Conversational Outcome Modeling With Turn-Level Sales Impact
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing systems struggle to accurately predict the likelihood of a successful sale outcome in sales conversations and determine the impact of individual conversation turns on this outcome, making it difficult for customer service representatives to optimize their strategies.
Innovation Solution
An intelligent prediction system using neural network models analyzes transcribed conversation data to generate point-in-time bind probabilities and segment impact scores, allowing for real-time assessment of conversation progress and strategic adjustments.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If natural language models are employed to guide virtual conversations and generate output, then the conversation flow is improved, but the ability to determine the impact of sale strategies is insufficient
Solution Approach 1:
The patent segments the conversation into individual speaker turn segments and analyzes each segment's impact on bind probability separately. This allows the system to provide both overall conversation guidance and specific feedback on the impact of individual sales strategies used at different points in the conversation.
Solution Approach 2:
The system implements feedback by generating speaker turn segment impact scores that quantify the effect of each conversational turn on the likelihood of successful sale outcome. This feedback mechanism enables sales representatives to understand the impact of their strategies and adjust accordingly.
2Measurement precision
If traditional prediction methods are used, then the system complexity is low, but the prediction accuracy of sale outcomes is insufficient
Solution Approach 1:
The patent transforms textual speaker turn segments into numerical vectors and uses these transformed parameters as input to neural network models. This parameter transformation enables the system to achieve high prediction accuracy by leveraging the pattern recognition capabilities of neural networks on structured numerical data derived from conversation text.
3Loss of information
If detailed analysis of each speaker turn is performed, then the insight into sale strategies is improved, but the processing time increases
Solution Approach 1:
The system performs preliminary vectorization of speaker turn segments and pre-processing of text data before the actual prediction analysis. This preliminary action prepares the data in advance, enabling faster processing during the prediction phase while maintaining detailed analysis capabilities.
Solution Approach 2:
The patent replaces manual analysis of conversational turns with automated neural network-based prediction models. This substitution eliminates the need for time-consuming manual evaluation while providing comprehensive analysis of each speaker turn's impact on sale outcomes.
Data Source
AI summary
An intelligent prediction system includes one or more processors, one or more memory components, and machine-readable instructions that cause the intelligent prediction system to: receive text data comprising a plurality of speaker turn segments of a transcription of a conversation, each speaker turn segment of the plurality of speaker turn segments representative of a turn in the conversation, the plurality of speaker turn segments collectively representative of the conversation up to a point of time, generate a point in time bind probability based on a speaker turn segment bind probability of a speaker turn segment at the point in time and memory data associated with the plurality of segments up to the point in time, and generate a speaker turn segment impact score at the point in time by subtracting an immediately preceding point in time bind probability from the point in time bind probability.