JPWO2015162723A1 - Behavior analysis device - Google Patents

Abstract

The present invention solves the problem of automatically determining purpose purchase and non-purpose purchase. In the store, equipped with a device that measures the stop time that is the time stopped in front of the product shelf, and the after-stop purchase rate that is the ratio of the number of people who stopped at the product shelf and the number of people who purchased the product on the product shelf A value obtained by applying a two-dimensional filter for the drop-in rate, which is the ratio of the number of visitors to the store and the number of people who have visited the product shelf, to the store as a whole, and applying the two-dimensional filter processing of the drop-in rate Is a post-filter stop-off rate, the apparatus includes a device for calculating a shelf having a long stop-off time, a low post-stop purchase rate, and a low post-filter stop-off rate.

Description

  The present invention relates to an apparatus for analyzing behavior of a customer or the like in a store or the like.

  In order to consider measures to improve sales at stores, it is important to understand the behavior of customers and other stores.

  As a prior art, a system for analyzing a customer's walking route and purchase behavior in a store is known.

JP 2003-223548

  In order to provide new value to customers at a store, it is necessary to find and purchase products that were not intended for purchase at the store. However, until now, it has been difficult to automatically determine the characteristics of a product in a store and the characteristics of a product shelf on which the product is to be placed, and thus it has been difficult to display products efficiently.

One aspect of the invention of the present application is a stop-by time that is a stoppage time in front of a product shelf in a store, and a purchase after stop-off that is a ratio of the number of people who stopped at the product shelf and the number of people who purchased the product on the product shelf. A two-dimensional filter that smoothes the drop-off rate, which is the ratio of the number of customers visiting the product shelf and the number of people who visited the product shelf, on a flat surface for the entire store. And a device for calculating a product shelf having a long drop-in time, a low purchase rate after drop-in, and a low drop-in rate after two-dimensional filtering.
In another aspect of the present invention, in a store, a stop-off time that is a stoppage time in front of a product shelf, a purchase after stop-off that is a ratio of the number of people who stopped at the product shelf and the number of people who purchased the product on the product shelf It is a behavior analysis device equipped with a device for measuring the rate, and has a device for smoothing the stop-off rate, which is the ratio of the number of customers visiting the product shelf and the number of people who visited the product shelf, for each product shelf in a flat manner throughout the store The first product shelf with a long drop-in time, low purchase rate after drop, low drop-off rate after smoothing, short drop-in time, high purchase rate after drop-off, and high drop-off rate after smoothing And a device for calculating a customer unit price improvement expected value when the product in the first product shelf and the product in the second product shelf are exchanged.
In the present invention, the length of the drop-off time and the level of the purchase rate after drop-off are used as indicators for indicating the tendency of the product characteristics. Therefore, there is no particular limitation as long as it can be determined by setting some standard. As a determination method, discrimination can be performed by providing one or a plurality of predetermined threshold values. Further, the digitized data may be rearranged and a certain ratio may be extracted from the upper or lower order. The same applies to the high and low drop-in rates.
Still another aspect of the present invention relates to an apparatus including an input device, an output device, a processing device, and a storage device. The input device has a function of inputting behavior history information from a portable sensor carried by a customer who has entered a predetermined geographical area while staying in the predetermined geographical area. The action history information includes information regarding continuous or discrete time and place. Based on this behavior history information, the processing device extracts stop-off information that indicates whether or not each customer has stopped at a plurality of predetermined locations in the geographical range. In addition, the processing device maps, for each of a plurality of predetermined locations, a stoppage rate that is a ratio of the total number of visitors and the number of customers who have stopped at the predetermined location, and further smoothes the stopover rate two-dimensionally. . Further, the processing device displays the smoothed drop-in rate data on the output device, stores it in the storage device, or both.
Still another aspect of the present invention relates to an apparatus including an input device, an output device, a processing device, and a storage device. The input device has a function of inputting behavior history information from a portable sensor carried by a customer who has entered a predetermined geographical area while staying in the predetermined geographical area. The action history information includes information regarding continuous or discrete time and place. Based on this action history information, the processing device extracts stop time information indicating the length of the stop time for each customer at each of the predetermined locations. The input device inputs purchase information indicating whether or not each customer has purchased a product arranged at a plurality of predetermined locations for each of these locations. Based on the purchase information and the action history information, the processing device calculates a post-stop purchase rate that is a ratio of the number of people who have stopped at a specific place and the number of people who have purchased a product placed at a predetermined place. The processor further labels and / or sorts the products located in a plurality of predetermined locations based on the length of the stop-off time and the purchase rate after the stop-off. Do.

