Response attribution valuation

• US 8,676,646 B2
• Filed: 07/25/2012
• Issued: 03/18/2014
• Est. Priority Date: 11/08/2010
• Status: Active Grant

• Current Assignee: HCL Technologies Limited
• Original Assignee: International Business Machines Corporation
• Inventors: Crites, Robert, Kennedy, Ruby
• Primary Examiner(s): Weiss, John G
• Assistant Examiner(s): PATEL, DIPEN M
• Application Number:
  US13/558,095
• Publication Number:
  US 20120290371A1
• Time in Patent Office: 601 Days
• Field of Search: 705/14.52
• US Class Current: 705/14.52
• CPC Class Codes: G06Q 30/02 Marketing; Price estimation or determination; Fundraising

Abstract

A computer-implemented method is described for determining a response attribution value that represents a credit assignment to a communication, for an associated response received from a recipient during a marketing campaign.

Claims

1. A computer-implemented method, comprising:

generating, using a computer hardware system, a set of stimulus-response pairs by pairing each individual stimuli from a plurality of stimulus with each individual response from a plurality of response; and

calculating, using the computer hardware system and for each pair in the set of stimulus-response pairs, a response attribution value, wherein

the response attribution value is based upon a combination of a first model and a second model,

the first model is a response rate model, and

the second model is a time delay model that weights output of the first model based on an amount of time delay between a stimuli and a response.

2. The method of claim 1, wherein

the response attribution value represents a likelihood that the stimuli caused the response.

3. The method of claim 1, wherein

the first model is automatically selected, for a particular stimulus-response pair, from a plurality of first models based upon a type of response in the particular stimulus-response pair.

4. The method of claim 1, wherein

the second model is based upon a power law distribution.

5. The method of claim 4, wherein

the first model generates a first response attribution value, and

the second model generates power law parameters using the first response attribution value from the first model.

6. The method of claim 5, wherein

the power law parameters include:

a minimum time before the response is received, and

a rate at which responses are received as a function of time.

7. The method of claim 5, wherein

the power law parameters for the stimuli are based upon a particular channel associated with the stimuli.

8. The method of claim 1, wherein

a single response causes response attribution values for multiple stimulus to receive credit.

9. The method of claim 8, wherein

the credit for the single response is differently weighted across the multiple stimulus.

10. The method of claim 1, wherein

the first model is based on a user definition of the response for the stimuli.