data, such as clusters of ideal customers or agent performance levels.
  • Stability and variance analysis – to determine the data set size or time frame required to build reliable predictive models; too little will not show reliable results and too much can mask significant variances.
  • What-if analysis and modeling – to develop different hypotheses that could impact business outcomes such as call routing strategies, scripting modifications, customer prioritization, offers or proposed problem resolutions.
  • Simulation modeling - to test the different hypotheses and determine potential return on investment.