ADVANTAGE OF THE INVENTION According to this invention, based on a customer's action history, the information for improving the arrangement | positioning of the goods in a store can be acquired automatically, and it is possible to provide a new value to a customer in a store. .

It is the block diagram which showed the implementation method of the behavioral analysis apparatus of this invention. It is a conceptual diagram explaining that sales improve by the action analysis apparatus of this invention. It is the conceptual diagram which showed the purchase pattern of purpose buying and non-purpose buying. It is the figure which applied the low pass filter to the drop-in rate of the sales floor. It is a perspective view explaining the device which measures a drop-in rate. It is the flowchart which showed the procedure which performs the measurement of a sales floor. It is a chart of an example of summary data which is a measurement result. It is a table | surface figure of an example of time from each shelf which is a measurement result. It is a table | surface figure of an example of the purchase amount classified by shelf which is a measurement result. It is a chart of an example of the purchase score according to shelf which is a measurement result. It is a chart of an example of the after-stop purchase rate which is an analysis result. It is a table | surface figure of an example of the flattening stoppage rate which is an analysis result. It is a chart of an example of a goods arrangement list which is an analysis result.

  Embodiments of the present invention will be described below with reference to the drawings. However, the present invention is not construed as being limited to the description of the embodiments below. Those skilled in the art will readily understand that the specific configuration can be changed without departing from the spirit or the spirit of the present invention.

  In order to provide new value to customers at a store, it is necessary to find and purchase products that were not intended for purchase at the store. For this purpose, it is important to specify a product when a customer purchases a product that was not intended for purchase. However, it has been difficult to automatically determine purpose purchases aimed at purchase and non-purpose purchases that have not been purposed so far. In addition, it is difficult to automatically determine the position where the non-purpose purchased product is to be placed. For this reason, it was difficult to arrange the non-purpose purchase product at an appropriate position.

  FIG. 1 is a system configuration diagram for performing behavior analysis in a store according to an embodiment of the present invention. The entire system can be realized by a computer or server including an input device, an output device, an information processing device, and a storage device. In the present embodiment, each processing unit to be described later is assumed to be executed by software operating on a computer, but may be configured by hardware. In addition, each database described later can be stored in a storage device.

  In FIG. 1, reference numeral 101 denotes a measurement result database that stores basic information for each customer ID, stop-by time per shelf, purchase price per shelf, purchase points per shelf, and the like. A data collection method will be described later with reference to FIG. The data stored in the measurement result database 101 will be described later with reference to FIG.

  Reference numeral 102 denotes a post-stop purchase rate analysis processing unit, 103 denotes a drop-in rate analysis processing unit, 104 denotes a product placement optimization processing unit, and 105 denotes a product placement list database indicating the best placement candidates for the product. The data stored in the product arrangement list database 105 will be described later with reference to FIG.

First, the post-stop purchase rate analysis processing unit 102 will be described. The post-stop purchase rate analysis processing unit 102 includes a stop-time database 121, a purchase price database 122, a ratio analysis process 123, and a post-stop purchase rate analysis result database 124.
Data stored in the drop-in time database 121 will be described later with reference to FIG. This data is the total of the stop-by time for each customer at each store.
Data stored in the purchase price database 122 will be described later with reference to FIG. The purchase amount 122 is a total of purchase amounts for each shelf for each customer visiting the store. The purchase price may be the number of purchases if the price difference for each product is ignored. The data in this case will be described later with reference to FIG.
The ratio analysis process 123 analyzes the ratio or distribution of values for each of the drop-in time 121 and the purchase amount 122, and determines whether the value is relatively large or small. Each determination result may be digitized.
Further, the ratio analysis process 123 calculates a post-stop purchase rate, which is a ratio purchased after the stop. The after-stop purchase ratio, for example, refers to the drop-in time database 121 and totals the number C of customers who stopped by each shelf (that is, the number of customers having stop-by-shelf time other than 0) and refers to the purchase price database 122 The number of customers B who purchased the product for each shelf (that is, the number of customers having a purchase price by shelf other than 0) may be tabulated, and B may be divided by C for each shelf. When calculating the number C of customers who stopped by each shelf, a threshold value (for example, 30 seconds or more) may be provided for the stop time.
In addition, based on these judgment results, the specific product on a specific shelf has a long customer visit time and the purchase rate after the visit is low, or the customer purchase time is short and the customer purchases after the visit. A numerical value that can identify whether the rate is high may be calculated. For example, if a value obtained by dividing the stop time by the after-stop purchase rate is used, a product having a tendency of non-purpose purchase product has a large value. At this time, both values may be weighted with an appropriate coefficient. These digitized data can be stored in the after-stop purchase rate analysis result database 124.
The determination result and the post-stop purchase rate are stored in the post-stop purchase rate analysis result 124. FIG. 11 shows an example of after-stop purchase rate data stored in the after-stop purchase rate analysis result database 124.

