Churn amelioration system and method therefor

• US 5,822,410 A
• Filed: 01/11/1996
• Issued: 10/13/1998
• Est. Priority Date: 01/11/1996
• Status: Expired due to Fees

• Current Assignee: Syniverse Technologies Incorporated (Syniverse Corp.)
• Original Assignee: GTE Telecommunications Services Incorporated
• Inventors: Kehoe, Richard J. Jr., Paredes, John, Tobin, Walter A., McCausland, William H.
• Primary Examiner(s): Kuntz, Curtis A.
• Assistant Examiner(s): TIEU, BINH KIEN
• Application Number:
  US08/585,232
• Time in Patent Office: 1,006 Days
• Field of Search: 379/111, 379/112, 379/114, 379/115, 379/201, 379/265, 379/211, 379/197, 379/230, 379/207, 379/92, 379/93, 379/216, 379/266, 379/309, 379/355, 379/267, 379/242, 379/243 , 379/245
• US Class Current: 379/114.01
• CPC Class Codes: H04M 15/00 Arrangements for metering, time-control or time indication ; Metering, charging or billing arrangements for voice wireline or wireless communications, e.g. VoIP

Abstract

A churn amelioration system (20) supports data feeds from mobile switching centers (32), billing systems (34), and customer service systems (36). The system (20) maintains an extensive database (42) of customer/unit data and of rules which indicate how to interpret the data. A churn history process (68) establishes rules for predicting the likelihood of churn. An administration process (86) establishes action-related and offer-related rules and assigns identified customer/units to selected organization representative user (ORU) in-boxes so that the identified customers may be proactively contacted by the ORUs. The ORUs interactively operate a churn amelioration process (126). The churn amelioration process (126) proactively or reactively identifies a specific customer/unit (128), then calculates churn vulnerability (140) and customer worth (150). The system (20) prompts ORUs with pre-approved potential offers (192) which may be extended to customers. These offers are configured in response to customer worth. The system (20) prompts ORUs with pre-approved potential actions (218) which may be extended to customers. These actions are configured in response to specified potential customer problems.

Claims

1. A method of providing customer churn amelioration information, said method comprising the steps of:

a) recording a plurality of customer proposals;

b) modeling churn history in a customer population, to define a customer churn likelihood for combinations of historical quanta and churn factor prediction data items;

c) identifying, for each of a multiplicity of customers, current quanta for said churn factor prediction data items;

d) relating said current quanta with said churn history to predict said customer churn likelihood for each of said multiplicity of customers;

e) identifying one of said multiplicity of customers in response to said predicted customer churn likelihood;

f) quantifying a customer worth for said one customer, said customer worth being responsive to a plurality of customer worth factors determined for said one customer; and

g) associating one of said customer proposals with said one customer in response to said customer worth for said one customer.

2. A method as claimed in claim 1 additionally comprising the steps of:

expressing said predicted customer churn likelihoods for said customers as vulnerabilities; and

graphically displaying said vulnerabilities at a video display terminal.

3. A method as claimed in claim 1 wherein:

said step e) additionally identifies a second one of said multiplicity of customers, said second customer being identified in response to said predicted customer churn likelihood for said second customer; and

said step g) additionally associates a second one of said customer proposals with said second customer in response to said customer worth for said second customer.

4. A method as claimed in claim 1 wherein said churn factor prediction data items include billing data.

5. A method as claimed in claim 1 wherein said step b) comprises the step of calculating a cardinal number which describes said churn likelihood.

6. A method as claimed in claim 1 wherein said step e) identifies said one customer in response to said predicted customer churn likelihood in combination with said customer worth.

7. A method as claimed in claim 1 wherein said step f) comprises the steps of:

identifying a plurality of data items that relate to customer worth;

associating a weight factor with each of said customer worth data items;

for said one customer, obtaining quantized values for said customer worth data items; and

for said one customer, calculating a summary statistic in response to said weight factors and said quantized values for said customer, said summary statistic describing said customer worth.

8. A method as claimed in claim 1 wherein said customer proposals are offers and said step g) comprises the steps of:

associating an offer value with each of said offers; and

matching said customer worth with one of said offer values to associate said one customer proposal.

9. A method as claimed in claim 1 wherein said customer proposals are offers and said step g) comprises the steps of:

associating an offer value with each of said offers;

presenting said offers having an offer value which matches said customer worth; and

receiving information indicating which one of said offers having an offer value matching said customer worth said one customer accepts.

