Analysis program, analysis system, computer device, and analysis method

Abstract

The purpose is to provide an analysis program that can more accurately estimate marketing results. A first effect, which is the effect of a specified cause, is estimated from the results of a consumer survey, and the first effect is corrected using the total number of people who came into contact with each of the specified causes and the number of unique people who came into contact with any of the specified causes, which can be calculated from the survey results, to estimate a second effect.

Description

[Technical field]

[0001] The present disclosure relates to an analysis program, an analysis system, a computer device, and an analysis method. [Background technology]

[0002] Conventionally, there is a method called Marketing Mix Modeling (hereinafter, also referred to as MMM) for statistically analyzing and modeling data on marketing measures so as to estimate the extent to which measures such as advertising have affected sales using sales data and time-series data on the amount of advertising spending on TV commercials, etc. The inventor works as a consultant using MMM.

[0003] There are also methods for reading causal relationships from observational data, particularly from real-world data that cannot be reproduced (for example, Non-Patent Document 1). [Prior art documents] [Non-patent literature]

[0004] [Non-Patent Document 1] "Iwanami Data Science Vol. 3," edited by the Iwanami Data Science Publication Committee, published in June 2016 Summary of the Invention [Problem to be solved by the invention]

[0005] However, the conventional analytical methods based on time-series data could not grasp the mechanism of consumer behavior, and were insufficient for estimating the effect of each marketing measure such as advertising on increasing sales. Therefore, the inventor devised a method that uses consumer survey data to grasp the mechanism of consumer behavior and estimate the effect, such as an increase in sales due to visiting a brand's store as a result of a measure such as a TV commercial.

[0006] An objective of at least one embodiment of the present disclosure is to provide an analysis program that can utilize consumer surveys to more accurately estimate marketing results as incremental sales. [Means for solving the problem]

[0007] According to a non-limiting aspect, the analysis program according to the present disclosure is an analysis program executable on a server device in an analysis system including a terminal device and a server device connectable to the terminal device via communication, the server device including: a first effect estimation means for estimating a first effect, which is an effect of a predetermined cause, from a result of a consumer survey; and a second effect estimation means for correcting the first effect by using a total number of contacts with each of the predetermined causes and a unique number of contacts with any of the predetermined causes, which can be counted from the survey results, to estimate a second effect. It is an analytical program that functions as a

[0008] From a non-limiting perspective, the analysis system of the present disclosure is an analysis system including a terminal device and a server device connectable to the terminal device via communication, and includes a first effect estimation means for estimating a first effect, which is the effect of a specified cause, from the results of a consumer survey, and a second effect estimation means for correcting the first effect using the total number of contacts with each of the specified causes and the unique number of contacts with any of the specified causes, which can be compiled from the survey results, to estimate a second effect.

[0009] From a non-limiting perspective, the analysis program of the present disclosure is an analysis program that causes a computer device to function as a first effect estimation means that estimates a first effect, which is the effect of a specified cause, from the results of a consumer survey, and a second effect estimation means that corrects the first effect using the total number of contacts with each of the specified causes and the unique number of contacts with any of the specified causes, which can be compiled from the survey results, to estimate a second effect.

[0010] From a non-limiting perspective, the analytical method of the present disclosure is an analytical method having a step of estimating a first effect, which is the effect of a specified cause, from the results of a consumer survey, and a step of correcting the first effect using the total number of contacts with each of the specified causes and the number of unique contacts with any of the specified causes, which can be compiled from the survey results, to estimate a second effect. Effect of the Invention

[0011] Each embodiment of the present disclosure addresses one or more of the deficiencies. [Brief description of the drawings]

