Methods and apparatus to incorporate saturation effects into marketing mix models

• US 9,721,271 B2
• Filed: 03/15/2013
• Issued: 08/01/2017
• Est. Priority Date: 03/15/2013
• Status: Active Grant

• Current Assignee: Nielsen Company (US) LLC (RR Donnelley & Sons Company)
• Original Assignee: The Nielsen Company LLC (MOOD Media Corporation)
• Inventors: Brixius, Nathan, Link, Ross
• Primary Examiner(s): Hamilton, Matthew L
• Application Number:
  US13/835,695
• Publication Number:
  US 20140278930A1
• Time in Patent Office: 1,600 Days
• Field of Search: 705 14
• CPC Class Codes: G06Q 30/0242 Determining effectiveness of advertisements
• CPC Class Codes: G06Q 30/0244 Optimization
• CPC Class Codes: G06Q 30/0246 Traffic

Abstract

Methods and apparatus to incorporate saturation effects into marketing mix models are disclosed. A disclosed method includes calculating adstocked gross rating points associated with an advertising campaign during segments of a period of time. The adstocked gross rating points are based on raw gross rating points corresponding to the advertising campaign. The raw gross rating points are delivered during the segments of the period of time. The example method further includes calculating an effective reach realized during the segments of the period of time for the advertising campaign. The effective reach realized is based on the adstocked gross rating points.

Claims

1. A method comprising:

calculating, by a processor, adstocked gross rating points associated with an advertising campaign during segments of a period of time, the adstocked gross rating points based on raw gross rating points corresponding to the advertising campaign, the raw gross rating points delivered during the segments of the period of time, the raw gross rating points defined by marketing mix input data stored in a marketing mix inputs database, the marketing mix input data including historical marketing data and historical sales data;

transforming, by the processor, the adstocked gross rating points into an effective reach realized during the segments of the period of time for the advertising campaign, the transformation based on determining a relationship indicative of a saturation effect on a reach of the advertising campaign that best fits the marketing mix input data, the effective reach realized indicative of a percentage of target audience members effectively exposed, during each segment, to an advertisement of the advertising campaign an average number of times sufficient to elicit a response to the advertisement, the effective exposure to the advertisement including actual exposure to the advertisement during each segment and a time lagged effect of exposure to the advertisement during previous segments;

generating, by the processor, a marketing mix model that automatically accounts for the saturation effect without manual adjustment, by one or more users, of a response curve associated with the advertising campaign, the accounting for the saturation effect based on a regression analysis of the effective reach realized during the segments of the period of time for the advertising campaign and the marketing mix input data retrieved from the marketing mix inputs database; and

reducing a computational burden on the processor by eliminating tuning of response curves after the generation of the marketing mix model and determining a best fitting saturation curve corresponding to the relationship to increase an accuracy of predictions of future sales associated with the advertising campaign and increase a confidence in plans for future marketing endeavors.

2. The method of claim 1, further including determining a first contribution of the advertising campaign to a lift in sales over the period of time based on the effective reach realized relative to a second contribution of another advertising campaign.

3. The method of claim 1, wherein the determining of the relationship includes looking up a plausible saturation curve stored in a saturation curve database.

4. The method of claim 1, wherein the determining of the relationship is based on an initial penetration associated with the advertising campaign and an initial effective frequency associated with the advertising campaign.

5. The method of claim 4, further including determining the initial penetration based on audience measurement data, the audience measurement data including at least one of demographic information corresponding to a target audience of the advertising campaign or media type information corresponding to a type of media associated with the advertising campaign.

6. The method of claim 4, further including determining the initial penetration and the initial effective frequency based on at least one of a geographic region associated with the advertising campaign, a type of product or service associated with the advertising campaign, a comparable penetration associated with a comparable campaign for related or competing products, a comparable effective frequency associated with the comparable campaign, or a complexity of the advertising campaign.

7. The method of claim 4, wherein the determining of the relationship includes:

determining a range of plausible penetrations based on the initial penetration;

determining a range of plausible effective frequencies based on the initial effective frequency;

calculating a range of plausible saturation curves based on the range of plausible penetrations and the range of plausible effective frequencies; and

analyzing the range of plausible saturation curves to identify the best fitting saturation curve with respect to the marketing mix input data associated with the advertising campaign.