  Next, the drop-in rate analysis processing unit 103 will be described. The drop-in rate analysis processing unit 103 includes a drop-in rate database 131, a two-dimensional low-pass filter process 132, and a drop-in rate analysis result database 133. The drop-in rate 131 is the ratio of the number of customers who visited by shelf among the number of customers visiting the store (for example, shown in FIG. 11). The drop-in rate for each shelf is affected by the product characteristics and shelf layout characteristics. The product characteristics are, for example, a high drop-in rate to a shelf with hit products. Further, the shelf arrangement characteristics include, for example, a high drop-in rate near the entrance. As described above, the drop-in rate is greatly different for each shelf due to various factors. It is the two-dimensional low-pass filter process 132 that smoothes these values. In the two-dimensional low-pass filter processing 132 of the present embodiment, smoothing is performed in units of several meters as an example. Thereby, it is possible to remove local fluctuations. The drop-in rate analysis result 133 stores the smoothed drop-in rate. The smoothed drop-in rate will be described later with reference to FIG.

  In the present embodiment, it is possible to acquire or visualize the drop-in rate that reduces the influence of the characteristics of the displayed product. That is, in this embodiment, the drop-in rate determined by factors other than the product characteristics, such as the shape of the store, the entrance / exit arrangement, and the product shelf arrangement, is extracted. For this reason, the two-dimensional low-pass filter process is performed as described above. With the two-dimensional low-pass filter, it is possible to smooth a local drop-in rate variation (for example, a high drop-in rate around a popular product) that depends on the arrangement of the product. The fact that the influence of the drop-in rate due to the product characteristics can be reduced also means that the drop-in rate distribution obtained by this embodiment is not significantly affected even if the arrangement of the product is changed.

  The characteristics of the low pass filter may be determined by the size and shape of the store, the characteristics of the product, and the display method. A filter (for example, 3 to 5 meters) that makes the product distribution invisible is assumed, assuming that the range in which one kind of product is arranged is several tens of centimeters to 1 meter wide, as in an ordinary supermarket or general store. That's fine. However, if the resolution is lowered too much, the distribution of the drop-in rate becomes invisible, so the filter characteristics should be such that necessary information can be obtained in consideration of the size of the store.

Next, the merchandise arrangement optimization processing unit 104 will be described. The product arrangement optimization processing unit 104 includes a calculation process 141 for a place with a high drop-in rate, a search process 142 for a non-purpose purchase product, and a calculation process 143 for a product arrangement change candidate.
In the calculation process 141 for a place where the drop-in rate is high, rearrangement is performed in descending order from the smoothed drop-off rate. For example, the shelves are sorted in the order of the smoothing stoppage rate shown in FIG.
In the non-purpose purchase merchandise search process 142, a merchandise shelf having a long customer visit time and a low after-stop purchase rate is searched. For this reason, in this embodiment, the determination result obtained by the ratio analysis process 123 is used. The digitized determination result is stored in the after-stop purchase rate analysis result database 124. For the search, the stored data obtained by quantifying the characteristic that “the customer's visit time is long and the purchase rate after visit is low” is rearranged in descending order.
In the product placement change candidate calculation processing 143, the customer's drop-in time is long, the purchase rate after drop is low, the customer's drop-in rate is low, the customer's drop-in time is short, the post-stop purchase rate is high, A product shelf with a high drop-in rate is calculated. As a simple realization example, the product of the product characteristic quantified by the ratio analysis process 123 described above and the smoothed reciprocal of the stop-off rate are calculated for each shelf, and rearranged in descending order. At this time, each value may be weighted. In this way, a product shelf with a long customer visit time, a low after-stop purchase rate, and a low customer visit rate has a large value.