[0012] [Figure 1] FIG. 2 is a block diagram showing a configuration of a server device corresponding to at least one of the embodiments of the present disclosure. [Diagram 2] 1 is a flowchart of a program execution process according to at least one of the embodiments of the present disclosure. [Diagram 3] FIG. 1 is a block diagram showing a configuration of an analysis system corresponding to at least one of the embodiments of the present disclosure. [Figure 4] FIG. 1 is a block diagram showing a configuration of an analysis system corresponding to at least one of the embodiments of the present disclosure. [Diagram 5] FIG. 2 is a diagram illustrating an example of survey data corresponding to at least one of the embodiments of the present disclosure. [Figure 6] FIG. 2 is a diagram illustrating an example of survey data corresponding to at least one of the embodiments of the present disclosure. [Figure 7] FIG. 2 is a diagram illustrating an example of survey data corresponding to at least one of the embodiments of the present disclosure. [Figure 8] FIG. 2 is a diagram illustrating an example of survey data corresponding to at least one of the embodiments of the present disclosure. [Figure 9] FIG. 2 is a diagram illustrating an example of survey data corresponding to at least one of the embodiments of the present disclosure. [Figure 10] FIG. 2 is a diagram illustrating an example of survey data corresponding to at least one of the embodiments of the present disclosure. [Figure 11] 1 is a flowchart of an execution process according to at least one of the embodiments of the present disclosure. [Figure 12] FIG. 2 is a diagram illustrating an example of cross-tabulation of contact rates by attribute derived from consumer survey data, corresponding to at least one of the embodiments of the present disclosure. [Figure 13] FIG. 11 is a diagram illustrating an example of aggregating single reach rates by attribute derived from consumer survey data, corresponding to at least one of the embodiments of the present disclosure. [Figure 14] FIG. 11 is a diagram showing contact rates by age group from multiple viewpoints, corresponding to at least one of the embodiments of the present disclosure. [Figure 15] FIG. 11 is a diagram for explaining correction of data regarding the contact rate of women in their twenties, corresponding to at least one of the embodiments of the present disclosure. [Figure 16] 11 is a diagram for explaining a method of calculating an overlap adjustment coefficient, corresponding to at least one of the embodiments of the present disclosure. FIG. [Figure 17] FIG. 11 is a diagram showing an example of a calculation result of a sales contribution amount for each measure, corresponding to at least one of the embodiments of the present disclosure. [Figure 18] FIG. 11 is a diagram showing an example of a calculation result of a sales contribution amount for each factor, corresponding to at least one of the embodiments of the present disclosure. [Figure 19] FIG. 11 is a diagram showing an example of a calculation result of a sales contribution amount for each measure and factor, corresponding to at least one of the embodiments of the present disclosure. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0013] Hereinafter, examples of embodiments of the present disclosure will be described with reference to the accompanying drawings. The following description of the effects is one aspect of the effects of the embodiments of the present disclosure, and is not limited to those described here. In addition, the contents described as an example of one embodiment may be omitted in other embodiments. Furthermore, the contents of operations and processes that are not related to the characteristic parts of each embodiment may be omitted. The order of each process constituting the flowchart described below is random to the extent that no contradiction or inconsistency occurs in the process contents.

[0014] [First embodiment] In the following, as a first embodiment, an analysis system including a terminal device and a server device connectable to the terminal device through communication will be described as an example. The terminal device and / or the server device may be connectable to each other through communication. The server device and / or the terminal device may be connectable to a distributed ledger network.

[0015] 1 is a block diagram showing a configuration of a server device 10 according to at least one of the embodiments of the present disclosure. The server device 10 may include at least a first effect estimating unit 101 and a second effect estimating unit 102.

[0016] The first effect estimation unit 101 has a function of estimating a first effect, which is an effect of a predetermined cause, from the survey results of a consumer survey. The second effect estimation unit 102 has a function of correcting the first effect using the total number of people who came into contact with each of the predetermined causes and the number of unique people who came into contact with any of the predetermined causes, which can be counted from the survey results, to estimate a second effect.

[0017] 2 is a flowchart of a program execution process corresponding to at least one of the embodiments of the present disclosure. The server device 10 estimates a first effect, which is an effect of a predetermined cause, from the results of a consumer survey (step S1). Next, the server device 10 corrects the first effect using the total number of contacts with each of the predetermined causes and the number of unique contacts with any of the predetermined causes, which can be counted from the survey results, to estimate a second effect (step S2), and ends the process.

