subject: The silver Bullet For Customer Retention [print this page] NICE Speech analytics enables companies to apply advanced technologiestranscription to translate speech into text, phonetic indexing to analyze and categorize calls by type, and emotion detection to identify even slight variations in tone and pitchto unlock the wealth of information contained in calls made to the call center. When applied to calls, speech analytics can reveal such insights as:
Competitive Information Did the customer mention a competitor? Was a competitive feature or rate plan mentioned?
Emotion Detection What emotions did the caller communicate? What were the agents emotions?
Call Disposition Why did the customer call? Was it a repeat call? Was it escalated? Was her issue resolved? Did she indicate an intention to terminate her contract?
Agent Efficacy How effective was the customer support agent? Did he answer the questions asked? Was the call managed properly? Did the agent talk over the customer or vice versa?
Trend Analysis How did customer behavior change over time? Is there an increase or decrease in customer dissatisfaction?
The use of speech analytics alone can predict churn with more than 80% accuracy, but only among the small subset of customers who contact the call center.
Transactional Churn Prediction
Combining aspects of speech analytics with historical, behavioral data captured by transactional billing, campaign management, customer service and support systems can yield insights into:
Retention Which customer is most likely to leave and why? How can he be saved?
Loyalty What behaviors are most strongly correlated with loyalty? How can the most loyal customers be nurtured?
Profitability Which customers will be receptive to cross-sell/up-sell offers?
Acquisition What potential customers should be targeted?
Advanced campaign management often includes customer-offer optimization, as a complement to transactional churn prediction. By modeling retention offers against churn-vulnerable customers, this approach has been shown to increase the acceptance of retention offers by as much as 700%, thereby stemming customer defections. However, it sheds no light on root causes of dissatisfaction or other information that can be used to proactively combat churn.
The Silver Bullet: Combining Speech Analytics with Transactional Predictive Models
By applying statistical predictive models to speech analytics-generated data gleaned from customer interactions and transactional data, companies can extrapolates the characteristics of customers who contact customer care to the entire customer base to improve churn prediction beyond the call center set. Combining the structured transactional data with unstructured voice data, rendered intelligible using speech analytics, churn prediction is as much as 20% to 30% more accurate. The marriage of this information also yields insight into the reasons for potential defection, helping to focus retention efforts and improve customer care.
To learn more about how speech analytics can be used to retain customers, download the NICE white paper, The Retention Silver Bullet: Using Speech Analytics to Predict Churn. [DEB1]The paper includes a powerful case study in which a US-based wireless carrier integrated its existing AnswerOn Solution for Customer Retention with NICE Interaction Analytics[DEB2], generating a 61% increase in churn-prediction accuracy and $1.86 million in projected churn-related savings in one year.
