Method and system for conducting online marketing research in a controlled manner

• US 8,234,146 B2
• Filed: 03/06/2008
• Issued: 07/31/2012
• Est. Priority Date: 12/17/2002
• Status: Expired due to Fees

• Current Assignee: International Business Machines Corporation
• Original Assignee: International Business Machines Corporation
• Inventors: Mittal, Parul A., Jain, Vivek, Modani, Natwar, Kothari, Ravi, Agrawal, Aseem
• Primary Examiner(s): Sterrett, Jonathan G
• Application Number:
  US12/043,245
• Publication Number:
  US 20080154707A1
• Time in Patent Office: 1,608 Days
• Field of Search: 705/7.29, 705/14.21
• US Class Current: 705/7.29
• CPC Class Codes: G06Q 30/02 Marketing; Price estimation or determination; Fundraising
• CPC Class Codes: G06Q 30/0201 Market modelling; Market analysis; Collecting market data
• CPC Class Codes: G06Q 30/0203 Market surveys; Market polls
• CPC Class Codes: G06Q 30/0204 Market segmentation
• CPC Class Codes: G06Q 30/0219 based on funds or budget

Abstract

The invention describes a method and system for conducting online marketing research keeping in consideration the specified budget for the experiment. The invention describes a methodology for effective data collection and optimised utilization of budget through the use of efficient sampling and grouping of users.

Claims

1. A computer-implemented method for conducting controlled online marketing research with budgetary control, said method comprising:

receiving user selection criteria, by a computer, for entry of a user into an online experiment, when said user visits an online site, where said online experiment is conducted;

determining, by said computer, that said user selection criteria, for entry of said user, match user selection criteria for one of a plurality of existing experimental groups and an existing control group stored in said computer system, said user selection criteria comprising;

tracking said user'"'"'s online browsing to match browsing events;

determining whether said user is a new user or an existing user in one of said plurality of existing experimental groups and said control group;

a number of users for each of said existing plurality of experimental groups and said control group;

an experimental time period for said online experiment; and

a budget for said online experiment, based on probabilistic costs of future user actions implemented by prediction tools;

determining, by said computer, that said user satisfies said user selection criteria for entry into said online experiment, matching said user to a previous participant of said online experiment, assigning said user to one of said plurality of existing experimental groups and said control group, other than that to which said previous participant belongs, entering said user into said online experiment, and conducting said online experiment; and

outputting, by said computer, an analysis of a change in a marketing variable with respect to a response variable given by said user measured during the experimental time period for said online experiment, said marketing variable comprising any of a price of product, a discount, an advertisement, a bundling of products, and a product design or packaging.

2. The method as claimed in claim 1, wherein each ideal representative profile selected for comparison must have at least one vacancy in the experimental or control group(s).

3. The method as claimed in claim 1, wherein the current user is assigned one of the available experimental or control groups such that there is minimal disparity among the relative size of the groups.

4. The method as claimed in claim 1, wherein the number of users assigned to an experiment or control group(s) for a given ideal representative profile is more than or equal to one.

5. The method as claimed in claim 1, wherein the ideal representative profiles are re-computed periodically and steps b(ii), (c), and (d) are repeated for the new set of users.

6. The method as claimed in claim 1, wherein for the remaining time period being less than a threshold, the threshold distance of users from the ideal representative profiles is increased.

7. A computer system for conducting controlled online marketing research with budgetary control, comprising:

a memory that stores user selection criteria for entry of a user into an online experiment; and

a processor that;

determines said user selection criteria, when said user visits on online site where said online experiment is conducted, match user selection criteria for one of an existing plurality of experimental groups and a control group, said user selection criteria comprising;

tracking said user'"'"'s online browsing to match browsing events;

determining whether said user is a new user or an existing user in one of said existing plurality of experimental groups and said control group;

a number of users for each of said existing plurality of experimental groups and said control group;

an experimental time period for said online experiment; and

a budget for said online experiment, based on probabilistic costs of future user actions implemented by prediction tools;

determines that said user satisfies said user selection criteria for entry into said online experiment, matching said user to a previous participant of said online experiment, assigning said user to one of said existing plurality of experimental groups and said control group, other than that to which said previous participant belongs, entering said user into said online experiment, and conducting said online experiment; and

outputs an analysis of a change in a marketing variable with respect to a response variable given by said user measured during the experimental time period for said online experiment, said marketing variable comprising any of a price of product, a discount, an advertisement, a bundling of products, and a product design or packaging.

8. The system as claimed in claim 7, wherein the representative profile generator is a computer implemented tool.

9. The system as claimed in claim 7, wherein the representative profile generator generates the ideal representative profiles periodically.

10. A non-transitory computer storage device containing computer readable programs code for causing a computer to perform a method for conducting controlled online marketing research with budgetary control, said method comprising:

receiving user selection criteria for entry of a user into an online experiment, when said user visits an online site, where said online experiment is conducted;

determining that said user selection criteria, for entry of said user, match user selection criteria for one of an existing plurality of experimental groups and a control group stored in said computer system, said user selection criteria comprising;

tracking said user'"'"'s online browsing to match browsing events;

determining whether said user is a new user or an existing user in one of said existing plurality of experimental groups and said control group;

a number of users for each of said existing plurality of experimental groups and said control group;

an experimental time period for said online experiment; and

a budget for said online experiment, based on probabilistic costs of future user actions implemented by prediction tools;

determining that said user satisfies said user selection criteria for entry into said online experiment, matching said user to a previous participant of said online experiment, assigning said user to one of said existing plurality of experimental groups and said control group, other than that to which said previous participant belongs, entering said user into said online experiment, and conducting said online experiment; and

outputting an analysis of a change in a marketing variable with respect to a response variable given by said user measured during the experimental time period for said online experiment, said marketing variable comprising any of a price of product, a discount, an advertisement, a bundling of products, and a product design or packaging.

11. The non-transitory computer program storage device as claimed in claim 10, each ideal representative profile being selected for comparison has at least one vacancy in the experimental or control group(s).

12. The non-transitory computer program storage device as claimed in claim 10, including computer readable program code means configured for assigning the current to one of the available experimental or control groups such that there is minimal disparity among the relative size of the groups.

13. The non-transitory computer program storage device as claimed in claim 10, including computer readable program code means configured for assigning more than 1 user to an experiment or control group(s) for a given ideal representative profile.

14. The computer non-transitory computer program storage device as claimed in claim 10, including computer readable program code means configured for setting the smart sampling parameters includes threshold distance, expected arrival rate of users, a spatial distribution of users along the user feature space and the expected response rate of the individual users as he desired parameters.

15. The non-transitory computer program storage device as claimed in claim 10, including computer readable program code means configured for recomputing the ideal representative profiles periodically and repeating steps (b)11, c, and (d) for the new set of users.

16. The non-transitory computer program storage device as claimed in claim 10, including computer readable program code means configured for dynamically adjusting the smart sampling based on the time period remaining for the experiment to be completed.