[0018] As one aspect of the first embodiment, an analysis program can be provided that can more accurately estimate marketing results as increased sales by utilizing consumer surveys.

[0019] In the first embodiment, the term "terminal device" refers to, for example, a stationary game machine, a portable game machine, a wearable terminal, a desktop or notebook personal computer, a tablet computer, a PDA, or a portable terminal such as a smartphone equipped with a touch panel sensor on the display screen. The term "server device" refers to, for example, a device that executes processing in response to a request from a terminal device.

[0020] In the first embodiment, a "cause" refers to, for example, a factor that has the effect of stimulating consumer consumption activity, and more specifically, examples of such factors that could be questions in a consumer survey include watching a television commercial, watching an online commercial, visiting a brand store, or seeing a push notification from an app.

[0021] In the first embodiment, the "total number of contacts" refers to, for example, the number of people who answered that they had been exposed to each of the causes. The "unique number of contacts" refers to, for example, the number of people who answered that they had been exposed to any of the causes. In other words, it refers to the number of people who answered that they had not been exposed to any of the causes, minus the number of people who answered that they had not been exposed to any of the causes, from the total number of people who answered the consumer survey questionnaire.

[0022] [Second embodiment] In the following, as a second embodiment, an analysis system including a terminal device and a server device connectable to the terminal device through communication will be described as an example. The terminal device and / or the server device may be connectable to each other through communication. The server device and / or the terminal device may be connectable to a distributed ledger network.

[0023] [Analysis system] 3 is a block diagram showing a configuration of an analysis system corresponding to at least one of the embodiments of the present disclosure. As shown in the figure, the analysis system (system) 1 includes a server device 10, a communication network 20, and terminal devices 30 (30A, 30B, . . . , 30N: N is any character) used by each of a plurality of users.

[0024] The configuration of the analysis system 1 is not limited to this, and for example, the server device 10 may be configured by multiple server devices, or may be configured by a virtual server device using cloud computing technology. Furthermore, the communication network 20 may be a distributed ledger network.

[0025] The server device 10 and the terminal device 30 are connected to each other so as to be able to communicate with each other. The terminal device 30 and the server device 10 do not need to be connected to each other at all times, but may be connected to each other as needed.

[0026] [Server device] As an example, the server device 10 includes at least a control unit, a RAM, a storage unit, and a communication interface, which are connected to each other via an internal bus. The control unit may include an internal timer. The communication interface may also be used to synchronize with an external server, thereby obtaining the actual time.

[0027] [Communication Network] The communication network 20 may be, for example, the Internet or a LAN, and may be any network that allows communication between connected devices.

[0028] [Terminal Device] The terminal device 30 includes, for example, a control unit, a RAM, a storage unit, a sound processing unit, a graphics processing unit, a communication interface, and an interface unit, which are connected to each other via an internal bus. The graphics processing unit is connected to a display unit. The display unit may have a display screen and a touch input unit that accepts input by a user touching the display unit.

[0029] The touch input unit may be capable of detecting the position of contact using any method, such as a resistive film method used in touch panels, a capacitive method, an ultrasonic surface acoustic wave method, an optical method, or an electromagnetic induction method, and the method is not important as long as the operation can be recognized by a user's touch operation. The touch input unit is a device capable of detecting the position of a finger or the like when the upper surface of the touch input unit is pressed or moved with a finger or a stylus.

[0030] An external memory (e.g., an SD card, etc.) can be connected to the interface unit. Data read from the external memory is loaded into the RAM, and arithmetic processing is performed by the control unit. The communication interface can be connected to a communication network wirelessly or via a wire, and is capable of receiving data via the communication network. Data received via the communication interface is loaded into the RAM, similar to data read from the external memory, and arithmetic processing is performed by the control unit.

[0031] The terminal device 30 may include a sensor such as a proximity sensor, an infrared sensor, a gyro sensor, or an acceleration sensor. The terminal device 30 may include an imaging unit that has a lens and captures an image through the lens. The terminal device 30 may be a terminal device that can be attached to the body (wearable).

