WO2017210786A1

WO2017210786A1 - Method and system for predictive modelling in advertising campaigns

Info

Publication number: WO2017210786A1
Application number: PCT/CA2017/050692
Authority: WO
Inventors: Harvir S. Bansal; Walter J. Ramdeholl; Avik Halder; Dwaipayan SINHA
Original assignee: B3Intelligence Ltd.
Priority date: 2016-06-06
Filing date: 2017-06-06
Publication date: 2017-12-14

Abstract

A computer-implemented method for determining the performance of at least one advertising campaign, the method having the steps of: estimating a contribution of at least one media channel associated with said at least one advertising campaign in generating lift in brand consciousness parameters; predicting a frequency of exposure in said at least one different media channel for which said lift is at its optimum; predict a budget for said at least one different media channel for which said lift is at its optimum; and determine the suitability of a current budget for each of said at least one different media channel associated with said campaign to generate said lift.

Description

METHOD AND SYSTEM FOR PREDICTIVE MODELLING

IN ADVERTISING CAMPAIGNS

FIELD OF THE INVENTION

[0001] The present invention relates to advertising campaigns, more particularly it relates to attribution analysis and predictive modelling methods.

BACKGROUND

[0002] Major brands frequently run advertising campaigns to promote their products and services. To determine the impact of such campaigns the brands conduct surveys to quantitatively estimate the increase in the recognition of their brand, generally known as "lift". Brand consciousness parameters, such as brand metrics and brand perceptions, are used to determine the impact of the campaigns on the target consumer or buyer. The brand metrics generally comprise brand awareness, likelihood to consider buying, brand familiarity, brand favorability, and likelihood to recommend, among others. Brand perceptions, on the other hand, represent different opinions about the brand, such as "I love this brand", "My friends recommend this brand", and "I would like to discuss this brand with some professional", and so forth.

[0003] The surveys for estimating lifts in these brand consciousness parameters are typically conducted on two groups of respondents, that is, a control group composed of respondents who have not been exposed to the brand campaign, and an exposed group composed of respondents who have been exposed to the brand campaign. These groups may be further divided into subgroups associated with a particular media channel, and are not necessarily mutually exclusive. As an example, if the proportion of the brand conscious people in the exposed group is higher than the control group, then there is a positive lift. However, negative lifts are also possible, and various reasons may be attributed to this, such as, overexposure to advertisements leading to a negative impact in the likelihood to recommend the brand.

[0004] While many approaches have been proposed to determine the impact of advertising campaigns on brands, none of them have been able to accurately and efficiently predict the impact of a particular campaign.

[0005] It is thus an object of the present invention to mitigate or obviate at least one of the above-mentioned disadvantages. SUMMARY OF THE INVENTION

[0006] In one of its aspects, there is provided a computing system for determining the performance of at least one advertising campaign, said system comprises instructions in data storage executable by a processor, said system comprising:

a media attribution analysis module comprising a first set of instructions executable by said processor to estimate a contribution of at least one media channel associated with said at least one advertising campaign in generating a lift in brand consciousness parameters;

a lift-frequency predictive modeller comprising a second set of instructions executable by said processor to predict a frequency of exposure in said at least one different media channel for which said lift is at its optimum;

a lift-budget predictive modeller comprising a third set of instructions executable by said processor to predict a budget for said at least one different media channel for which said lift is at its optimum; and

a budget efficiency analysis module comprising a fourth set of instructions executable by said processor to determine the suitability of a current budget for each of said at least one different media channel associated with said campaign to generate said lift.

[0007] In another of its aspects, there is provided a computer-implemented method for determining the performance of at least one advertising campaign, the method having the steps of: estimating a contribution of at least one media channel associated with said at least one advertising campaign in generating a lift in brand consciousness parameters;

predicting a frequency of exposure in said at least one different media channel for which said lift is at its optimum;

predicting a budget for said at least one different media channel for which said lift is at its optimum; and

determining the suitability of a current budget for each of said at least one different media channel associated with said campaign to generate said lift.

[0008] In another of its aspects, there is provided a computer-implemented method for determining an impact of an advertisement associated with a brand, said brand being associated with brand consciousness parameters, the method having the steps of: at first processing device, selecting a control group of respondents having no access to said advertisement and selecting an exposed group of respondents for engagement with said advertisement;

generating a survey with questions associated with brand consciousness parameters; associating said advertisement with an identifier;

presenting said advertisement to said exposed group;

causing said exposed group to answer said survey to generate exposed group survey data; generating control group survey data from responses to said survey by said control group not exposed to said advertisement;

receiving said exposed group survey data and said control group survey data and generating a survey dataset; and fitting a regression model for said brand consciousness parameters as independent variables, and demographic variables from said survey dataset as covariates to compute a lift corresponding to said impact of said advertisement on awareness of said brand.

[0009] In another of its aspects, there is provided at a processing device, a computer- implemented method for determining an impact of an advertisement associated with a brand, said brand being associated with brand consciousness parameters, the method having the steps of: selecting a control group of respondents and an exposed group of respondents; and generating a survey with questions associated with brand consciousness parameters;

transmitting said advertisement from said processing device to an apparatus for presentation of said advertisement to said exposed group;

automatically determining a frequency of exposure to said advertisement for each member of said exposed group, and when said frequency of exposure exceeds a predetermined threshold to a member of said exposed group, then said member is classified as an outlier and removed from said exposed group;

computing an average frequency of exposure to said exposed group;

generating data to facilitate construction of a model for brand consciousness parameters; generating a predictive model from said data, whereby said predictive model predicts a frequency of exposure in said at least one different media channel for which a lift corresponding to said impact of said advertisement is at its optimum; and

plotting, on a display, said lift against said frequency of exposure. [0010] In another of its aspects, there is provided a computer-implemented method for determining a budget for which a lift corresponding to said impact of said advertisement associated with a brand is optimal, said brand being associated with brand consciousness parameters, the method having the steps of:

in said processing device, selecting a control group of respondents and an exposed group of respondents; and generating a survey with questions associated with brand consciousness parameters;

computing an average frequency of exposure to said exposed group;

generating data to facilitate construction of a model for brand consciousness parameters; generating a predictive model from said data, whereby said predictive model predicts a frequency of exposure in said at least one different media channel for which said lift corresponding to said impact of said advertisement is at its optimum;

plotting, on a display, said lift against said frequency of exposure; and

linking said frequency of exposure to said budget.

