System and method for assessing customer segmentation strategies

• US 8,000,995 B2
• Filed: 03/22/2006
• Issued: 08/16/2011
• Est. Priority Date: 03/22/2006
• Status: Active Grant

• Current Assignee: SAS Institute Incorporated
• Original Assignee: SAS Institute Incorporated
• Inventors: Bulkley, Jack T., Chu, Chengwen R.
• Primary Examiner(s): Jarrett; Scott L
• Assistant Examiner(s): WALKER III, GEORGE H
• Application Number:
  US11/386,918
• Publication Number:
  US 20070226039A1
• Time in Patent Office: 1,973 Days
• Field of Search: 705/7.33
• US Class Current: 705/7.33
• CPC Class Codes: G06Q 30/02 Marketing; Price estimation or determination; Fundraising
• CPC Class Codes: G06Q 30/0204 Market segmentation

Abstract

In a novel computer-implemented method and system for assessing segmentation strategies, at least two models are selected for a plurality of segments. Segment performance of the segmentation strategy segments according to selected models is measured. Aggregate segmentation strategy performance data is obtained by aggregating segment performance for each segmentation strategy. Segmentation strategy performance indicia are generated to compare the aggregate segmentation strategy performance data of at least two of the segmentation strategies.

Claims

1. A computer-implemented method for assessing segmentation strategies, comprising:

defining a plurality of segmentation strategies, wherein a segmentation strategy partitions a population of consumers or clients into a plurality of segments, and wherein a segment groups members of the population according to one or more attributes;

associating, using one or more data processors, a model with each segment of a segmentation strategy, wherein the associated model is an algorithm that is used to determine a performance score for each segment, the performance score that predicts a consumer or client related activity for a population segment, wherein the consumer or client related activity comprises one or more of a response of the population segment to a marketing activity, loyalty of the population segment to a brand, account acquisitions by the population segment or fraudulent activity by the population segment;

selecting an evaluation criterion, wherein an associated model uses the evaluation criterion to determine the performance score for a segment;

aggregating, using the one or more data processors, the performance scores to generate a strategy performance score for each segmentation strategy, wherein aggregating includes sorting and binning data associated with each segmentation strategy;

selecting a segmentation strategy performance criterion comprising one or more of lift over a baseline or lift relative to an existing segmentation strategy;

determining, using the one or more data processors, whether one or more of the segmentation strategies exceeds a performance criterion based on the strategy performance scores;

based on a determination that none of the segmentation strategies exceed the performance criterion, performing one or more of;

redefining the plurality of segmentation strategies;

or

selecting an alternate evaluation criterion for use by the associated model in determining the performance score for each segment of the population;

comparing, using the one or more data processors, the strategy performance scores, wherein comparing includes determining a best strategy performance score representing the segmentation strategy predicted to produce a best overall response from the population; and

displaying a result of the comparison.

2. The method of claim 1, further comprising:

evaluating the plurality of segmentation strategies using a decision matrix.

3. The method of claim 1, wherein the associated model for each segment is a champion model for that segment.

4. The method of claim 1, further comprising:

generating a chart of the strategy performance scores for at least two of the segmentation strategies.

5. The method of claim 1, further comprising:

generating a table of the strategy performance scores for at least two of the segmentation strategies.

6. The method of claim 1, further comprising:

determining a best segmentation strategy by generating a cumulative rank for each segmentation strategy based on the strategy performance score for that segmentation strategy.

7. A system for assessing segmentation strategies, comprising:

one or more processors;

one or more computer readable storage mediums containing instructions configured to cause the one or more processors to perform operations including;

defining a plurality of segmentation strategies, wherein a segmentation strategy partitions a population of consumers or clients into a plurality of segments, and wherein a segment groups members of the population according to one or more attributes;

associating a model with each segment of a segmentation strategy, wherein the associated model is an algorithm that is used to determine a performance score for each segment, the performance score that predicts a consumer or client related activity for a population segment, wherein the consumer or client related activity comprises one or more of a response of the population segment to a marketing activity, loyalty of the population segment to a brand, account acquisitions by the population segment or fraudulent activity by the population segment;

selecting an evaluation criterion, wherein an associated model uses the evaluation criterion to determine the performance score for a segment;

aggregating the performance scores to generate a strategy performance score for each segmentation strategy, wherein aggregating includes sorting and binning data associated with each segmentation strategy;

selecting a segmentation strategy performance criterion comprising one or more of lift over a baseline or lift relative to an existing segmentation strategy;

determining whether one or more of the segmentation strategies exceeds a performance criterion based on the strategy performance scores;

based on a determination that none of the segmentation strategies exceed the performance criterion, performing one or more of;

redefining the plurality of segmentation strategies;

or

selecting an alternate evaluation criterion for use by the associated model in determining the performance score for each segment of the population;

comparing the strategy performance scores, wherein comparing includes determining a best strategy performance score representing the segmentation strategy predicted to produce a best overall response from the population; and

displaying a result of the comparison.

8. The system of claim 7, wherein the operations further comprise:

evaluating the plurality of segmentation strategies using a decision matrix.

9. The system of claim 7, wherein the associated model for each segment is a champion model for that segment.

10. The system of claim 7, wherein the operations further comprise:

generating a chart of the strategy performance scores of at least two of the segmentation strategies.

11. The system of claim 7, wherein the operations further comprise:

generating a table of the strategy performance scores of at least two of the segmentation strategies.

12. A computer program product for assessing segmentation strategies, tangibly embodied in a machine readable storage medium, including instructions configured to cause a data processing apparatus to:

define a plurality of segmentation strategies, wherein a segmentation strategy partitions a population of consumers or clients into a plurality of segments, and wherein a segment groups members of the population according to one or more attributes;

associate a model with each segment of a segmentation strategy, wherein the associated model is an algorithm that is used to determine a performance score for each segment, the performance score that predicts a consumer or client related activity for a population segment, wherein the consumer or client related activity comprises one or more of a response of the population segment to a marketing activity, loyalty of the population segment to a brand, account acquisitions by the population segment or fraudulent activity by the population segment;

select an evaluation criterion, wherein an associated model uses the evaluation criterion to determine the performance score for a segment;

aggregate the performance scores to generate a strategy performance score for each segmentation strategy, wherein aggregating includes sorting and binning data associated with each segmentation strategy;

selecting a segmentation strategy performance criterion comprising one or more of lift over a baseline or lift relative to an existing segmentation strategy;

determine whether one or more of the segmentation strategies exceeds a performance criterion based on the strategy performance scores;

based on a determination that none of the segmentation strategies exceed the performance criterion, performing one or more of;

redefine the plurality of segmentation strategies;

or

select an alternate evaluation criterion for use by the associated model in determining the performance score for each segment of the population;

compare the strategy performance scores, wherein comparing includes determining a best strategy performance score representing the segmentation strategy predicted to produce a best overall response from the population; and

display a result of the comparison.