[0032] [Function Description] The following describes functions of the analysis system 1 according to the second embodiment. Fig. 4 is a block diagram showing the configuration of an analysis system corresponding to at least one of the embodiments of the present disclosure.

[0033] The analysis system 1 may include a contact estimation unit 201, a first effect estimation unit 202, a second effect estimation unit 203, and a sales contribution estimation unit 204.

[0034] The contact estimation unit 201 has a function of estimating the number of contacts who came into contact with a predetermined cause from the survey results of a consumer survey. The first effect estimation unit 202 has a function of estimating a first effect, which is the effect of a predetermined cause, from the survey results of a consumer survey.

[0035] The second effect estimation unit 203 has a function of correcting the first effect using the total number of people who came into contact with each of the predetermined causes and the number of unique people who came into contact with any of the predetermined causes, which can be counted from the survey results, to estimate the second effect. The sales contribution estimation unit 204 has a function of estimating the degree of sales contribution for each attribute based on each of the causes, using the number of people who came into contact with the predetermined cause, the second effect, the average purchase price, and the average number of purchases estimated from a probability model of a predetermined consumer purchase.

[0036] [Consumer survey data] 5 to 10 are diagrams for explaining examples of survey data corresponding to at least one of the embodiments of the present disclosure. Data that is answered by consumers in a questionnaire format is collected in an analyzable amount.

[0037] It is preferable that the consumer survey data include the degree of favorability toward the brand as shown in Figure 5, the timing of exposure to the brand as shown in Figure 6, the degree of passive exposure to the brand as shown in Figure 7, the degree of active exposure to the brand as shown in Figure 8, the respondent's exposure time to the media as shown in Figure 9, and future intention to use the brand as shown in Figure 10.

[0038] In addition, the consumer survey data may be controlled so that, as shown in FIG. 5, a brand for which the customer answers "I don't know" is treated as an invalid answer for the brands in the subsequent questions in FIGS.

[0039] It is preferable for consumer survey questions to include questions that address passive attitudes toward the brand from the consumer's perspective, as shown in Figure 7, and questions that address active attitudes toward the brand, as shown in Figure 8.

[0040] [Execution process] 11 is a flowchart of an execution process corresponding to at least one of the embodiments of the present disclosure. The server device 10 preferably reads data on the results of a consumer survey conducted in advance. In addition, the data on the survey results of the consumer survey preferably includes information on the attributes of the respondents and, if the respondents have purchased the product, information on the average purchase price.

[0041] The consumer survey data may be the answered data read as it is, or the data that has been pre-edited may be read in. There is no particular limitation on the form as long as the effects of the present disclosure are achieved.

[0042] Below, we will provide a detailed explanation using responses to a survey of approximately 6.16 million women in their 20s regarding television commercials for restaurant chains.

[0043] The analysis system 1 estimates the number of people who came into contact with the predetermined cause from the survey results (step S21). In the verified example, the contact rate of women in their 20s was about 59%, so it is estimated that about 3.63 million people came into contact.

[0044] Next, the server device 10 estimates a first effect, which is the effect of a predetermined cause, from the results of the consumer survey (step S22).

[0045] [First effect estimation process] The estimation of the first effect will be described. Fig. 12 is a diagram for explaining an example of cross-tabulation of contact rates by attribute derived from consumer survey data, which corresponds to at least one of the embodiments of the present disclosure. The data shown as "Measures" in Fig. 12 is a cross-tabulation of the answers to the questions shown in Fig. 7, and the data shown as "Factors" in Fig. 12 is a cross-tabulation of the answers to the questions shown in Fig. 8.

[0046] Next, the "single reach rate" is calculated. The single reach rate is the percentage of respondents who answered that they had only one type of contact for each measure and factor. FIG. 13 is a diagram for explaining an example of calculating the single reach rate by attribute derived from consumer survey data, which corresponds to at least one of the embodiments of the present disclosure.

