System, method for deploying computing infrastructure, and method for identifying customers at risk of revenue change

• US 20060064370A1
• Filed: 09/17/2004
• Published: 03/23/2006
• Est. Priority Date: 09/17/2004
• Status: Active Grant

• Current Assignee: LinkedIn Corporation (Microsoft Corporation)
• Original Assignee: International Business Machines Corporation
• Inventors: Mojsilovic, Aleksandra, Ray, Bonnie Kathryn, Schreiber, Craig F., Wiggins, Valerie C.
• Granted Patent: US 7,870,047 B2
• Time in Patent Office: Days
• US Class Current: 705/38
• CPC Class Codes: G06Q 40/00 Finance; Insurance; Tax strategies; Processing of corporate or income taxes
• CPC Class Codes: G06Q 40/03 Credit; Loans; Processing thereof
• CPC Class Codes: G06Q 40/06 Asset management; Financial planning or analysis
• CPC Class Codes: G06Q 40/08 Insurance

Abstract

A system (and method, and method for deploying computing infrastructure) for identifying a customer at risk of revenue change (e.g., revenue loss or revenue gain) including computing a risk score based on a plurality of risk factors associated with the customer.

Claims

1. A method for identifying a customer at risk of revenue change, comprising:

computing a risk score based on a plurality of risk factors associated with said customer.

2. The method according to claim 1, wherein said computing said risk score comprises computing a plurality of risk scores based on each of said plurality of risk factors.

3. The method according to claim 1, wherein said plurality of risk factors comprises a plurality of mutually-related risk factors.

4. The method according to claim 2, further comprising:

predicting an overall risk score based on said plurality of risk scores.

5. The method according to claim 1, further comprising:

predicting an amount of revenue at risk based on said risk score.

6. The method according to claim 1, further comprising:

predicting a probability of revenue change based on said risk score.

7. The method according to claim 1, wherein said revenue change comprises at least one of revenue loss and revenue gain.

8. The method according to claim 1, further comprising:

identifying sources of change in revenue of said customer based on said risk score.

9. The method according to claim 1, further comprising:

performing an analysis of publicly available information of said customer and extracting comparative data for said customer from said analysis.

10. The method according to claim 1, further comprising:

performing an analysis of internal data of said customer and extracting comparative data for said customer based on said analysis.

11. The method according to claim 1, wherein said method further comprises:

evaluating said plurality of risk factors; and

combining said plurality of risk factors into an overall risk score.

12. The method according to claim 1, wherein said method further comprises:

inputting said plurality of risk factors;

accumulating data based on historical data; and

building a mathematical model based on said plurality of risk factors and said historical data.

13. The method according to claim 1, further comprising:

receiving said plurality of risk factors;

receiving data based on historical data; and

building a mathematical model based on said plurality of risk factors and said historical data.

14. The method according to claim 12, wherein said mathematical model adapts to said historical data.

15. The method according to claim 1, further comprising:

identifying historic examples of at least one of said customer and other customers;

extracting said plurality of risk factors associated with said customer based on analysis of at least one of publicly available information and internal data of said customer; and

predicting a probability that said customer will take an action which will result in a revenue change greater than a predetermined threshold.

16. A method for identifying at least one account at risk of revenue change, comprising:

building a mathematical predictive model that predicts a probability of at least one example from a set of historical positive examples of stable accounts and a set of historical negative examples of unstable accounts, based on a set of contract detail features for a set of historical positive examples of stable accounts and a set of historical negative examples of unstable accounts within a predetermined period of time;

extracting a set of features for said at least one account within at least a portion of said predetermined period of time;

applying said predictive mathematical model to said extracted features for said at least one account; and

computing a risk score including a probability that said at least one account belongs to one of said set of historical positive examples and said set of historical negative examples.

17. The method according to claim 16, wherein said mathematical predictive model is optimized to produce a best prediction against said set of historical positive examples and said set of historical negative examples.

18. The method according to claim 16, wherein said stable accounts comprise accounts without revenue loss within at least a portion of said predetermined period of time;

and

wherein said unstable accounts comprise accounts including at least one of revenue loss within at least a portion of said predetermined period of time and accounts including a terminated contract, a re-scoped contract, a renegotiated contract, and an unplanned decrease in revenue to a provider on a predetermined date within said predetermined period of time.

19. A method for forecasting an amount of revenue change, comprising:

building a mathematical predictive model for predicting the amount of revenue change for a historical example of accounts and revenue streams to a service provider prior to a date within a predetermined period of time;

extracting a set of features including at least one of contract detail features, internal information, and external information, for a plurality of accounts to be tracked with respect to revenue change, wherein said set of features for said plurality of accounts are extracted during a period of time preceding a date for which a forecast is sought;

applying said mathematically predictive model to the extracted set of features for each account of said plurality of accounts;

computing a revenue forecast for each account of said plurality of accounts; and

computing risk components representing a contribution of at least one feature of said set of features for said plurality of accounts to an overall revenue change.

20. The method according to claim 19, further comprising:

constructing a set of historical examples of accounts and revenue streams to said service provider prior to a date within a predetermined period of time; and

for each of said historical examples, constructing a set of historical example features including at least one of contract detail features, internal information, and external information.

21. A method for identifying an account at risk of revenue change, comprising:

building a mathematical predictive model for predicting a probability of one example of a set of historical examples of accounts, for each of a set of categories of client accounts, for a date within a predetermined period of time preceding a date for which identification is sought, belonging to one of said set of categories of client accounts,

wherein said one example includes a set of features including at least one of contract detail features, internal information, and external information, and

wherein said mathematical predictive model is optimized to produce a best prediction against said set of categories of clients;

extracting another set of features for a plurality of accounts to be tracked against said risk of revenue change, wherein said another set of features are extracted during a period of time preceding a date for which identification is sought;

applying the mathematical predictive model to said extracted another set of features for each account of said plurality of accounts;

computing at least one of a risk score and a probability that an account is within one category of said set of categories of clients, wherein at least one of said risk score and said probability represents a score indicating a risk of revenue change of said account; and

computing risk components representing a contribution of at least one feature of said set of features for said account to an overall account risk score for said account.

22. A system of identifying at least one account at risk of revenue change, comprising:

a computing unit that computes a risk score based on a plurality of risk factors associated with said at least one account.

23. The system according to claim 22, further comprising:

a predicting unit that predicts at least one of a risk of revenue change of said at least one account and an amount of revenue at risk based on said risk score.

24. A system of identifying an account at risk of revenue change, comprising:

means for computing a risk score for said at least one account based on a plurality of risk factors; and

means for predicting at least one of a risk of revenue change of said at least one account based on said risk score and an amount of revenue at risk of revenue change based on said risk score.

25. A signal-bearing medium tangibly embodying a program of machine-readable instructions executable by a digital processing apparatus to perform a method for identifying a customer at risk of revenue change, the method comprising:

computing a risk score based on a plurality of risk factors associated with said customer.

26. A method of deploying computing infrastructure in which computer-readable code is integrated into a computing system, and combines with said computing system to perform a method for identifying a customer at risk of revenue change, the method comprising:

computing a risk score based on a plurality of risk factors associated with said customer.