[0011] In another of its aspects, there is provided an apparatus comprising:

a processing system; and

one or more computer readable storage media including program instructions stored on the one or more computer readable media that, when executed by the processing system, direct the processing system to at least:

select a control group of respondents having no access to said advertisement and select an exposed group of respondents for engagement with said advertisement;

generate a survey with questions associated with brand consciousness parameters;

associate said advertisement with an identifier; present said advertisement to said exposed group;

cause said exposed group to answer said survey to generate exposed group survey data; generate control group survey data from responses to said survey by said control group not exposed to said advertisement;

receiving said exposed group survey data and said control group survey data and generate a survey dataset; and fitting a regression model for said brand consciousness parameters as independent variables, and demographic variables from said survey dataset as covariates to compute a lift corresponding to said impact of said advertisement on awareness of said brand.

[0012] In another of its aspects, there is provided an apparatus for determining an impact of an advertisement associated with a brand, said brand being associated with brand consciousness parameters, the apparatus comprising:

a processing system; and

select a control group of respondents and an exposed group of respondents;

transmit said advertisement exposed group of respondents;

automatically determine a frequency of exposure to said advertisement for each member of said exposed group, and when said frequency of exposure exceeds a predetermined threshold to a member of said exposed group, then said member is classified as an outlier and removed from said exposed group;

compute an average frequency of exposure to said exposed group;

generate data to facilitate construction of a model for said brand consciousness parameters;

generate a predictive model from said data, whereby a frequency of exposure in said at least one different media channel for which said lift is at its optimum is predicted.

[0013] Advantageously, the methods and systems provide analyses capable of determining the best media for promoting the brand, and can also predict the optimal budget and frequency of exposure and identify the most budget efficient media channel. Accordingly, these methods and systems assist a brand to recognize whether it should assign additional resources to an ongoing advertising ("ad") campaign for a desirable positive lift, or reduce the exposure of the ad campaign to avoid brand saturation, or negative impacts on popular opinion, thereby resulting in negative lift.

BRIEF DESCRIPTION OF THE DRAWINGS

[0014] Several exemplary embodiments of the present invention will now be described, by way of example only, with reference to the appended drawings in which:

[0015] Figure 1 shows an exemplary computing system;

[0016] Figure 2 shows an exemplary environment in which a method and system for predictive modelling operate;

[0017] Figure 3 shows a high level flow diagram illustrating exemplary process steps for media attribution analysis;

[0018] Figure 4 is an exemplary chart illustrating the contribution of five different channels to brand favorability;

[0019] Figure 5 shows a high level flow diagram illustrating exemplary process steps for lift- frequency predictive modelling;

[0020] Figure 6 shows an exemplary chart of lift-frequency analysis of some brand perceptions;

[0021] Figure 7 shows a high level flow diagram illustrating exemplary process steps for lift- budget predictive modelling;

[0022] Figure 8 shows a budget efficiency chart;

[0023] Figure 9 shows a high level flow diagram illustrating exemplary process steps for budget efficiency analysis;

[0024] Figure 10 shows an exemplary budget efficiency analysis; and

[0025] Figure 11 shows a magnitude-efficiency matrix.

DETAILED DESCRIPTION OF EXEMPLARY EMBODF ENTS

[0026] Various embodiments of the disclosure are discussed in detail below. While specific implementations are discussed, it should be understood that this is done for illustration purposes only. A person skilled in the relevant art will recognize that other components and configurations may be used without parting from the spirit and scope of the disclosure.

[0027] With reference to Figure 1 , exemplary computing system 10 includes a general-purpose computing device, including processing unit (CPU or processor) 12 and system bus 11 that couples various system components including system memory 13 such as read only memory (ROM) 14 and random access memory (RAM) 15 to processor 12. System 10 can include cache 16 of high speed memory connected directly with, in close proximity to, or integrated as part of processor 12. System 10 copies data from memory 13 and/or storage device 18 to the cache 16 for quick access by processor 12. In this way, the cache provides a performance boost that avoids processor 12 delays while waiting for data. These and other modules can control or be configured to control processor 12 to perform various actions. Other system memory 13 may be available for use as well. Memory 13 can include multiple different types of memory with different performance characteristics. It can be appreciated that the methods and system may operate on computing device 10 with more than one processor 12 or on a group or cluster of computing devices networked together to provide greater processing capability. Processor 12 can include any general purpose processor and a hardware module or software module, such as module 1 20a, module 2 20b, and module 3 20c stored in storage device 18, configured to control processor 12 as well as a special-purpose processor where software instructions are incorporated into the actual processor design. Processor 12 may essentially be a completely self-contained computing system, containing multiple cores or processors, a bus, memory controller, cache, etc. A multi-core processor may be symmetric or asymmetric.