[0047] Next, causal inference is used to derive the increment in penetration rate (lift rate) for each cause. For example, a simple comparison between people who have seen a TV commercial and those who have not will result in selection bias, so causal inference is used to adjust for the bias. Propensity score analysis is one method that uses causal inference. As propensity score analysis is not the essence of the contents of this disclosure, a detailed explanation will be omitted.

[0048] In the model disclosed here, rather than focusing on predictive ability, we place importance on whether the covariates that cause bias in the estimation of effects are balanced between the two groups. More specifically, if the sample size is small when analyzing by age and sex (a rough guide is 40), the balance between the two groups becomes unstable, so if the sample size is 40 or less, a process is performed to substitute the average effect of the policy or factor.

[0049] In this disclosure, the value estimated by ATT (Average Treatment effect on the Treated) was used as the increment in the penetration rate (lift rate). ATT is the weight value of the IPTW estimator, which calculates the propensity score of each sample using the covariates in the propensity score analysis, and aggregates it by weighting. The covariates used are data such as whether or not the person has children, marital status, and type of residence, which can be obtained as basic attributes of the Internet survey, as well as brand favorability (Figure 5), media contact time (Figure 9), and usage intention (Figure 10), which are obtained from the survey responses. The favorability and usage intention are excluded from intermediate variables that play an intermediate role between cause and effect. In the specific example of this disclosure, it was estimated that the usage rate increased due to TV commercials for approximately 4.22% of the approximately 3.63 million contacts.

[0050] Returning to the explanation of Fig. 11, the analysis system 1 corrects the first effect using the total number of people who came into contact with each of the predetermined causes and the number of unique people who came into contact with any of the predetermined causes, which can be counted from the investigation results, to estimate the second effect (step S23).

[0051] [Second effect estimation process] First, the analysis system 1 uses the response data of the consumer survey to compile the "total reach," "unique user reach," "total single reach," and "overlapping unique user reach" by age and gender.

[0052] FIG. 14 is a diagram showing contact rates by age group from multiple perspectives, corresponding to at least one of the embodiments of the present disclosure. The reach represents the total contact rate of each measure. The unique user reach represents the proportion of respondents who were contacted by any of the measures. The overlapping unique user reach represents the value obtained by subtracting the total reach from the unique user reach.

[0053] If the effect of each measure (increase in penetration rate, etc.) is, for example, 10%, and there is no overlapping reach of the penetration rate of each measure, the calculation can be done as follows: Population of each age and gender x Reach (contact rate) (%) x 10% = Number of people with penetration rate lift. However, in reality, there are overlapping responses, so the estimation results are overestimated, and the results of marketing cannot be evaluated correctly.

[0054] As a specific example, a method for correcting the value of the first effect when the number of unique users reached by women in their twenties is 4.647 million will be described. The number of unique users reached is a value derived from response data of consumer survey data. FIG. 15 is a diagram for explaining correction of data related to the contact rate of women in their twenties, corresponding to at least one of the embodiments of the present disclosure.

[0055] The increase in the number of people due to the penetration rate for single reach can be estimated by multiplying the population of each age and gender, the reach (contact rate), and ATT to estimate the penetration lift number. The penetration lift number taking into account overlaps is calculated by multiplying the total overlap reach number by the overlap adjustment coefficient to match the total number of unique user reach, which is approximately 4.647 million people.

[0056] The overlap adjustment coefficient is a value obtained by dividing the overlap unique user reach by the overlap reach for each age group and gender. FIG. 16 is a diagram for explaining a method for calculating the overlap adjustment coefficient, corresponding to at least one of the embodiments of the present disclosure. The corrected number of contacts is calculated by the sum of the value obtained by multiplying the population by the single reach (%) and the value obtained by multiplying the population by the overlap reach (%) and then multiplying the result by the overlap coefficient. The effect adjustment coefficient is calculated by dividing the corrected number of contacts by the actual number of contacts.