  The reason why an increase in sales can be expected with the above configuration will be described with reference to FIG. In FIG. 2, the horizontal axis represents the customer drop-in rate, and the vertical axis represents the post-stop purchase rate, and the characteristics are plotted for each product. Here, the product A is a product with a large number of purpose purchases, a product with a high customer drop-in rate, and a high purchase rate after drop-in. On the other hand, the product B is a product with a large number of non-purpose purchases, and is a product with a low customer drop-in rate and a low post-stop purchase rate. The result of replacing these two types of product arrangements is the right side of the figure. Although the product A has been moved to a place with a low drop-in rate, it is a product with a large number of purpose-purchased items, indicating that the drop-in rate of customers does not decrease much. In addition, since the purchase rate after dropping does not change significantly, the sales of the product A do not change so much. Here, since the post-stop purchase rate depends on the characteristics of the product, it is considered that the change in the arrangement does not change greatly. On the other hand, since the product B has moved to a place with a high drop-in rate, the drop-in rate increases. Although the after-stop purchase rate does not change significantly, the sales of the product B increases as the drop-in rate increases. Thus, the sales as a store increase by changing the arrangement of the two types of products.

  With reference to FIG. 3, a method for determining whether a product has a large amount of purpose purchases or a non-purpose purchase is described. FIG. 3 is a graph in which the horizontal axis represents the customer visit time and the vertical axis represents the post-stop purchase rate. Here, the purpose buying pattern is characterized by a short drop-in time and a high after-stop purchase rate. A typical example is a case where what is going to be purchased is decided before coming to the store, and the decision to purchase a product is quick at the store and the purchase rate is high. On the other hand, the non-purpose buying pattern is characterized by a long drop-in time and a low after-stop purchase rate. A typical example is a case where purchase of a product is not decided before coming to the store, and purchase is decided after seeing the product at the store. The time until the decision is long and the purchase rate is low. Here, the purpose purchase and the non-purpose purchase are called, but it is not always necessary to accurately determine the purpose purchase or the non-purpose purchase. The reason for this is that the product placement and exchange shown in FIG. 2 can be expected to increase sales by moving to a place where the drop-in rate is high if the drop-in time is long and the purchase rate is low. In other words, regardless of whether or not it is non-purpose purchase, it is possible to expect an increase in sales by improving the arrangement of products that are confused for purchase. In other words, the purpose purchase is when the drop-in time is short and the purchase rate after drop is high, and the case when the drop-in time is long and the purchase rate is low is considered non-purpose purchase.

A method for finding a place with a high drop-in rate will be described with reference to FIG. FIG. 4 is a floor map in which a shelf with a high drop-in rate for each shelf is painted in a dark color, and a shelf with a low drop-in rate is painted in a light color. The drop-in rate of the merchandise shelf is affected by the characteristics of the merchandise and the characteristics of the shelves in the store. The characteristic of the product is, for example, a high drop-in rate for daily necessities. Further, the shelf arrangement characteristics include, for example, a high drop-in rate near the main flow line. As described above, the drop-in rate is greatly different for each shelf due to various factors. If this is the case, there is a problem that it is difficult to determine where to move because the drop-in rate changes due to the movement of the product when the product arrangement is locally replaced. Therefore, two-dimensional low-pass filter processing is performed to perform smoothing on a plane. The smoothed drop-in rate is the lower diagram of FIG. By smoothing, the place where the drop-in rate is high becomes clear, and it can be seen that there are three places here. By using the smoothed drop-in rate, even if the product arrangement is locally changed, the change in the drop-in rate due to the product movement can be kept small.
A method of measuring the stop-by time for each shelf will be described with reference to FIG. FIG. 5 is a method of measuring the time at which the vehicle stopped by the communication means using infrared rays. Here, reference numeral 501 denotes a flow line detection sensor device worn by a customer who visits the store, 502 denotes an infrared beacon that transmits an ID by infrared, and 503 denotes a product shelf. In this embodiment, the flow line detection sensor device 501 includes an infrared sensor, an ID reception function, a memory, and the like. By arranging an infrared beacon 502 that transmits different IDs for each product shelf, it is possible to measure which shelf the traffic line detection sensor device 501 has received, and how long it has stopped. is there. When infrared rays are used for the communication means, the directivity of infrared rays is strong, so that it is possible to measure the time when the customer is facing the product shelf depending on whether infrared rays are detected. By using this time as a drop-in time, it is possible to exclude a time when the vehicle simply passes by or is doing another thing without facing the product shelf. For this reason, it is possible to more accurately measure the time when there is a high possibility that the user is at a loss in purchasing the product. The measurement result is accumulated in the memory of the flow line detection sensor device 501.

