The purpose of speech analytics is automatically mine for customer intelligence and performance optimization data within the context of a recorded voice interaction. Used in conjunction with a contact center’s recording technology, speech analytics scans recorded conversations with the objective of finding key words or phrases that can offer the user insights into such common business and performance factors
According to Saddletree Research, there are essentially three business use cases for analyzing speech.
1. Categorization (Proactive)
2. Discovery (Proactive)
3. Search and Exploration (Reactive)
Categorization – In the speech categorization process, calls are categorized based upon phrases used by agents and customers during the call. During this process, key agent performance indicators are measured via the recognition of critical phrases that occur during the call and are flagged for review and action as appropriate. Speech categorization can drive contact center performance management through the assurance of agent performance against key performance indicators (KPIs). As a result, speech categorization delivers a high degree of business value.
Discovery- Discovery automatically uncovers trends or events occurring within conversations. These are typically trends or events of which the organization may not have previously been aware. Discovery can be useful for proactively uncovering important trends or issues before they become major problems for the organization.
Search and Exploration – Search-and-exploration, is the process of searching for individual words or phrases spoken within calls, generally on an ad-hoc basis. Search is useful within many analysis workflows and can be used for ad-hoc exploration of conversation content or to test hypotheses.
Written by Paul Stockford, The Evolution of Speech Analytics From Word Spotting to Driving Business Value, discusses Speech Analytics Technological Approaches for each case, the profile of today’s speech analytics user with a discussion of Utopy SpeechMiner 7.0