Evaluation of Client Status for Likelihood of Churn

• US 20110218955A1
• Filed: 03/08/2010
• Published: 09/08/2011
• Est. Priority Date: 03/08/2010
• Status: Active Grant

• Current Assignee: Micro Focus LLC (Open Text Corporation)
• Original Assignee: Hewlett-Packard Development Company, L.P. (HP Inc.)
• Inventors: DYER, Justin S., TANG, Hsiu-Khuern
• Granted Patent: US 8,478,709 B2
• Time in Patent Office: Days
• US Class Current: 706/52
• CPC Class Codes: G06F 15/16 Combinations of two or more digital computers each having at least an arithmetic unit, a program unit and a register, e.g. for a simultaneous processing of several programs coordinating program control therefor G06F9/52; in regulating and control system G05B
• CPC Class Codes: G06N 5/02 Knowledge representation; Symbolic representation
• CPC Class Codes: G06Q 10/04 Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem" market predictions or forecasting for commercial activities G06Q30/0202
• CPC Class Codes: G06Q 30/02 Marketing; Price estimation or determination; Fundraising

Abstract

System, including method, apparatus, and computer-readable media, for evaluating client status for a likelihood of churn. Client data may be received, with the client data representing events from a set of different event types performed by clients. Parameters of a statistical model that describes client behavior may be estimated using a computer and based on the client data. A churn type of event may be encoded in the statistical model as an absorbing state of a stochastic process, with a time of transition to the absorbing state modeled as being infinite. At least one of the parameters may correspond to the churn type of event. A likelihood of churn may be calculated for a plurality of the clients at one or more time points using the statistical model and its estimated parameters.

Claims

1. A method of evaluating client status, comprising:

receiving client data representing events from a set of different event types performed by clients;

estimating, using a computer and based on the client data, parameters of a statistical model that describes client behavior, wherein a chum type of event is encoded in the statistical model as an absorbing state of a stochastic process, with a time of transition to the absorbing state modeled as being infinite, and wherein at least one of the parameters corresponds to the churn type of event; and

calculating a likelihood of churn for a plurality of the clients at one or more time points using the statistical model and its estimated parameters.

2. The method of claim 1, wherein estimating includes basing the statistical model on an assumption that the events for each client form a sequence of event types that follows a Markov chain.

3. The method of claim 2, wherein basing includes making an assumption that the sequence of event types is time-homogeneous.

4. The method of claim 1, wherein estimating includes basing the statistical model on an assumption that a probability distribution of a time interval between a given pair of consecutive events is dependent on the event type for a temporally first member of the given pair of consecutive events.

5. The method of claim 1, wherein estimating includes basing the statistical model on an assumption that a probability distribution of a time interval between a given pair of consecutive events is dependent on the event type for both members of the given pair of consecutive events.

6. The method of claim 1, wherein estimating includes decoupling an estimation of one or more parameters associated with predicting a sequence of events for each client from an estimation of one or more parameters associated with predicting time intervals between events for each client.

7. The method of claim 6, wherein estimating includes (a) estimating one or more parameters associated with predicting a sequence of events for each client and (b) estimating one or more parameters associated with predicting time intervals between events, and wherein (a) and (b) are performed separately and iteratively.

8. The method of claim 1, further comprising a step of sending a communication to each client for which the likelihood meets a predefined condition.

9. An article comprising at least one computer readable storage medium containing instructions executable by a computer to perform a method of evaluating client status, the method comprising:

receiving client data representing events from a set of different event types performed by clients;

estimating, using a computer and based on the client data, parameters of a statistical model that describes client behavior, the statistical model being based on an assumption that the events for each client form a sequence of event types that follows a Markov chain, wherein a churn type of event is encoded in the statistical model as an absorbing state of a stochastic process, with a time of transition to the absorbing state modeled as being infinite, and wherein at least one of the parameters corresponds to the churn type of event; and

calculating a likelihood of churn for a plurality of the clients at one or more time points using the statistical model and its estimated parameters.

10. The article of claim 9, wherein estimating includes making an assumption that the sequence of event types is time-homogeneous.

11. The article of claim 9, wherein estimating includes basing the statistical model on an assumption that a probability distribution of a time interval between a given pair of consecutive events is dependent on the event type for a temporally first member of the given pair of consecutive events.

12. The article of claim 9, wherein estimating includes basing the statistical model on an assumption that a probability distribution of a time interval between a given pair of consecutive events is dependent on the event type for both members of the given pair of consecutive events.

13. The article of claim 9, wherein estimating includes decoupling an estimation of one or more parameters associated with predicting a sequence of events for each client from an estimation of one or more parameters associated with predicting time intervals between events for each client.

14. The article of claim 13, wherein estimating includes (a) estimating one or more parameters associated with predicting a sequence of events for each client and (b) estimating one or more parameters associated with predicting time intervals between events, and wherein (a) and (b) are performed separately and iteratively.

15. The article of claim 9, further comprising a step of selecting clients for which the likelihood meets a predefined condition, to enable sending a communication to the clients selected.

16. An apparatus for evaluating client status, comprising:

a storage medium to receive client data representing events from a set of different event types performed by clients;

a parameter estimation routine that estimates parameters of a statistical model describing client behavior based on the client data, wherein a chum type of event is encoded in the statistical model as an absorbing state of a stochastic process, with a time of transition to the absorbing state modeled as being infinite, and wherein at least one of the parameters corresponds to the churn type of event; and

a likelihood calculator that calculates a likelihood of churn for a plurality of the clients at one or more time points using the statistical model and its estimated parameters.

17. The apparatus of claim 16, wherein the parameter estimation routine bases the statistical model on an assumption that a probability distribution of a time interval between a given pair of consecutive events is dependent on the event type for at least one member of the given pair of consecutive events.

18. The apparatus of claim 16, wherein the parameter estimation routine decouples an estimation of one or more parameters associated with predicting a sequence of events for each client from an estimation of one or more parameters associated with predicting time intervals between events for each client.

19. The apparatus of claim 16, wherein the parameter estimation routine (a) estimates one or more parameters associated with predicting a sequence of events for each client and (b) estimates one or more parameters associated with predicting time intervals between events, and wherein (a) and (b) are performed separately and iteratively.

20. The apparatus of claim 16, further comprising a client selector that selects clients for which the likelihood of churn meets a predefined condition.