[0028] System bus 11 may be any of several types of bus structures including a memory bus or memory controller, a peripheral bus, and a local bus using any of a variety of bus architectures. A basic input/output (BIOS) stored in ROM 14 or the like, may provide the basic routine that helps to transfer information between elements within computing device 10, such as during start-up. Computing device 10 further includes storage devices 18 such as a hard disk drive, a magnetic disk drive, an optical disk drive, a solid state drive, a tape drive or the like. Storage device 18 can include software modules 20a, 20b, 20n for controlling processor 12. Other hardware or software modules are contemplated. The storage device 18 is connected to system bus 11 by a drive interface. The drives and the associated computer readable storage media provide non-volatile storage of computer readable instructions, data structures, program modules and other data for computing device 10. In one aspect, a hardware module that performs a particular function includes the software component stored in a non-transitory computer-readable medium in connection with the necessary hardware components, such as processor 12, bus 11, display 22, and so forth, to carry out the function. The basic components are known to those of skill in the art and appropriate variations are contemplated depending on the type of device, such as whether device 10 is a handheld computing device, a desktop computer, or a computer server.

[0029] Although the exemplary embodiment described herein employs hard disk 18, it should be appreciated by those skilled in the art that other types of computer readable media which can store data that are accessible by a computer, such as magnetic cassettes, flash memory cards, digital versatile disks, cartridges, random access memories (RAMs) 15, read only memory (ROM) 14, a cable or wireless signal containing a bit stream and the like, may also be used in the exemplary operating environment. Non-transitory computer-readable storage media expressly exclude media such as energy, carrier signals, electromagnetic waves, and signals per se.

[0030] To enable user interaction with computing device 10, input device 24 represents any number of input mechanisms, such as a microphone for speech, a touch-sensitive screen for gesture or graphical input, keyboard, mouse, motion input, speech and so forth. Output device 22 can also be one or more of a number of output mechanisms known to those of skill in the art. In some instances, multimodal systems enable a user to provide multiple types of input to communicate with computing device 10. Communications interface 26 generally governs and manages the user input and system output. There is no restriction on operating on any particular hardware arrangement and therefore the basic features here may easily be substituted for improved hardware or firmware arrangements as they are developed.

[0031] For clarity of explanation, the illustrative system embodiment is presented as including individual functional blocks, including functional blocks labeled as a "processor" or processor 12. The functions these blocks represent may be provided through the use of either shared or dedicated hardware, including, but not limited to, hardware capable of executing software and hardware, such as processor 12, that is purpose-built to operate as an equivalent to software executing on a general purpose processor. For example, the functions of one or more processors, presented in Figure 1, may be provided by a single shared processor or multiple processors. (Use of the term "processor" should not be construed to refer exclusively to hardware capable of executing software.) Illustrative embodiments may include microprocessor and/or digital signal processor (DSP) hardware, read-only memory (ROM) 14 for storing software performing the operations discussed below, and random access memory (RAM) 15 for storing results. Very large scale integration (VLSI) hardware embodiments, as well as custom VLSI circuitry in combination with a general purpose DSP circuit, may also be provided. [0032] The logical operations of the various embodiments are implemented as: (1) a sequence of computer implemented steps, operations, or procedures running on a programmable circuit within a general use computer, (2) a sequence of computer implemented steps, operations, or procedures running on a specific-use programmable circuit; and/or (3) interconnected machine modules or program engines within the programmable circuits. System 10, shown in Figure 1, can practice all or part of the recited methods, can be a part of the recited systems, and/or can operate according to instructions in the recited non-transitory computer-readable storage media. Such logical operations can be implemented as modules configured to control processor 12 to perform particular functions according to the programming of the module. For example, Figure 1 illustrates three modules 20a, 20b and 20n which are modules configured to control processor 12. These modules 20a, 20b and 20n may be stored on the storage device 18 and loaded into RAM 15 or memory 13 at runtime or may be stored, as would be known in the art, in other computer-readable memory locations.

[0033] Computer system 10 can be of varying types including a workstation, server, computing cluster, blade server, server farm, or any other data processing system or computing device. Due to the ever-changing nature of computers and networks, the description of computer system 10 depicted in Figure 1 is intended only as a specific example for purposes of illustrating some implementations. Many other configurations of computer system 10 are possible having more or fewer components than the computer system depicted in Figure 1.

[0034] Figure 2 shows a top-level component architecture diagram of an exemplary environment, generally identified by reference numeral 30, for which the methods and systems for predictive modelling operate. As shown, Figure 2 illustrates environment 30, in which a user interacts with computing system 32, such as an application server, through user computer 34 communicatively coupled thereto via communication medium 35, or network, e.g., the Internet, and/or any other suitable network. The computers of environment 30 comprise the features of the general-purpose computing device 10, as described above, and may include, but are not limited to: a mini computer, a handheld communication device, e.g. a tablet, a mobile device, a smart phone, a smartwatch, a wearable device, a personal computer, a server computer, a series of server computers, and a mainframe computer. Application server 32 is associated with one or more databases 36, which may be any type of data repository or combination of data repositories, which store records or other representations of data associated with the surveys, ad campaigns, brand metrics, and so forth.

[0035] Application server 32 comprises brand analysis engine 38 for determining the performance of ad campaigns. Brand analysis engine 38 comprises a plurality of modules, such as, media attribution analysis module 40, lift-frequency predictive modeller 42, lift-budget predictive modeller 44 and budget efficiency analysis module 46. Generally, media attribution analysis module 40 comprises instructions in data storage 18, executable by processor 12 to cause processor 12 to estimate the contribution of different media channels involved in an ad campaign in generating lift in the brand consciousness parameters through various techniques, such as regression analysis. Lift-frequency predictive modeller 42 comprises instructions in data storage 18, executable by processor 12 to cause processor 12 to predict the frequency of exposure in media for which the lift is at its optimum, such as digital media. Lift-budget predictive modeller 44 comprises instructions in data storage 18, executable by processor 12 to cause processor 12 to predict the budget for media channels for which lift is at its optimum, such as digital media channels. Budget efficiency analysis module 46 comprises instructions in data storage 18, executable by processor 12 to cause processor 12 to determine the suitability of the current budget for each of the different media channels involved in the campaign to generate lift.

