Customer journey management

• US 10,404,860 B2
• Filed: 03/14/2019
• Issued: 09/03/2019
• Est. Priority Date: 09/29/2015
• Status: Active Grant

• Current Assignee: Nice Ltd
• Original Assignee: Nice Ltd
• Inventors: Piaggio, Natalia Beatriz, Newnham, Leonard Michael
• Primary Examiner(s): Deane, Jr., William J
• Application Number:
  US16/353,288
• Publication Number:
  US 20190208060A1
• Time in Patent Office: 173 Days
• Field of Search: 37926501-26514, 37926601-2661, 379309
• CPC Class Codes: G06Q 30/0201 Market modelling; Market analysis; Collecting market data
• CPC Class Codes: G06Q 30/0203 Market surveys; Market polls
• CPC Class Codes: H04M 3/5141 Details of processing calls and other types of contacts in an unified manner unified messaging in packet-switching networks H04L51/56
• CPC Class Codes: H04M 3/5191 interacting with the Internet

Abstract

Systems and methods of managing customer journeys are implemented using one or more processors in a computing system. Each journey may comprise a succession of interactions at interaction points such as telephone conversations, responses to an interactive voice response “IVR” system and viewing a web page. Customer journey scores are determined for customers at one or more interaction points along the customer journey and the customer journey score is used to determine whether and when an intervention should take place. Models for determining customer journey scores may be created for a set of customers based on one or both of subjective and objective data relating to a subset of the set of customers that have made some or part of the journey, e.g. customers that have responded to polls. An intervention may take place during the journey or after completion of the journey.

Claims

1. A method of using a predictive model to manage customer journeys, the method comprising using one or more processors in a computer server:

receiving data defining a plurality of customer journeys, each customer journey comprising a succession of logged events representing a historical trail of previous actions performed by a particular customer in a computing system, each event corresponding to an interaction at an interaction point between a customer device and a server or other device;

at each of a plurality of interaction points of a particular customer in a customer journey;

retrieving from computer data storage a combination of independent variables representing objective and survey data relating to the customer;

using a predictive model to determine a dependent variable representing a customer journey score for the particular customer at the interaction point based on the combination of independent variables representing objective and survey data relating to the customer;

determining if the dependent variable representing the customer journey score is above a threshold; and

if the dependent variable representing the customer journey score is above the threshold, sending information to be displayed to the customer.

2. The method of claim 1, wherein the predictive model is built by executing a continuous loop in which each iteration of the continuous loop comprises:

collecting survey data from a sample of customers; and

using machine learning to build the predictive model based on the accumulated survey data.

3. The method of claim 2, wherein the predictive model is built using machine learning to learn that one or more of the independent variables in the predictive model do not affect the dependent variable and ceasing to retrieve the one or more of the independent variables in future uses of the predictive model to determine the dependent variable.

4. The method of claim 1, wherein the predictive model is built by storing, in computer data storage, data structures including one or more location identifiers for one or more predictive models, a set of values for a plurality of independent variables, a set of values for one or more dependent variables, suggested action rules for one or more interventions, and computer-executable instructions for performing the suggested action rules.

5. The method of claim 1, wherein if the dependent variable representing the customer journey score is above the threshold, comprising intervening in the customer journey to change the customer journey itself by sending the device information to modify web content.

6. The method of claim 1, wherein if the dependent variable representing the customer journey score is above the threshold, comprising intervening in the customer journey by sending information to be displayed in a pop-up window on the customer device.

7. The method of claim 1, wherein the information to be displayed to the customer is an email containing a personalized discount or offer.

8. The method of claim 1, wherein the survey data is collected from a sample of customers triggered at predetermined interaction points during the customers'"'"' journeys.

9. The method of claim 1, wherein the dependent variable is determined for the particular customer for whom there is no survey data by extrapolating obtained survey data to the particular customer who was not polled.

10. The method of claim 1, wherein the predictive model is defined at each of a plurality of interaction points based on one or more independent variables representing survey data obtained from customers within a number of journey points from the interaction point.

11. The method of claim 1 comprising, using said one or more processors, determining the dependent variable as a weighted sum of the combination of independent variables.

12. The method of claim 1 comprising, using said one or more processors:

retrieving from computer data storage descriptive data relating to the customer; and

using the predictive model to determine the dependent variable at each interaction point based on a combination of independent variables representing objective, survey and descriptive data.

13. The method of claim 1, wherein the dependent variable representing the customer journey score is an estimate of customer satisfaction.

14. The method of claim 1, wherein the customer journey comprises interactions occurring over multiple communication channels.

15. The method of claim 1, wherein the survey data is obtained by polling a random sampling of customers at an interaction point along the journey.

16. The method of claim 1, wherein said determining if the dependent variable representing the customer journey score is above a threshold is triggered by an interaction at a predetermined interaction point by the particular customer in the customer journey.

17. The method of claim 1 comprising, using said one or more processors, retrieving the predictive model from among a set of multiple predictive models for determining multiple respective dependent variables, each of the multiple respective dependent variables being a measure of a different customer journey score.

18. The method of claim 17 comprising, using said one or more processors, selecting the predictive model associated with a particular location of the interaction point in the customer journey.

19. The method of claim 18, wherein the predictive model associated with the particular location is built by filtering customers from the plurality of customers that have visited the particular location and using only data representing previous actions performed by the filtered customers.

20. A system for using a predictive model to manage customer journeys, the system comprising:

a computer server comprising one or more processors configured to;

receive data defining a plurality of customer journeys, each customer journey comprising a succession of logged events representing a historical trail of previous actions performed by a particular customer in a computing system, each event corresponding to an interaction at an interaction point between a customer device and a server or other device,

at each of a plurality of interaction points of a particular customer in a customer journey;

retrieve from computer data storage a combination of independent variables representing objective and survey data relating to the customer,

use a predictive model to determine a dependent variable representing a customer journey score for the particular customer at the interaction point based on the combination of independent variables representing objective and survey data relating to the customer,

determine if the dependent variable representing the customer journey score is above a threshold, and

if the dependent variable representing the customer journey score is above a threshold, send information to be displayed to the customer.

21. The system of claim 20, wherein the predictive model is built by executing a continuous loop in which each iteration of the continuous loop comprises:

collect survey data from a sample of customers, and

use machine learning to build the predictive model based on the accumulated survey data.

22. The system of claim 21, wherein the predictive model is built using machine learning to learn that one or more of the independent variables in the predictive model do not affect the dependent variable and cease to retrieve the one or more of the independent variables in future uses of the predictive model to determine the dependent variable.

23. The system of claim 20 comprising computer data storage configured to store data structures including one or more location identifiers for one or more predictive models, a set of values for a plurality of independent variables, a set of values for one or more dependent variables, suggested action rules for one or more interventions, and computer-executable instructions for performing the suggested action rules.

24. The system of claim 20, wherein if the dependent variable representing the customer journey score is above the threshold, changing the customer journey itself by sending the device information to modify web content.

25. The system of claim 20, wherein the information to be displayed to the customer is an email containing a personalized discount or offer.

26. The system of claim 20, wherein the one or more processors are configured to:

retrieve from computer data storage descriptive data relating to the customer, and

use the predictive model to determine the dependent variable at each interaction point based on a combination of independent variables representing objective, survey and descriptive data.

27. The system of claim 20, wherein the one or more processors are configured to retrieve the predictive model from among a set of multiple predictive models for determining multiple respective dependent variables, each of the multiple respective dependent variables being a measure of a different customer journey score.