8. The method of claim 4, wherein the determining of the relationship includes:

generating a simulated advertising campaign based on the initial penetration and the initial effective frequency;

determining data points indicative of simulated reach against simulated gross rating points associated with the simulated advertising campaign; and

calculating an equation that fits the data points, the equation corresponding to the relationship.

16. The method of claim 1, wherein the determining of the relationship includes an initial regression analysis to test different saturation curves for the advertising campaign against the marketing mix input data while other advertising campaigns represented in the marketing mix input data are each assigned a default saturation curve.

9. A tangible machine readable storage medium comprising instructions, which when executed by a processor, cause a machine to at least:

calculate adstocked gross rating points associated with an advertising campaign during segments of a period of time, the adstocked gross rating points based on raw gross rating points corresponding to the advertising campaign, the raw gross rating points delivered during the segments of the period of time, the raw gross rating points defined by marketing mix input data stored in a marketing mix inputs database, the marketing mix input data including historical marketing data and historical sales data;

transform the adstocked gross rating points into an effective reach realized during the segments of the period of time for the advertising campaign, the transformation based on determining a relationship indicative of a saturation effect on a reach of the advertising campaign that best fits the marketing mix input data, the effective reach realized indicative of a percentage of target audience members effectively exposed, during each segment, to an advertisement of the advertising campaign an average number of times sufficient to elicit a response to the advertisement, the effective exposure to the advertisement including actual exposure to the advertisement during each segment and a time lagged effect of exposure to the advertisement during previous segments;

generate a marketing mix model that automatically accounts for the saturation effect without manual adjustment, by one or more users, of a response curve associated with the marketing mix model, the accounting for the saturation effect based on a regression analysis of the effective reach realized during the segments of the period of time for the advertising campaign and the marketing mix input data retrieved from the marketing mix inputs database; and

reduce a computational burden on the machine by eliminating tuning of response curves after the generation of the marketing mix model and determining a best fitting saturation curve corresponding to the relationship to be saved for use as an initial estimate in determining a second best fitting saturation curve for a second advertising campaign thereby increasing a computational efficiency for the determination of the second best fitting saturation curve.

10. The storage medium of claim 9, wherein the instructions further cause the machine to determine a first contribution of the advertising campaign to a lift in sales over the period of time based on the effective reach realized relative to a second contribution of another advertising campaign.

11. The storage medium of claim 9, wherein the determining of the relationship includes looking up a plausible saturation curve stored in a saturation curve database.

12. The storage medium of claim 9, wherein the instructions further cause the machine to determine an initial estimate of a penetration associated with the advertising campaign based on audience measurement data, the audience measurement data including at least one of demographic information corresponding to a target audience of the advertising campaign or media type information corresponding to a type of media associated with the advertising campaign.

13. The storage medium of claim 9, wherein the instructions further cause the machine to determine an initial estimate of a penetration associated with the advertising campaign and an initial estimate of an effective frequency associated with the advertising campaign based on at least one of a geographic region associated with the advertising campaign, a type of product or service associated with the advertising campaign, a comparable penetration associated with a comparable campaign for related or competing products, a comparable effective frequency associated with the comparable campaign, or a complexity of the advertising campaign.

14. The storage medium of claim 9, wherein the instructions further cause the machine to:

determine a range of plausible penetrations based on an initial estimate of a penetration associated with the advertising campaign;

determine a range of plausible effective frequencies based on an initial estimate of an effective frequency associated with the advertising campaign;

calculate a range of plausible saturation curves based on the range of plausible penetrations and the range of plausible effective frequencies; and

analyze the range of plausible saturation curves to identify the best fitting saturation curve with respect to the marketing mix input data associated with the advertising campaign.

15. The storage medium of claim 9, wherein the instructions further cause the machine to:

generate a simulated advertising campaign based on an initial estimate of a penetration associated with the advertising campaign and an initial estimate of an effective frequency associated with the advertising campaign;

determine data points indicative of simulated reach against simulated gross rating points associated with the simulated advertising campaign; and

calculate an equation that fits the data points, the equation corresponding to the relationship.