[0057] In the model disclosed herein, the single reach of each media is left as it is. The method described above uses the "number of people reached" to calculate the number of unique users reached by overlapping parts, and the "effect adjustment coefficient (corrected number of contacts / actual number of contacts)" x "penetration lift rate (estimated by ATT)" to estimate the lift number of users for each measure, which makes it possible to estimate the corrected penetration rate, which is the second effect.

[0058] A specific example of the present disclosure will be described with reference to FIG. 15 and FIG. 17. FIG. 17 is a diagram showing an example of the calculation result of the sales contribution amount for each measure, which corresponds to at least one of the embodiments of the present disclosure. As shown in FIG. 15, 13 types of measures were asked about in the survey, and about 75.3% of the respondents in the entire age group had contact with any of them. In this case, as a correction to return the number of contacts to the unique number for the measure name "TV commercial", the lift rate is multiplied by the "effect adjustment coefficient (corrected number of contacts / actual number of contacts)" of about 84.4%, so that the 4.22% estimated as the first effect can be corrected to about 3.55%. The value after this correction is the second effect.

[0059] Returning to the explanation of Fig. 11, the analysis system 1 estimates the degree of sales contribution of each attribute based on each of the causes using the number of contacts of the predetermined cause, the second effect, the average purchase price, and the average number of purchases estimated from a predetermined probability model of consumer purchases (step S24), and ends the process.

[0060] [Sales contribution estimation function] The second effect estimation function allowed us to estimate the number of users who used the lift for each measure. Next, we calculated the increased number of times and the average cost per use (also called the average purchase cost or average cost per use).

[0061] It is preferable to use purchase data obtained by one's company to determine the increased number of times, but when targeting brands other than one's own company, it is preferable to derive the number of times using a probability model of consumer purchases based on consumer surveys. As a probability model of consumer purchases, for example, the "Gamma Poisson recency model" shown in Figure 6, which performs analysis from recency data asking survey subjects about the time of their last purchase, can be used. These models have already been studied, and existing probability models will suffice for the probability models used in this disclosure, so a detailed explanation will be omitted here.

[0062] Next, we calculate the average cost per use. This can be calculated from the response data of the consumer survey.

[0063] In this disclosure, sales contribution refers to the lift amount for each measure, calculated by multiplying the estimated number of lifts by users for each measure, the increased number of lifts, and the average unit price per use calculated from the survey.

[0064] Fig. 18 is a diagram showing an example of a calculation result of the sales contribution amount for each factor, corresponding to at least one of the embodiments of the present disclosure. In the model of the present disclosure, the contents analyzed by the measures (Fig. 17) and the contents analyzed by the factors (Fig. 18) may be combined to estimate the effect of the measures → factors → sales (increase in penetration rate × number of times × unit price).

[0065] Fig. 19 is a diagram showing an example of the calculation result of the sales contribution amount via each factor with the measure being a TV commercial, corresponding to at least one of the embodiments of the present disclosure. The lift amount due to TV commercials directly estimated using the model of Fig. 17 (measure → sales) was approximately 582 million yen. On the other hand, the model of the present disclosure shown in Fig. 19 has a total lift amount of approximately 1.273 billion yen from the measure (TV commercial) → factor → sales, which is different from the lift amount in Fig. 17. There may be a discrepancy between the directly estimated model and the model estimated from the assistance via factors.

[0066] By conducting an analysis of measures → factors → sales, it is possible to grasp the communication structure by conducting a detailed analysis of the factors through which the measures contributed to sales.

[0067] As one aspect of the second embodiment, an analysis program capable of more accurately estimating marketing results can be provided.

[0068] In the second embodiment, the “terminal device,” “server device,” “cause,” “total number of contacts,” and “unique number of contacts” may adopt the contents described in the first embodiment to the extent necessary.

