Method and System for Conducting Online Marketing Research in a Controlled Manner
First Claim
Patent Images
1. A method for conducting controlled online marketing research with budgetary control comprising the steps of:
- a. identifying and specifying the experiment parameters,b. implementing smart sampling to determine whether the current user should be selected for the market research by;
(i) determining the ideal representative user profiles so as to maximise the expected information gain,(ii) comparing the current user'"'"'s profile with existing ideal representative user profiles, wherein a set of users with similar profiles are associated with the same ideal representative profile,c. assigning the selected user to any one of the available experimental or control groups,d. administering the experiment to the selected user if the specified budget is sufficient,e. repeating steps b(ii), (c), and (d) until the specified budget is expended and/or a termination condition is satisfied, andf. analyzing and reporting the results,wherein the smart sampling technique uses any known active learning algorithm or a distance metric selected from a set of metrics.
0 Assignments
0 Petitions
Accused Products
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 utilisation of budget through the use of efficient sampling and grouping of users.
39 Citations
33 Claims
-
1. A method for conducting controlled online marketing research with budgetary control comprising the steps of:
-
a. identifying and specifying the experiment parameters, b. implementing smart sampling to determine whether the current user should be selected for the market research by; (i) determining the ideal representative user profiles so as to maximise the expected information gain, (ii) comparing the current user'"'"'s profile with existing ideal representative user profiles, wherein a set of users with similar profiles are associated with the same ideal representative profile, c. assigning the selected user to any one of the available experimental or control groups, d. administering the experiment to the selected user if the specified budget is sufficient, e. repeating steps b(ii), (c), and (d) until the specified budget is expended and/or a termination condition is satisfied, and f. analyzing and reporting the results, wherein the smart sampling technique uses any known active learning algorithm or a distance metric selected from a set of metrics. - View Dependent Claims (2, 3, 4, 5, 6, 7)
-
-
8. A method for conducting controlled online marketing research with budgetary control comprising the steps of:
-
a. identifying and specifying the experiment parameters, b. comparing the current user'"'"'s profile with existing representative user profiles, wherein a set of users associated with the representative profile have one to one correspondence, c. creating a new representative user profile with current user'"'"'s profile, for user that does not match with any of the existing representative user profiles or associating the current user with the closest representative user profile, d. assigning the current user to any one of the available experimental or control groups, e. administering the experiment to the selected user if the specified budget is sufficient, f. repeating steps (b), (c), (d) and (e) until the specified budget is expended and/or a termination condition is satisfied, and g. analyzing and reporting the results, 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. - View Dependent Claims (9, 10, 11, 12)
-
-
13. A system for conducting controlled online marketing research with budgetary control comprising:
-
a. parameter specifier that identifies the experiment parameters, b. representative profile generator that determines the ideal representative user profiles so as to maximise the expected information gain, c. smart sampler to determine whether the current user should be selected for the market research by; (i) determining the ideal representative user profiles so as to maximise the expected information gain, (ii) comparing the current user'"'"'s profile with existing ideal representative user profiles, wherein a set of users with similar profiles are associated with the same ideal representative profile, d. assignor that assigns the current user to any one of the available experimental or control groups, e. administering means that administers the experiment to the selected user if the specified budget is sufficient, f. analyzer that analyzes and reports the results, wherein the smart sampler is a computer implemented tool that uses any known active learning algorithm or a distance metric selected from a set of predetermined metrics. - View Dependent Claims (14, 15, 16)
-
-
17. A system for conducting controlled online marketing research with budgetary control comprising of:
-
a. parameter specifier for specifying the experiment parameters, b. user selector for comparing the current user'"'"'s profile with existing representative user profiles, wherein a set of users associated with the representative profile have one to one correspondence, c. representative profile generator for creating a new representative user profile with the current user'"'"'s profile, for users that do not match with any of the existing representative user profiles or associating the current user with the closest representative user profile, d. assignor for assigning the current user to any one of the available experimental or control groups, e. administering means for administering the experiment to the selected user if the specified budget is sufficient, f. analyzer for analyzing and reporting the results, wherein the representative profile generator uses an algorithm that dynamically determines whether a new representative profile would be created for the current user or not. - View Dependent Claims (18, 19, 20, 21)
-
-
22. A computer program product residing on computer readable media containing computer readable programs code for causing a computer to conduct controlled online marketing research with budgetary control comprising:
-
a. computer readable program code means configured for identifying and specifying the experiment parameters, b. computer readable program code configure for implementing smart sampling for determining whether the current user should be selected for the market research by; (i) determining the ideal representative user profiles so as to maximise the expected information gain, (ii) comparing the current user'"'"'s profile with existing ideal representative user profiles, wherein a set of users with similar profiles are associated with the same ideal representative profile, c. computer readable program code means configured for assigning the selected user to any one of the available experimental or control groups, d. computer readable program code means configured for administering the experiment to the selected user if the specified budget is sufficient, e. computer readable program code means configured for repeating steps (b) ii, (c), and (d) until the specified budget is expended and/or the required data is collected, and f. computer readable program code means configured for analyzing and reporting the results, g. 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 the desired parameters. - View Dependent Claims (23, 24, 25, 26, 27, 28, 29)
-
-
30. A computer program product for conducting controlled online marketing research with budgetary control comprising:
-
a. computer readable program code means configured for identifying and specifying the experiment parameters, b. computer readable program code means configured for comparing the current user'"'"'s profile with existing representative user profiles, wherein a representative profile is associated with a set of users whose profiles have one to one correspondence, c. computer readable program code means configured for creating a new representative user profile with current user'"'"'s profile, for user that do not match with any of the existing representative user profiles or associating the current user with the closest representative user profile, d. computer readable program code means configured for assigning the current user to any one of the available experimental or control groups, e. computer readable program code means configured for administering the experiment to the selected user if the specified budget is sufficient, f. computer readable program code means configured for repeating steps (b), (c), (d) and (e) until the specified budget is expended and/or the required data is collected, and g. computer readable program code means configured for analyzing and reporting the results, h. including computer readable program code means configured for assigning the current user to one of the available experimental or control groups such that there is minimal disparity among the relative size of the groups. - View Dependent Claims (31, 32, 33)
-
Specification