[0036] It should be understood that brand analysis engine 38 as depicted is merely provided for illustrative purposes and may have more, or less modules, and the modules may vary in their functionality or in how the functionality is implemented. One or more of the components and/or one or more additional components of the example environment of Figure 2 may each include memory for storage of data and software applications, a processor for accessing data and executing applications, and components that facilitate communication over a network. In some implementations, the components may include hardware that shares one or more characteristics with exemplary computer system 10 that is illustrated in Figure 1.

[0037] It should be noted that although application server 32 has been described as having brand analysis engine 38 comprising a plurality of modules, such as, media attribution analysis module 40, lift-frequency predictive modeller 42, lift-budget predictive modeller 44 and budget efficiency analysis module 46, user computer 34 may include brand analysis engine 38 having a plurality of modules, such as, media attribution analysis module 40, lift-frequency predictive modeller 42, lift-budget predictive modeller 44 and budget efficiency analysis module 46, configured to operate as a stand-alone solution. Accordingly, brand analysis engine 38 may be included as an add-on to an existing platform to provide the above-noted functionality.

[0038] Referring to Figure 3, an exemplary flowchart of an overview of exemplary process steps for a media attribution analysis method is shown. The method comprises the steps of associating an advertisement corresponding to a campaign with a unique identifier, and/or campaign with a unique identifier, and receiving survey data (step 100), and fitting regression models for brand metrics and perceptions as the independent variables and the demographic variables from the survey data as covariates (step 102). While demographic variables are survey- specific, the most common demographic variables are age, gender, household/personal income range, etc. In some surveys, the demographic variables are usually coded as categorical variables, and there may be exceptions such as the age information, which may be continuous data, but that can also be coded as categorical data using different ranges. As an example, categorical variables that have more than two levels, are recoded into two levels, where the top categorical variable or top two categorical variables are considered as a positive response and coded as "1", and the rest are coded as "0" (step 104). Generally, however, the top categorical variable is recoded for 3-level questions, and top two categorical variables are recoded for 5 levels or more. Next, regression modelling is performed (step 106), and may employ a logistic regression model defined as:

[0039] where p is the probability for a respondent to give a positive answer when asked an attitudinal question about the brand, and k is the highest level of the demographic factor i. kj is the h ighest level of the demograph ic factor ί^' .

Ί if for ^'tli demographic factor the level is /

L 0 otherwise

1 if the respondent belongs to the exposed group

0 if the respondent belongs to the control group

0 is th e intercept and a and β are the coefficients of the regression model. [0040] In step 108, the contribution to lift is computed by employing the following formula in which p is defined as: = ST

¹ 1 + e^X?

where X is the vector of all the variables involved in the regression (x's and y's from the logistic regression equation) and β is the estimate of the coefficient β.

[0041] Thus the contribution to lift by a certain media channel is p(y = 1 | other covariates fixed) - p(y\ = 0 | other covariates fixed).

[0042] Next, the computed contributions to lift are converted into percentage contribution (step 1 10), so that the total combination of all media channels for each brand metric and brand perception is 100%, in order to identify the strongest driver of lift for the brand metric and brand perception (step 1 12). An exemplary chart illustrating the contribution of five different channels to brand favorability is shown in a chart of Figure 4, in which the interaction of television (TV), outdoor billboards and digital media had the biggest contribution to the lift of brand favorability.

[0043] Now looking at Figure 5, there is shown a high level flow diagram illustrating exemplary process steps for a method for lift-frequency predictive modelling. Accordingly, the method comprises the steps of associating an advertisement corresponding to a campaign with an identifier and/or campaign with a unique identifier, tracking any digital media used in a campaign (step 200), and the number of times an individual is exposed to the advertisement of the brand in consideration is quantified, and the frequency of exposure is determined (step 202). While some individuals may actually be exposed a large number of times, even 1 ,000 to 2,000 times, the number of such individuals is substantially small, and therefore such individuals are considered to be anomalous, or outliers. After removing the outliers (step 204), the average frequency of exposure (F_avg) is computed (step 206) as

where fi is the frequency of exposure, and Wi is the sample size for the frequency of exposure .

[0044] Once the sample sizes for each frequency of exposure with the outliers removed are obtained, non-parametric models are deployed for estimating the linkage between frequency and lifts (step 208). Thus the analysis is data-driven, rather than fitting data on a-priori assumptions of functional forms. In one example, polynomial splines (B-Spline) are employed to analyze the data and provide an approximation of the function which describes the link between frequency and lifts. Generally, a model is created for each of the brand metrics and brand perceptions.

[0045] From the B-Spline curves, a point representing the maximum lift for a brand metric and brand perception is determined (step 210). If a curve shows the same maximum lift for two different frequencies, the lower one is chosen to be the optimal frequency for that brand metric and brand perception. However, the global optima for each curve may occur at different frequencies. In that case, the frequency where most of the brand metrics and brand perceptions achieve maxima or near maxima together is considered the optimal one. Figure 6 shows an exemplary chart of lift-frequency analysis of some brand perceptions, in which the current frequency of exposure is at 5, and the optimal frequency for three brand perceptions (("Has knowledgeable sales associates" (curve E), "Offers products that help me perform at my best" (curve B), "Offers the highest quality products" (curve A)) is 8, and for the other three brand perceptions ("Offers something different than other retailers" (curve D), "Provides great value for my money" (curve C), "Offers products that my kids want" (curve F)), the optimal frequency is about 1.3, or approximated to 1, as frequency cannot be fractional.