  A data collection method will be described with reference to FIG. FIG. 6 shows a procedure in which a customer at the store wears the flow line detection sensor device 501 and collects after shopping. First, a request for mounting the flow line detection sensor device 501 is made. The flow line detection sensor device 501 has a form that is lowered from the neck with a string, for example. This asks the customer for survey cooperation at the store. Next, you will have a normal shopping. Here, the flow line detected by the flow line detection sensor device 501 and the stay situation are measured. The measured data is temporarily stored in a memory in the flow line detection sensor device 501. A typical example of the measured data is the coordinate data in the store and data indicating the time existing at the coordinates. Or the data which shows the goods shelf number in a store and the time when the customer existed around the goods shelf may be sufficient. Next, the flow line detection sensor device 501 is connected to the cradle at the cash register, and the action data accumulated in the memory is collected. On the other hand, at the cash register, purchase data including a product ID and price data is collected from tags attached to the product. These data are accumulated in the measurement result database 101, and the action data and the purchase data can be matched. This can be verified later with the POS data, for example by measuring the passage time of the cash register, or keeping a customer's membership card number, or keeping a receipt number. Alternatively, you prepare the database to indicate which items in advance which product shelf has been display. With these, it is possible to check whether or not a product has been purchased for each shelf. Finally, the flow line detection sensor device is collected at the time of leaving the store. After collection, the data in the sensor device is deleted.

  In the above example, the method of collecting the data in the flow line detection sensor device 501 at the cash register together is shown. As another example, the flow line detection sensor device 501 has a transmission function, the information collected by the flow line detection sensor device 501 is sent to the management server in real time, and the sent information is stored in the measurement result database 101. You may comprise.

  The method for measuring the stop-by time of each shelf is not limited to infrared communication means, but location measurement using RFID, distance measurement using laser, image recognition using a camera, and the like may be used.

  FIG. 7 is summary data for each customer as a result of performing in-store behavior measurement. This data is stored in the measurement result database 101 of FIG. A unique customer ID is attached to each customer, and data such as store entry time, store exit time, and stay time in the store, which is the difference between them, is recorded. In addition, sex, age group, etc. may be recorded as personal information. Moreover, you may implement a questionnaire and summarize the result.

  FIG. 8 shows data on the drop-in time for each customer as a result of the in-store behavior measurement. This data is stored in the drop-in time database 121 of FIG. Assign a unique customer ID to each customer. Individual names are assigned to the shelves. Here, “shelf 1”, “shelf 2”, “shelf 3”,... The data of the time of stopping at each shelf is recorded for each customer. It can be seen that the customer ID 1234 has a long drop-in time of the shelf 6.

  FIG. 9 is data on the purchase amount for each shelf for each customer as a result of the in-store behavior measurement. This data is stored in the purchase price database 122 of FIG. Assign a unique customer ID to each customer. Individual names are assigned to the shelves. Here, “shelf 1”, “shelf 2”, “shelf 3”,... Data on the amount of money purchased on each shelf for each customer is recorded. It can be seen that the customer ID 1234 has a high purchase price for the shelf 8.

  FIG. 10 shows data on the number of purchase points by shelf for each customer as a result of measuring the behavior in the store. This data is stored in the purchase price database 122 of FIG. 1 instead of the data of FIG. 9 or together with the data of FIG. Assign a unique customer ID to each customer. Individual names are assigned to the shelves. Here, “shelf 1”, “shelf 2”, “shelf 3”,... Data on the number of products purchased on each shelf for each customer is recorded. It can be seen that the customer ID 1234 has purchased one item each of the shelf 1 and the shelf 8.