[0069] In the second embodiment, the "average purchase price" refers to, for example, the total amount required to acquire a product or service divided by the quantity held or the number of times provided. The "degree of sales contribution" refers to, for example, the degree of contribution to an increase in sales, and may be expressed as a numerical value or a percentage. The "reach" refers to, for example, the actual number of people who came into contact with a specific site or brand during a certain period of time, or the percentage of that number to the total number of Internet users, and is also called the number of contacts. The "causal inference" refers to, for example, the idea of ​​statistically estimating causal relationships, which are the relationship between causes and the results that result from them, from input data and output data. The "propensity score" refers to, for example, a statistical method of balancing used to estimate causal effects by adjusting covariates in observational studies in which random allocation is difficult and various confounds are likely to occur.

[0070] In the second embodiment of the present disclosure, the analysis system 1 includes all the functions, but is not limited to this. For example, the functions may be distributed to the server device 10 and the terminal device 30.

[0071] The contents described in each of the above embodiments can be appropriately designed by those skilled in the art by rearranging them.

[0072] [Note] The above description of the embodiments has been provided so as to enable a person having ordinary skill in the art to which the invention pertains to practice the following invention.

[0073] [1] An analysis program executable on a server device in an analysis system including a terminal device and a server device connectable to the terminal device through communication, The server device, A first effect estimation means for estimating a first effect, which is the effect of a predetermined cause, from the results of a consumer survey; A second effect estimation means for correcting the first effect by using the total number of people who have come into contact with each of the predetermined causes and the number of unique people who have come into contact with any of the predetermined causes, which can be counted from the survey results, to estimate a second effect. An analytical program that acts as a

[0074] [2] The survey results include respondent attributes and average purchase price. The server device further includes: A sales contribution estimation function that estimates the degree of sales contribution for each attribute based on each of the causes using the number of contacts of the predetermined cause, the second effect, the average purchase price, and an average purchase frequency estimated from a predetermined probability model of consumer purchases. The analysis program according to [1],

[0075] [3] The predetermined cause includes a content related to a measure and a content related to a factor, The analysis program described in [2], wherein the sales contribution estimation function estimates the degree of sales contribution for each attribute using the second effect estimated for each content of the measure and the second effect estimated for each content of the factor.

[0076] [4] The server device further includes: A contact number estimation means for estimating the number of people who have come into contact with a predetermined cause based on the results of the investigation The analysis program according to [1] or [2], which functions as

[0077] [5] The analysis program according to claim 1 or 2, wherein the first effect estimation function estimates the first effect using causal inference.

[0078] [6] The analysis program described in [1] or [2], wherein the first effect estimation function estimates the first effect using a propensity score.

[0079] [7] The analysis program according to [1] or [2], which uses a Gamma Poisson recency model as the probability model of the specified consumer purchase.

[0080] [8] A server device having the program described in [1] or [2] installed.

[0081] [9] An analysis system including a terminal device and a server device connectable to the terminal device through communication, a first effect estimation means for estimating a first effect, which is the effect of a predetermined cause, from the results of a consumer survey; A second effect estimation means for correcting the first effect by using a total number of people who have come into contact with each of the predetermined causes and a unique number of people who have come into contact with any of the predetermined causes, which can be counted from the survey results, to estimate a second effect; An analysis system comprising:

[0082]

[10] An analysis program executable on a terminal device in an analysis system including a terminal device and a server device connectable to the terminal device through communication, The terminal device, A first effect estimation means for estimating a first effect, which is the effect of a predetermined cause, from the results of a consumer survey; A second effect estimation means for correcting the first effect by using the total number of people who have come into contact with each of the predetermined causes and the number of unique people who have come into contact with any of the predetermined causes, which can be counted from the survey results, to estimate a second effect. An analytical program that acts as a

[0083]

[11] A terminal device having the program described in

[10] installed.