10. A method as claimed in claim 1 additionally comprising the steps of:

identifying a plurality of data items that relate to potential customer problems;

for said one customer, obtaining quantized values for said potential customer problem data items;

calculating a summary statistic for each of said plurality of potential customer problem data items with respect to said one customer; and

presenting said summary statistics for said one customer to a user.

11. A method as claimed in claim 10 wherein said customer proposals include offers and actions, and said method additionally comprises the steps of:

associating actions with said potential customer problem data items;

identifying one of said potential customer problem data items; and

receiving information indicating which one of said actions associated with said identified potential customer problem data items said one customer accepts.

12. A method as claimed in claim 1 wherein said step f) quantifies customer worth for each of said multiplicity of customers, and said step e) comprises the steps of:

identifying a plurality of customer service representatives;

filtering said multiplicity of customers in response to said customer worth of each customer;

assigning said filtered customers to said customer service representatives; and

selecting, through a process assigned to one of said customer service representatives, said one customer front among said filtered customers assigned to said one customer service representative.

13. A method of providing customer churn amelioration information, said method comprising the steps of:

a) modeling churn history in a customer population, to define customer churn likelihoods for combinations of historical quanta and churn factor prediction data items;

b) computing a cardinal number for each of said customer churn likelihoods;

c) identifying, for each of a multiplicity of customers, current quanta for said churn factor prediction data items;

d) relating said current quanta with said churn history to obtain said cardinal number characterizing a predicted customer churn likelihood for each of said multiplicity of customers;

e) recording a plurality of customer proposals;

f) identifying one of said multiplicity of customers in response to said cardinal number for said predicted customer churn likelihood; and

g) associating one of said customer proposals with said one customer.

14. A method as claimed in claim 13 wherein said step g) comprises the steps of:

identifying a plurality of data items that relate to potential customer problems;

associating one or more of said customer proposals with each of said potential customer problem data items;

calculating a summary statistic for each of said plurality of potential customer problem data items with respect to each of said customers;

presenting said summary statistics for said one customer to a user;

selecting one of said potential customer problem data items; and

receiving information indicating which one of said customer proposals associated with said selected potential customer problem data item said one customer accepts.

15. A method as claimed in claim 13 additionally comprising the step of quantifying a customer worth for each of said multiplicity of customers so that said step f) identifies said one customer in response to said cardinal number for said predicted customer churn likelihood in combination with said customer worth.

16. A method as claimed in claim 15 wherein said customer proposals include offers and said step g) comprises the steps of:

associating an offer value with each of said offers; and

matching said customer worth with one of said offer values to associate said one customer proposal.

17. A method as claimed in claim 13 wherein said step f) comprises the steps of:

identifying a plurality of customer service representatives;

filtering said multiplicity of customers in response to at least said predicted customer churn likelihood;

assigning said filtered customers to said customer service representatives; and

selecting, through a process assigned to one of said customer service representatives, said one customer from among said filtered customers assigned to said one customer service representative.

18. A method of providing customer churn amelioration information, said method comprising the steps of:

a) predicting a likelihood of customer churn for each of a multiplicity of customers;

b) computing a cardinal number which characterizes said predicted customer churn likelihood for each of said multiplicity of customers;

c) quantifying a customer worth for each of said multiplicity of customers;

d) recording a plurality of customer proposals;

e) filtering said multiplicity of customers in response to at least one of said predicted customer churn likelihood number and said customer worth quantity;

f) identifying a plurality of customer service representatives;

g) assigning said filtered customers to said customer service representatives;

h) selecting, through a process assigned to one of said customer service representatives, one of said customers from among said filtered customers assigned to said one customer service representative; and

i) associating one of said customer proposals with said one customer.

19. A method as claimed in claim 18 additionally comprising the step of graphically displaying said customer worth at a video display terminal.

20. A method as claimed in claim 18 additionally comprising the steps of:

expressing said predicted customer churn likelihoods for said customers as vulnerabilities; and

graphically displaying said vulnerabilities at a video display terminal.

21. A method as claimed in claim 18 wherein said customer proposals include offers and said step i) comprises the steps of:

associating an offer value with each of said offers; and

matching said customer worth for said one customer with one of said offer values to associate said one customer proposal.