  FIG. 11 is data of the drop-in rate for each customer and the purchase rate after the drop-down of the analysis result of the in-store behavior measurement. This data is stored in the after-stop purchase rate analysis result database 124 of FIG. Individual names are assigned to the shelves. Here, “shelf 1”, “shelf 2”, “shelf 3”,... Data on the drop-in rate and purchase rate after drop is recorded for each shelf. Here, the drop-in rate is a value obtained by dividing the number of customers visiting the store among the customers who have measured the number by the number of customers visiting the store. Further, the after-stop purchase rate is a value obtained by dividing the number of customers who have purchased a product on the shelf among the customers who have been measured by the number of customers who have visited the shelf among the customers who have been measured.

  FIG. 12 is data of the drop-in rate and the smoothed drop-in rate for each customer in the analysis results of the in-store behavior measurement. This data is stored in the drop-in rate analysis result database 133 of FIG. Individual names are assigned to the shelves. Here, “shelf 1”, “shelf 2”, “shelf 3”,... The data of the drop-in rate and the smoothed drop-in rate are recorded for each shelf. Here, the drop-in rate is a value obtained by dividing the number of customers visiting the store among the customers who have measured the number by the number of customers visiting the store. Further, the smoothed stoppage rate is a value obtained by performing a two-dimensional filter process with the stoppage rate arranged in a plane.

  FIG. 13 is data of a product arrangement list as an analysis result of performing in-store behavior measurement. This data is stored in the drop-off product arrangement list database 105 of FIG. The target product, the replacement product candidate, and the expected value for improving the customer unit price are shown in the order in which the sales increase can be expected by exchanging the product arrangement. Here, the expected value of sales improvement is shown by the unit price of customers. As a simple example, data calculated in the product arrangement change candidate calculation process 143 and rearranged in descending order of the values are used. As described above, in the calculation process 143, calculation is performed so that a product shelf with a long customer visit time, a low purchase rate after visit, and a low customer visit rate has a large value. That is, a product that is highly likely to be a non-purpose purchase product is placed on a product shelf with a low drop-in rate, and is ranked higher. In FIG. 13, the products A, B, and C are cases where a large value is obtained by the calculation process 143, and the products X, Y, and Z are cases where a small value is obtained. Here, for example, it is shown that 160 yen improvement can be expected as a customer unit price by switching the arrangement of the product A and the product X. Here, product A is a product with a long drop-in time but a low purchase rate after drop-off, and a low drop-in rate after smoothing, and product X has a short drop-in time but a high purchase rate after drop-off, and a drop-off after smoothing It is a product with a high rate.

  Here, how to obtain the expected value for improving the customer unit price will be described. The arrangement of the product A and the product X is exchanged. It is assumed that the drop-in rate of the product A becomes the drop-in rate of the product X from the value before the rearrangement. Assuming that the after-stop purchase rate does not change, multiply the difference between the drop-in rate for product X and the drop-in rate for product A by the post-stop purchase rate for product A, and then multiply the product unit price to find an increase in customer unit price. . Here, it is assumed that the sales of the product X do not change.

I will supplement each assumption. First, it was assumed that the drop-in rate of the product A is the drop-in rate of the product X from the value before the rearrangement. The assumption is reasonable because it does not change significantly. Next, the post-stop purchase rate is assumed not to change, but the post-stop purchase rate largely depends on the characteristics of the product, and the assumption is reasonable because it does not change greatly in the product layout. Finally, it is assumed that the sales of the product X do not change. However, since the product X has a short stop-off time and a high purchase rate after the stop-off, it is presumed that the product X has a lot of purpose buying. For this reason, it is considered that sales do not depend greatly on product placement, and the assumption is reasonable.
As described above, it is possible to expect an improvement in the sales of the store by rearranging the products based on the list of FIG.

As described above, a product with a large amount of purpose-buying rarely decreases in both the drop-in rate and the post-stop purchase rate even if it is brought to a place where the drop-in rate of the shelf is not high. On the other hand, if you bring products with many non-purpose purchases to a place with a high drop-in rate of shelves, the number of purchases can be increased and sales can be expected to improve.
According to this example, products with a short drop-in time and a high after-stop purchase rate tend to be purpose-purchased, whereas products with a long drop-in time and a low after-stop purchase rate tend to be products with many non-purpose purchases. As a result of the discovery, it is possible to automatically determine whether a product should be placed in a place with a high drop-in rate or a product should be placed in a place with a low drop-in rate.
According to the behavior analysis apparatus of the present embodiment, products that are not intended for purchase can be found in the store, and new value can be provided to the customer in the store.
In the present invention, the evaluation method or the quantification method of each element of the shelf on which the product is displayed, the length of the drop-off time, the size of the purchase rate after the drop-off, the size of the drop-off rate after the smoothing is as described above. It is not limited to examples. Various other statistical methods can also be used. Also, each element can be weighted. In short, if a specific product is placed in a specific product shelf (location), is the product shelf a product with a long stop-off time but a low purchase rate after the stop and a low stop-off rate after smoothing? However, it is only necessary to be able to evaluate the tendency of goods shelves with a short drop-in time but a high purchase rate after drop-in and a high drop-in rate after smoothing.