[0084]

[12] An analysis method executed in a server device of an analysis system including a terminal device and a server device connectable to the terminal device through communication, comprising: A step of estimating a first effect, which is an effect of a predetermined cause, from the results of a consumer survey; A step of correcting the first effect using the total number of contacts with each of the predetermined causes and the number of unique contacts with any of the predetermined causes, which can be counted from the survey results, to estimate a second effect; The analytical method has the following features:

[0085]

[13] An analysis method to be executed in an analysis system including a terminal device and a server device connectable to the terminal device through communication, A step of estimating a first effect, which is an effect of a predetermined cause, from the results of a consumer survey; A step of correcting the first effect using the total number of contacts with each of the predetermined causes and the number of unique contacts with any of the predetermined causes, which can be counted from the survey results, to estimate a second effect; The analytical method has the following features:

[0086]

[14] Computer equipment A first effect estimation means for estimating a first effect, which is the effect of a predetermined cause, from the results of a consumer survey; A second effect estimation means for correcting the first effect by using the total number of people who have come into contact with each of the predetermined causes and the number of unique people who have come into contact with any of the predetermined causes, which can be counted from the survey results, to estimate a second effect. An analytical program that acts as a

[0087]

[15]

[14] A computer device having the analysis program described in

[14] installed.

[0088]

[16] A step of estimating a first effect, which is an effect of a predetermined cause, from the results of a consumer survey; A step of correcting the first effect using the total number of contacts with each of the predetermined causes and the number of unique contacts with any of the predetermined causes, which can be counted from the survey results, to estimate a second effect; The analytical method has the following features: [Industrial Applicability]

[0089] According to one embodiment of the present disclosure, it is useful to provide an analysis program that can more accurately estimate marketing results. [Explanation of symbols]

[0090] 1: Analysis system 10: Server device 20: Communication Networks 30: Terminal device

Claims

1. An analysis program executable on a server device in an analysis system including a terminal device and a server device connectable to the terminal device through communication, The server device, A first effect estimation means for estimating a first effect, which is the effect of a predetermined cause, from the results of a consumer survey; A second effect estimation means for correcting the first effect by using the total number of people who have come into contact with each of the predetermined causes and the number of unique people who have come into contact with any of the predetermined causes, which can be counted from the survey results, to estimate a second effect. An analytical program that acts as a

2. The survey results include respondent attributes and average purchase price. The server device further includes: A sales contribution estimation function that estimates the degree of sales contribution for each attribute based on each of the causes using the number of contacts of the predetermined cause, the second effect, the average purchase price, and an average purchase frequency estimated from a predetermined probability model of consumer purchases. The analysis program according to claim 1 ,

3. The predetermined cause includes a content related to a measure and a content related to a factor, The analysis program according to claim 2 , wherein the sales contribution estimation function estimates the degree of sales contribution for each attribute using the second effect estimated for each content of the measure and the second effect estimated for each content of the factor.

4. An analysis system including a terminal device and a server device connectable to the terminal device through communication, a first effect estimation means for estimating a first effect, which is the effect of a predetermined cause, from the results of a consumer survey; A second effect estimation means for correcting the first effect by using a total number of people who have come into contact with each of the predetermined causes and a unique number of people who have come into contact with any of the predetermined causes, which can be counted from the survey results, to estimate a second effect; An analysis system comprising:

5. Computer equipment A first effect estimation means for estimating a first effect, which is the effect of a predetermined cause, from the results of a consumer survey; A second effect estimation means for correcting the first effect by using the total number of people who have come into contact with each of the predetermined causes and the number of unique people who have come into contact with any of the predetermined causes, which can be counted from the survey results, to estimate a second effect. An analytical program that acts as a

6. a first effect estimation means for estimating a first effect, which is the effect of a predetermined cause, from the results of a consumer survey; A second effect estimation means for correcting the first effect by using a total number of people who have come into contact with each of the predetermined causes and a unique number of people who have come into contact with any of the predetermined causes, which can be counted from the survey results, to estimate a second effect; A computer device comprising:

7. An analysis method for an analysis system including a terminal device and a server device connectable to the terminal device through communication, the method comprising: A step of estimating a first effect, which is an effect of a predetermined cause, from the results of a consumer survey; A step of correcting the first effect using the total number of contacts with each of the predetermined causes and the number of unique contacts with any of the predetermined causes, which can be counted from the survey results, to estimate a second effect; The analytical method has the following features: