System and method for AB testing based on communication content
Systems and methods include identifying sets of communications using AB testing based on a difference in metadata values between the sets for a first type of metadata, determining at least one similarity between communications in a first set based on metadata values of a second type of metadata, determining at least one similarity between the communications in the second set based on metadata values of the second type of metadata or a third type of metadata, comparing the at least one similarity from the first set to the at least one similarity of the second set to identify differences in metadata values or types of metadata, and determining, based on the differences, a suggested action for an agent for a future communication with respect to at least one of a desired script, a target phrase, or a desired acoustic characteristic.
1. A method, executed by a computer-based communications analytics facility, comprising:
performing an analysis of acoustic characteristics and language characteristics of a plurality of communications between customers and agents;
assigning respective metadata values to communications of the plurality of communications based on the analysis of the acoustic characteristics and the language characteristics, wherein each metadata value relates to a type of metadata comprising at least one of: a category of a communication, an overall call performance, a score, a successful sale, a successful upselling, a script, a word, a phrase, a tone, a satisfaction, a dissatisfaction, a duration of the communication, a duration of silence, an agitation level, a compliment, an agent, an agent group, or an agent ownership;
identifying at least a first set of communications and a second set of communications of the plurality of communications using AB testing based on a difference in metadata values between the first set and the second set for a first type of metadata;
determining at least one similarity between the communications in the first set based on metadata values of a second type of metadata;
determining at least one similarity between the communications in the second set based on metadata values of the second type of metadata or a third type of metadata;
comparing the at least one similarity from the first set to the at least one similarity of the second set to identify differences in metadata values or types of metadata; and
determining, based on the differences, a suggested action for an agent for a future communication with respect to at least one of a desired script, a target phrase, or a desired acoustic characteristic.
2. The method of claim 1 , wherein the first type of metadata relates to an agent score.
3. The method of claim 1 , wherein the first type of metadata relates to an agent success with respect to sales.
4. The method of claim 3 , wherein the second type of metadata relates to words of the communications.
5. The method of claim 1 , wherein the first set of communications is associated with use of a first script by one or more agents and the second set of communications is associated with use of a second script by the one or more agents.
6. The method of claim 5 , wherein the second type of metadata relates to an agent score of the one or more agents using the first script.
7. The method of claim 5 , wherein the second type of metadata relates to a successful sale of the one or more agents.
8. The method of claim 5 , wherein the third type of metadata is used to determine the at least one similarity between the communications in the second set, and wherein the third type of metadata relates to an agent score of the one or more agents using the second script.
9. The method of claim 1 , wherein multiple factors are used to determine multiple similarities between the communications in the first set.
10. The method of claim 1 , wherein multiple factors are used to determine multiple similarities between the communications in the second set.
11. The method of claim 1 , wherein the plurality of communications is filtered according to a time frame.
12. The method of claim 1 , wherein the first set of communications is based on a first group of agents and the second set of communications is based on a second group of agents.