[0046] Figure 7 shows a high level flow diagram illustrating exemplary process steps for lift- budget predictive modelling, in which the distribution of the frequency of exposure of an uniquely identified advertisement is employed to build a predictive model to find the optimal budget. The budget is linked to the frequency of exposure by modelling the random distribution of the frequency (step 300), and then linking the budget to a distribution parameter (step 302). In one example, a logarithmic distribution defines the frequency distribution which is a single parameter distribution that can be tied to the budget, in which a connection μ is employed to estimate single parameter p in the logarithmic model, that is

ln( l - p) l -

[0047] Changes in the budget are used to generate frequency distributions for those budget levels (step 304). These distributions are repeatedly sampled to determine the mean frequency levels for a particular budget level (step 306), and the upper and lower bounds of the frequency that can be used for calculating confidence intervals around the frequency (step 308), and consequently the lift. Assuming budge t_nei,v

= budgets ^{X avg}oid

[0048] the samples for the new frequency distribution are generated and the sample means are calculated for each new frequency distribution (step 310). Accordingly, the average of the sample means becomes the new value for the mean frequency. Using the method for lift-frequency predictive modelling, described above with reference to Figure 5, the model to predict average frequency for a given budget is now computed (step 312). Therefore, these two models are combined obtain a model to predict lift for a given budget. Figure 8 shows a budget analysis chart for the same campaign shown in Figure 6, for which the current spend is about $1,400,000. Some striking features of the chart include the observations that the budget curves look similar to the frequency curves of Figure 6, and after $4,500,000, all the curves become flat, showing diminishing returns with further increase of budget. Similar to the frequency analysis, three of the perceptions ("Has knowledgeable sales associates" (curve E), "Offers products that help me perform at my best" (curve B), "Offers the highest quality products" (curve A)), require a higher budget of around $4,000,000, and three others ("Offers something different than other retailers" (curve D), "Provides great value for my money" (curve C), "Offers products that my kids want" (curve F)), require a lower budget of $700,000.

[0049] The budget efficiency analysis is performed by following the exemplary process steps of Figure 9. In step 400, the magnitude of a media channel, defined as the lift for a particular brand metric/perception expressed as a percentage of the metric/perception proportion in the control group, is determined, that is

Lift for the brand metric/perception bv the channel

Magnitude of a channel = : _; ^"— : x 100%

Control metric/ perception prop ortion or the channel

[0050] For example, if 80% of the control group of TV favor a brand, and 98% of the exposed group for TV favor the brand, then the Magnitude of TV in Brand Favorability is (98- 80)/80=0.225 or 22.5%. Next, the efficiency of a channel, defined as magnitude of a channel per unit spend on the channel, is determined (step 402), that is

Magnitude of t e channel

Efficiency of a channel =

Unit spend of the channel [0051] If the unit spend is $1,000,000, and in the previous example the spend in TV is 2, 000,000, then the Efficiency of TV is 22.5/2=1 1.25%/million dollars. Next, the Efficiency Frontier™, defined as the average efficiency of all the media channels, is determined (step 404),

[0052] where Ei the efficiency for the i media channel, and N is the total number of channels. Figure 10 shows an exemplary budget efficiency analysis chart for a certain brand perception "Has knowledgeable sales associates". The Efficiency Frontier is 8.19%/million dollars, and three media channels are performing better than the average, while the others are performing poorly.

[0053] Figure 1 1 shows a magnitude-efficiency matrix in which the performances of each channel for each brand metric/perception can be assessed. The graph has two main axes, one at the average magnitude of all channel-metric (or channel-perception) combinations, and another one at the average efficiency. The axes divide the graph into four quadrants: 1 st Quadrant: High Magnitude - High Efficiency; 2nd Quadrant: Low Magnitude - High Efficiency; 3rd Quadrant: Low Magnitude - Low Efficiency; and 4th Quadrant: High Magnitude - Low Efficiency. The matrix shows which channel-metric (or channel-perception) combination performs the best, and combinations in the 1 st Quadrant are the best performers. Accordingly, the matrix shows that the interaction of Digital and Radio performs best for the brand perception "Provides great value for money", with the highest magnitude and efficiency. The matrix also contains pie charts, which show the proportion of the presence of each media channel quadrant-wise.

[0054] One or more of the components and/or one or more additional components of the example environment of Figure 2 may each include memory for storage of data and software applications, a processor for accessing data and executing applications, and components that facilitate communication over a network. In some implementations, the components may include hardware that shares one or more characteristics with the example computer system that is illustrated in Figure 1.

[0055] Database 36 may be, include or interface to, for example, the Oracle™ relational database sold commercially by Oracle Corp. Other databases, such as Informix™, DB2 (Database 2), Sybase or other data storage or query formats, platforms or resources such as OLAP (On Line Analytical Processing), SQL (Standard Query Language), a storage area network (SAN), Microsoft Access™ or others may also be used, incorporated or accessed in the invention. Alternatively, database 36 is communicatively coupled to application server 32.

[0056] Embodiments within the scope of the present disclosure may also include tangible and/or non-transitory computer-readable storage media for carrying or having computer- executable instructions or data structures stored thereon. Such non-transitory computer-readable storage media can be any available media that can be accessed by a general purpose or special purpose computer, including the functional design of any special purpose processor as discussed above. By way of example, and not limitation, such non-transitory computer-readable media can include RAM, ROM, EEPROM, CD-ROM or other optical disk storage, magnetic disk storage or other magnetic storage devices, solid state drives, or any other medium which can be used to carry or store desired program code means in the form of computer-executable instructions, data structures, or processor chip design. When information is transferred or provided over a network or another communications connection (either hardwired, wireless, or combination thereof) to a computer, the computer properly views the connection as a computer-readable medium. Thus, any such connection is properly termed a computer-readable medium. Combinations of the above should also be included within the scope of the computer-readable media.