In the first embodiment (FIG. 1), an example is shown in which the processing up to calculation of the product arrangement change candidate is performed based on the collected data.
Another embodiment of the present invention is a drop-in rate analysis system in which the measurement result database 101 and the drop-in rate analysis processing unit 103 are the main components in the configuration of FIG. The other configurations in FIG. 1 can be omitted. That is, the drop-in rate analysis system of the present embodiment can extract points with high drop-in rates shown in the lower part of FIG. This drop-in rate is smoothed as described above, and the influence of the product characteristics is reduced. Therefore, even if the place where the product is placed is changed, it is hardly affected. By displaying the image shown in FIG. 4 on the output device, the manager can grasp a point at which the customer is likely to stop, so that it can be used as a reference for the product arrangement.

Another embodiment of the present invention is a post-stop purchase rate analysis system having the measurement result database 101, post-stop purchase rate analysis processing unit 102, and non-purpose purchase product search processing 142 in the configuration of FIG. It is. The other configurations in FIG. 1 can be omitted. In other words, the post-stop purchase rate analysis system of this embodiment classifies the product characteristics based on the customer's stop time and post-stop purchase rate from the data of the stop time database 121 and the purchase price database 122.
As a method for extracting merchandise shelves with a long customer visit time and a low purchase rate after visit, Example 1 shows a method of sorting trends by quantifying trends. As another method, one or a plurality of threshold values are provided for the stop time and the post-stop purchase rate, and the product characteristics can be classified by the threshold values. For example, a product with a stop time of 30 seconds or less and a post-stop purchase rate of 80% or more is flagged as a target purchase product. In addition, products with a stop time of 3 minutes or more and a post-stop purchase rate of 30% or less are flagged as non-purpose purchase products. These flags are associated with the product type ID to create a database and stored in the storage device.
The contents of the database can be displayed on the output device at any time. In the display, a product with a long stop-off time and a low after-stop purchase rate and a product with a short stop-off time and a high after-stop purchase rate can be displayed in an identifiable manner. In this way, since the product characteristics can be grasped, it can be arranged on the product shelf according to the characteristics of the products, and sales can be improved.
In the above-described embodiment, the arrangement of products on the product shelves in the store has been described as an example. The concept of the present invention is not limited to this, and can be applied, for example, to store-type arrangements in stores in shopping malls and shopping streets. That is, even in the case of a store in a shopping mall, the drop-in rate varies from store to store depending on the structure of the shopping mall, and depending on the product trend handled by each store, there is a strong tendency to purchase products (for example, a supermarket that handles daily necessities) This is because it is considered that there are stores and other stores (for example, souvenir shops).
In the above description, the position, size, shape, range, and the like of each component illustrated in the drawings and the like may not represent the actual position, size, shape, range, or the like in order to facilitate understanding of the invention. . For this reason, the present invention is not necessarily limited to the position, size, shape, range, and the like disclosed in the drawings and the like.
The present invention is not limited to the embodiments described above, and includes various modifications. For example, a part of the configuration of one embodiment can be replaced with the configuration of another embodiment, and the configuration of another embodiment can be added to the configuration of one embodiment. In addition, it is possible to add, delete, and replace other configurations for a part of the configuration of each embodiment.

It is possible to automatically provide information that is effective for examining the display arrangement of products in a store and the arrangement of stores in a shopping street.

101 Measurement Result Data 102 Post-Stop Purchase Rate Analysis Processing 103 Stop-off Rate Analysis Processing 104 Product Placement Optimization Processing 105 Product Placement List Data 501 Flow Line Detection Sensor Device 502 Infrared Beacon 503 Product Shelf

Claims (10)

In the store, equipped with a device that measures the stop time that is the time stopped in front of the product shelf, and the after-stop purchase rate that is the ratio of the number of people who stopped at the product shelf and the number of people who purchased the product on the product shelf A behavior analysis device,
A device for applying a two-dimensional filter that obtains a drop-in rate that is the ratio of the number of customers visiting the store and the number of people who visited the product shelf for each product shelf, and smoothing the obtained drop-in rate on a plane in the entire store,
A behavior analysis apparatus comprising: a device that calculates a product shelf having a long drop-in time, a low after-stop purchase rate, and a low drop-in rate after the two-dimensional filter processing. The behavior analysis device according to claim 1,
A behavior analysis apparatus comprising: a device that calculates a product shelf having a short drop-in time, a high post-stop purchase rate, and a high post-filter drop-in rate. The behavior analysis device according to claim 1,
The behavior analysis device is characterized in that the stop-off time is measured based on a time of stopping in front of a product shelf detected by infrared communication. In the store, equipped with a device that measures the stop time that is the time stopped in front of the product shelf, and the after-stop purchase rate that is the ratio of the number of people who stopped at the product shelf and the number of people who purchased the product on the product shelf A behavior analysis device,
Obtain a drop-in rate that is the ratio of the number of visitors to the store shelf and the number of people who visited the product shelf for each product shelf, and equipped with a device that smoothes the obtained drop-off rate in a flat manner throughout the store,
A first product shelf with a long drop-in time, a low purchase rate after the drop-in, and a low drop-in rate after the smoothing;
A second merchandise shelf with a short drop-in time, a high after-stop purchase rate, and a high drop-in rate after smoothing;
A behavior analysis apparatus comprising: a device for calculating an expected value for improving a customer unit price when a product in the first product shelf and a product in the second product shelf are exchanged. The behavior analysis apparatus according to claim 4,
The behavior analysis device is characterized in that the stop-off time is measured based on a time of stopping in front of a product shelf detected by infrared communication. The behavior analysis apparatus according to claim 4,
The apparatus for performing smoothing performs a two-dimensional low-pass filter process. An apparatus comprising an input device, an output device, a processing device, and a storage device,
The input device has a function of inputting behavior history information from a portable sensor carried by a customer who has entered a predetermined geographical area while staying in the predetermined geographical area;
The behavior history information includes information on continuous or discrete time and place,
The processing device has a function of extracting stop information for each location indicating whether or not each customer has stopped at a plurality of predetermined locations in the geographical range based on the behavior history information,
The processing device maps, for each of the plurality of predetermined places, a drop-in rate that is a ratio of the total number of visitors and the number of customers who have stopped at the predetermined place, and further smoothes the drop-off rate two-dimensionally Has the function to
The processing device displays the smoothed drop-in rate data on the output device, stores the data in the storage device, or both. The behavior analysis apparatus according to claim 7,
The processing device has a function of extracting the stop time information indicating the length of the drop-in time at which each customer has stopped at the plurality of predetermined locations based on the behavior history information for each location,
The input device has a function of inputting purchase information indicating, for each location, whether each customer has purchased a product arranged in the plurality of predetermined locations,
The processing device calculates a post-stop purchase rate based on the purchase information and the action history information, which is a ratio of the number of people who have stopped at the specific place and the number of people who have purchased the products arranged at the predetermined place. Has the function to
Based on the length of the drop-in time and the purchase rate after the drop-off, the products arranged in the plurality of predetermined locations are labeled and classified or sorted, or both of them are performed. A behavior analysis device characterized by The behavior analysis device according to claim 8,
Based on the length of the drop-in time and the purchase rate after the drop-in, the product having a long drop-in time and a low purchase rate after the drop-in, and a short drop-in time length and the purchase rate after the drop-in A behavior analysis apparatus that displays high products on the output device in an identifiable manner. The behavior analysis device according to claim 8,
Product shelves with infrared transmitters are arranged in the plurality of predetermined locations,
The portable sensor comprises an infrared receiver;
According to the action history information indicating that the customer carrying the portable sensor faces the product shelf and the infrared receiver detects the infrared ray from the infrared transmitter, Determine whether you have visited multiple predetermined locations,
The processing device is configured such that the customer carrying the portable sensor faces the merchandise shelf, and each customer has the action history information indicating the length of time that the infrared receiver detects infrared rays from the infrared transmitter. A behavior analysis apparatus characterized by determining the length of a stop-by time when the user stops at the plurality of predetermined locations.

Images