[0057] Computer-executable instructions include, for example, instructions and data which cause a general purpose computer, special purpose computer, or special purpose processing device to perform a certain function or group of functions. Computer-executable instructions also include program modules that are executed by computers in stand-alone or network environments. Generally, program modules include routines, programs, components, data structures, objects, and the functions inherent in the design of special-purpose processors, etc. that perform particular tasks or implement particular abstract data types. Computer-executable instructions, associated data structures, and program modules represent examples of the program code means for executing steps of the methods disclosed herein. The particular sequence of such executable instructions or associated data structures represents examples of corresponding acts for implementing the functions described in such steps.

[0058] Certain embodiments described herein may be implemented as logic or a number of modules, engines, components, or mechanisms. A module, engine, logic, component, or mechanism (collectively referred to as a "module") may be a tangible unit capable of performing certain operations and configured or arranged in a certain manner In certain exemplary embodiments, one or more computer systems (e.g., a standalone, user, or server computer system) or one or more components of a computer system (e.g., a processor or a group of processors) may be configured by software (e.g., an application or application portion) or firmware (note that software and firmware can generally be used interchangeably herein as is known by a skilled artisan) as a module that operates to perform certain operations described herein.

[0059] Those of skill in the art will appreciate that other embodiments of the disclosure may be practiced in network computing environments with many types of computer system configurations, including personal computers, hand-held devices, multi-processor systems, microprocessor-based or programmable consumer electronics, network PCs, minicomputers, mainframe computers, and the like. Embodiments may also be practiced in distributed computing environments where tasks are performed by local and remote processing devices that are linked (either by hardwired links, wireless links, or by a combination thereof) through a communications network. In a distributed computing environment, program modules may be located in both local and remote memory storage devices.

[0060] The various embodiments described above are provided by way of illustration only and should not be construed to limit the scope of the disclosure. Those skilled in the art will readily recognize various modifications and changes that may be made to the principles described herein without following the example embodiments and applications illustrated and described herein, and without departing from the spirit and scope of the disclosure.

Claims

CLAIMS:

1. A computer-implemented method for determining an impact of an advertisement associated with a brand, said brand being associated with brand consciousness parameters, the method having the steps of:

at a processing device, selecting a control group of respondents having no access to said advertisement and selecting an exposed group of respondents for engagement with said advertisement;

presenting said advertisement to said exposed group;

2. The computer-implemented method of claim 1, wherein said brand consciousness parameters comprise at least one of brand metrics and brand perceptions.

3. The computer-implemented method of claim 2, wherein said at least one of brand metrics comprises at least one of brand awareness, brand favourability, brand familiarity,

recommendation intent and purchase intent.

4. The computer-implemented method of claim 3, wherein said advertisement is presented to said exposed group via at least one of a plurality of channels.

5. The computer-implemented method of claim 4, wherein said at least one of plurality of channels comprises television.

6. The computer-implemented method of claim 5, wherein said processing device determines a contribution of said television channel to a lift corresponding to said impact of said advertisement.

7. The computer-implemented method of claim 4, wherein said at least one of plurality of channels comprises digital.

8. The computer-implemented method of claim 7, wherein said processing device determines a contribution of said digital channel to said lift.

9. The computer-implemented method of claim 4, wherein said at least one of plurality of channels comprises out of home media (OOH), and wherein said processing device determines a contribution of said out of home media (OOH) channel to said lift.

10. The computer-implemented method of claim 4, wherein said at least one of plurality of channels comprises radio, and wherein said processing device determines a contribution of said radio to said lift.

11. The computer-implemented method of claim 4, wherein said lift is determined by combining lifts for each of said at least one of plurality of channels.

12. The computer-implemented method of any one of claims 6, 8, 9 and 10, wherein said lift corresponds to a combined lift due to said advertisement in said television, digital and out of home media (OOH) channels.

13. The computer-implemented method of claim 2, wherein said at processing device estimates a contribution of at least one media channel associated with at least one advertising campaign in generating said lift in said brand consciousness parameters.

14. The computer-implemented method of claim 13, wherein said at processing device comprises a memory device having embodied therein said exposed group survey data, control group survey data; and said processing device in communication with said memory device, said processor configured to receive said exposed group survey data and said control group survey data and generate a survey dataset; and estimate a contribution of at least one media channel associated with said at least one advertising campaign in generating said lift in brand

consciousness parameters.

15. The computer-implemented method of claim 14, wherein a logistic regression model is employed, said logistic regression model defined as:

in which p is the probability for said respondent to give a positive answer when asked an attitudinal question about said brand, and

kj is the h ighest level of the demographic factor i .

f^' l if for fth demographic factor the level is _j

ί 0 otherwise

Ί if the respondent belongs to the exposed group

, 0 if the respondent belongs to the control group

γ₀ is the intercept and a and β are the coefficients of the regression model.

16. The computer-implemented method of claim 15, wherein said contribution to said lift is computed by employing a formula in which p is defined as:

17. The computer-implemented method of claim 16, wherein said computed contributions to lift are converted into percentage contribution, in order to identify the strongest driver of lift for each of said brand metrics and said brand perceptions.

18. The computer-implemented method any one of claims 1 to 16, wherein said demographic variables comprise at least one of age, gender, household income, and personal income.

19. At a processing device, a computer-implemented method for determining an impact of an advertisement associated with a brand, said brand being associated with brand consciousness parameters, the method having the steps of:

selecting a control group of respondents and an exposed group of respondents; and generating a survey with questions associated with brand consciousness parameters;

computing an average frequency of exposure to said exposed group;

plotting, on a display, said lift against said frequency of exposure.

20. The computer-implemented method of claim 19, wherein said data is analyzed using polynomial splines to provide an approximation of a function which describes a link between said frequency of exposure and said lift.

21. The computer-implemented method of claim 20, wherein said polynomial splines comprise B-Spline functions.

22. The computer-implemented method of claim 21, wherein said brand consciousness parameters comprise brand metrics and brand perceptions.

23. The computer-implemented method of claim 22, wherein said predictive model is created for said brand perceptions.

24. The computer-implemented method of claim 22, wherein said brand metrics comprises at least one of brand awareness, brand favorability, brand familiarity, recommendation intent and purchase intent.

25. The computer-implemented method of claim 24, wherein said predictive model is created for at least one of said brand metrics.

26. The computer-implemented method of any one of claims 22 to 25, wherein said predictive model for at least one of said brand metrics and each of said brand perceptions comprises a B- Spline curve.

27. The computer-implemented method of claim 26, wherein a point representing a maximum lift for each of said at least one of said brand metrics is determined on said B-Spline curve corresponding to said each of said at least of one of said brand metrics.

28. The computer-implemented method of claim 26, wherein a point representing a maximum lift for each of said brand perceptions is determined on said B-Spline curve corresponding to said each of said brand perceptions.

29. The computer-implemented method of any one of claims 27 and 28, wherein when said B- Spline curve shows the same maximum lift for two different frequencies, said two frequencies comprising a low frequency and a high frequency, then said low frequency corresponds to an optimal frequency for each of said at least of one of said brand metrics and each of said brand perceptions.

30. The computer-implemented method of claim 29, wherein when said B-Spline curve shows a global optima for each curve at different frequencies, then a frequency at which most of said at least of one of said brand metrics and most of said brand perceptions achieve maxima or near maxima together is said optimal frequency.

31. A computer-implemented method for determining a budget for which a lift corresponding to said impact of said advertisement associated with a brand is optimal, said brand being associated with brand consciousness parameters, the method having the steps of:

at said processing device, selecting a control group of respondents and an exposed group of respondents; and generating a survey with questions associated with brand consciousness parameters;

computing an average frequency of exposure to said exposed group;

plotting, on a display, said lift against said frequency of exposure; and

linking said frequency of exposure to said budget.

32. The computer-implemented method of claim 31, wherein said data is analyzed using polynomial splines to provide an approximation of a function which describes a link between said frequency of exposure and said lift.

33. The computer-implemented method of claim 32, wherein said polynomial splines comprise B-spline functions.

34. The computer-implemented method of claim 33, wherein said brand consciousness parameters comprise brand metrics and brand perceptions.

35. The computer-implemented method of claim 34, wherein said predictive model is created for said brand metrics and said brand perceptions.

36. The computer-implemented method of any one of claims 33 to 35, wherein said predictive model for each of said brand metrics and each of said brand perceptions comprises a B-Spline curve.

37. The computer-implemented method of claim 36, wherein a point representing a maximum lift for each of said at least one of said brand metrics is determined on said B-Spline curve corresponding to said each of said at least of one of said brand metrics.

38. The computer-implemented method of claim 37, wherein a point representing a maximum lift for each of said brand perceptions is determined on said B-Spline curve corresponding to said each of said brand perceptions.

39. The computer-implemented method of claim 38, wherein when said B-Spline curve shows the same maximum lift for two different frequencies, said two frequencies comprising a low frequency and a high frequency, then said low frequency corresponds to an optimal frequency for each of said at least of one of said brand metrics and each of said brand perceptions.

40. The computer-implemented method of claim 39, wherein when said B-Spline curve shows a global optima for each curve at different frequencies, then a frequency at which most of said at least of one of said brand metrics and most of said brand perceptions achieve maxima or near maxima together is said optimal frequency.

41. The computer-implemented method of claim 40, wherein said frequency of exposure is linked to said budget by modelling a random distribution of said frequency and subsequently linking said budget to a distribution parameter.

42. The computer-implemented method of claim 41, wherein a logarithmic distribution defines said frequency distribution in which said parameter distribution can be tied to a budget level.

43. The computer-implemented method of claim 42, wherein changes in said budget are used to generate frequency distributions for each of said budget levels.

44. The computer-implemented method of claim 43, wherein said frequency distributions for said budget levels are repeatedly sampled to determine a mean frequency for a particular budget level.

45. The computer-implemented method of claim 44, wherein said upper and lower bounds of said frequency is used for calculating confidence intervals around said frequency and said lift.

46. The computer-implemented method of claim 45, wherein an average frequency for a given budget is determined.

47. The computer-implemented method of claim 46, wherein for a media channel, determining the magnitude of said media channel by computing said lift for each of said brand metrics and said brand perceptions expressed as a percentage of said brand metric and said brand perception proportion in said control group.

48. The computer-implemented method of claim 47, computing the magnitude of said media channel per unit spend on said media channel to determine the efficiency of said media channel.

49. The computer-implemented method of claim 44, on a display, plotting said lift against said budget.

50. An apparatus comprising:

a processing system; and

associate said advertisement with an identifier;

present said advertisement to said exposed group;

51. The apparatus of claim 50, wherein said brand consciousness parameters comprise at least one of brand metrics and brand perceptions.

52. The apparatus of claim 51, wherein said at least one of brand metrics comprises at least one of brand awareness, brand favorability, brand familiarity, recommendation intent and purchase intent.

53. The apparatus of claim 52, wherein said advertisement is presented to said exposed group via at least one of a plurality of channels.

54. The apparatus of claim 53, wherein said at least one of plurality of channels comprises television, and said apparatus determines a contribution of said television channel to a lift corresponding to said impact of said advertisement.

55. The apparatus of claim 53, wherein said at least one of plurality of channels comprises a digital channel, wherein said apparatus determines a contribution of said digital channel to said lift.

56. The apparatus of claim 53, wherein said at least one of plurality of channels comprises radio.

57. The apparatus of claim 56, wherein said apparatus determines a contribution of said radio to said lift.

58. The apparatus of claim 53, wherein said at least one of plurality of channels comprises out of home media (OOH).

59. The apparatus of claim 58, wherein said apparatus determines a contribution of said out of home media (OOH) channel to said lift.

60. The apparatus of claim 53, wherein said lift is determined by combining lifts for each of said at least one of plurality of channels.

61. The apparatus of claim 54, 55, 57 and 59, wherein said lift corresponds to a combined lift due to said advertisement in said television, digital and out of home media (OOH) channels.

62. The apparatus of claim 61, wherein said at apparatus estimates a contribution of at least one media channel associated with at least one advertising campaign in generating said lift in said brand consciousness parameters.

63. The apparatus of claim 62, wherein a logistic regression model is employed, said logistic regression model defined as:

AT ,- is the h ighest level of the demographic factor i .

_ j^'l if for :^"t demographic factor the level is /

Λ^'

ί 0 otherwise

1 if the respondent belongs to the exposed group

0 if the respondent belongs to the control group

}'₀ is the intercept and a and β are the coefficients of the regression model.

64. The apparatus of claim 63, wherein said contribution to said lift is computed by employing a formula in which p is defined as:

Ό = where X is the vector of all the va ria bles involved in the regression (x's and y's from the logistic regression equation) a nd β is the estimate of the coefficient β.

65. The apparatus of claim 64, wherein said computed contributions to lift are converted into percentage contribution, in order to identify the strongest driver of said lift for each of said brand metrics and said brand perceptions.

66. An apparatus for determining an impact of an advertisement associated with a brand, said brand being associated with brand consciousness parameters, the apparatus comprising: a processing system; and

select a control group of respondents and an exposed group of respondents;

generate a survey with questions associated with brand consciousness parameters; transmit said advertisement exposed group of respondents;

compute an average frequency of exposure to said exposed group;

67. The apparatus of claim 66, wherein said data is analyzed using polynomial splines to provide an approximation of a function which describes a link between said frequency of exposure and said lift.

68. The apparatus of claim 67, wherein said polynomial splines comprise B-spline functions.

69. The apparatus of claim 68, wherein said brand consciousness parameters comprise brand metrics and brand perceptions.

70. The apparatus of claim 69, wherein said predictive model is created for said brand perceptions.

71. The apparatus of claim 70, wherein said brand metrics comprises at least one of brand awareness, brand favorability, brand familiarity, recommendation intent and purchase intent.

72. The apparatus of claim71, wherein said predictive model is created for at least one of said brand metrics.

73. The apparatus of any one of claims 69 to 72, wherein said predictive model for at least one of said brand metrics and each of said brand perceptions comprises a B-Spline curve.

74. The apparatus of claim 73, wherein a point representing a maximum lift for each of said at least one of said brand metrics is determined on said B-Spline curve corresponding to said each of said at least of one of said brand metrics.

75. The apparatus of claim 73, wherein a point representing a maximum lift for each of said brand perceptions is determined on said B-Spline curve corresponding to said each of said brand perceptions.

76. The apparatus of any one of claims 74 and 75, wherein when said B-Spline curve shows the same maximum lift for two different frequencies, said two frequencies comprising a low frequency and a high frequency, then said low frequency corresponds to an optimal frequency for each of said at least of one of said brand metrics and each of said brand perceptions.

77. The apparatus of claim 76, wherein when said B-Spline curve shows a global optima for each curve at different frequencies, then a frequency at which most of said at least of one of said brand metrics and most of said brand perceptions achieve maxima or near maxima together is said optimal frequency.

78. The apparatus of claim 77, wherein a budget for which a lift corresponding to said impact of said advertisement associated with a brand is optimal is determined by linking said frequency of exposure to said budget by modelling a random distribution of said frequency and subsequently linking said budget to a distribution parameter.

79. The apparatus of claim 78, wherein a logarithmic distribution defines said frequency distribution in which said parameter distribution can be tied to a budget level.

80. The apparatus of claim 79, wherein changes in said budget are used to generate frequency distributions for each of said budget levels.

81. The apparatus of claim 80, wherein said frequency distributions for said budget levels are repeatedly sampled to determine a mean frequency for a particular budget level.

82. The apparatus of claim 81, wherein said upper and lower bounds of said frequency is used for calculating confidence intervals around said frequency and said lift.

83. The apparatus of claim 82, wherein an average frequency for a given budget is determined.

84. The apparatus of claim 83, wherein for a media channel, determining the magnitude of said media channel by computing said lift for each of said brand metrics and said brand perceptions expressed as a percentage of said brand metric and said brand perception proportion in said control group.

85. The apparatus of claim 84, computing the magnitude of said media channel per unit spend on said media channel to determine the efficiency of said media channel.

86. A computing system for determining the performance of at least one advertising campaign, said system comprises instructions in data storage executable by a processor, said system comprising:

87. A computer-implemented method for determining the performance of at least one advertising campaign, the method having the steps of:

estimating a contribution of at least one media channel associated with said at least one advertising campaign in generating a lift in brand consciousness parameters;