Sales promotion support equipment

Abstract

To make it easy to obtain sales promotion contents that match a sales strategy for a commodity, and to easily obtain other sales promotion contents even if the contents do not lead to expected results. [Solution] The system has a first content generation means that inputs one of the products for sale and one of the features into the trained machine learning model (including a form in which the trained machine learning model references an external database) generated by a model generation means that generates a machine learning model that outputs the sales promotion content when the products for sale and the features are input by training a training dataset that is a combination of the products for sale, features based on management-related data, and sales promotion content, and generates one of the sales promotion content.

Description

[Technical Field]

[0001] Regarding sales promotion techniques. [Background technology]

[0002] For companies, increasing sales is an important element that supports the foundation of management and is essential for achieving sustainable growth. Therefore, as part of a strategy aimed at increasing sales, the introduction of sales promotion content that stimulates customer purchasing desire is considered. This is expected to effectively convey the appeal of products and services and increase competitiveness in the market.

[0003] Under such circumstances, with regard to sales promotion technology, for example, Patent Document 1 proposes a technology for appropriately providing content for promoting sales of recommended products. [Prior art documents] [Patent documents]

[0004] [Patent Document 1] Patent No. 7671895 Summary of the Invention [Problem to be solved by the invention]

[0005] However, the above-mentioned conventional technology has a problem in that it is difficult to easily obtain the intended sales promotion content.

[0006] In view of the above problems, the present invention aims to provide a sales promotion support device that makes it easy to obtain sales promotion content that matches a product's sales strategy, and that makes it easy to obtain other sales promotion content even if that content does not lead to the expected results. [Means for solving the problem]

[0007] One embodiment of the disclosed sales promotion support device includes a first content generation means for inputting one of the target products for sale and one of the feature quantities into a trained machine learning model (including an embodiment in which the trained machine learning model references an external database) that generates the one of the target products for sale and one of the feature quantities by training a training data set that is a combination of the target products for sale and feature quantities based on management-related data and sales promotion content, and generating the one of the sales promotion content. a content implementation means for performing processing to implement the one sales promotion content in sales channels including any one or more of brick-and-mortar stores, online sales, mail-order sales, catalog sales, television shopping, live commerce, sales via mobile apps, and sales via SNS for the one sales target product; an evaluation information acquisition means for acquiring sales performance for the one sales target product from a management information system; a performance evaluation means for comparing the sales performance with a predetermined target value and determining whether the target has been achieved; a model re-learning means for, after generating the one sales promotion content, re-learning the trained machine learning model by training the training dataset corrected by data acquired after processing by the model generation means; a second content generation means for, if the target has not been achieved, inputting the one sales target product and the one feature into the trained machine learning model after the re-learning (including a form in which the trained machine learning model references an external database), and generating another sales promotion content; and a model storage means for storing parameters of the trained machine learning model. The present invention is characterized by having the following. [Effects of the Invention]

[0008] The disclosed sales promotion support device makes it easy to obtain sales promotion content that matches a product's sales strategy, and even if that content does not lead to the expected results, makes it easy to obtain other sales promotion content. [Brief explanation of the drawings]

[0009] [Figure 1] 1 is a diagram showing an overview of a sales promotion support device according to an embodiment of the present invention; [Figure 2] 1 is a functional block diagram of a sales promotion support device according to an embodiment of the present invention. [Figure 3] 1 is a diagram illustrating an example of the hardware configuration of a sales promotion support device according to an embodiment of the present invention. [Figure 4] 10 is a flowchart showing a flow of an example of processing by the sales promotion support device according to the present embodiment. DETAILED DESCRIPTION OF THE INVENTION

[0010] DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS An embodiment of the present invention will be described with reference to the drawings. (Operation principle of the sales promotion support device according to this embodiment)

[0011] The operating principle of a sales promotion support device 100 according to this embodiment (hereinafter simply referred to as "this device") will be described using Figures 1 and 2. Figure 1 is a diagram showing the connection relationship between this device 100 and other devices, and Figure 2 is a functional block diagram of this device 100.

[0012] As shown in FIG. 1, the device 100 is connected to a management information system 290 via a communication network 310. The communication network 300 may be either wired or wireless. The management information system 290 is a system that stores and provides sales records 280 related to products 210 for sale, and the sales records 280 are, for example, information related to sales amounts, sales volume, etc. The management information system 290 may be, for example, a POS (Point of Sales) system.

[0013] As shown in Figure 2, the present device 100 has a model storage means 110, a learning data storage means 120, a model generation means 130, a first content generation means 140, a content implementation means 150, an evaluation information acquisition means 160, a result evaluation means 170, a model re-learning means 180, and a second content generation means 190. The model storage means 110 stores model parameters that define the operation of the trained machine learning model 260, which will be described later.

[0014] The training data storage means 120 stores a training data set 250, which is training data for training the machine learning model 260 and is a combination of the products for sale 210, features 220 based on management-related data, and sales promotion content 240. Note that the features 220 based on management-related data may be features based on purchase data 230.

[0015] The features 220 based on management-related data include, for example, sales and profit-related data (sales data, profit data), sales and product management-related data (product master and specifications, shelf layout data, sales results and sales promotion effects), inventory and purchasing-related data (inventory status, order and purchasing data), customer and marketing-related data (customer ID data (ID-POS), marketing measure results), store operation and personnel management-related data (store performance, staff and shifts), expense and finance-related data (store and headquarters expenses, financial indicators), external environment data (trade area / population data, location and selling price information of competing stores, weather, temperature, disaster information, local event / school event information), and instruction and communication-related data (work instructions and policies from superiors, reports and proposals from the field, collaboration history with headquarters (Slack, etc.)).

[0016] Sales data includes sales by product / category, store / time period, average customer spend / number of items purchased, and monthly / weekly sales trends. Profit data includes gross profit / gross profit margin (by SKU / category), impact of discounts and disposal, and profit impact rankings (high-margin products). Product master specifications include product code / classification / cost / selling price / gross profit margin, manufacturer / supplier / order unit information. Sales performance and promotional effectiveness include sales volume trends (by SKU), response to flyers / apps / coupons, and adoption rates of trending and new products. Inventory status includes inventory quantity / inventory value by SKU, inventory turnover / retained inventory list, and inventory difference data (theoretical inventory vs. actual inventory). Ordering and purchasing data includes order history / order frequency / delivery delays, order quantity vs. sales quantity (accuracy), and automatic order correction history and rationale.

[0017] In addition, customer ID data (ID-POS) includes frequency of visits / purchase history / average purchase amount, purchasing trends by age and gender, and tendency to sell specific products together (basket analysis), while marketing measures results include coupon usage rate / return visit rate, sales effect by campaign, number of responses and evaluations on SNS / reviews. Store performance includes sales by store / comparison with previous year / square meter efficiency, number of customers visiting / number of transactions processed at the register.

[0018] Includes store satisfaction (surveys and reviews), staff and shifts include working hours / operating efficiency by staff, staffing by time period sales, and training history / work proficiency. Store and headquarters expenses include trends in costs such as utility costs, sales promotion costs, disposal losses, labor costs / outsourcing costs, consumables and repair costs, and financial indicators include management indicators such as profit and loss statement (PL) items, cash flow, ROA and ROE.

[0019] In addition, trade area and population data includes surrounding population / household attributes, topography, and foot flow data; competitor location and selling price information includes distance, number of stores, flyer information, and sale information; and external factors include weather, temperature, disaster information, and local / school event information. Work instructions and policies from superiors include instructions to increase sales of key products, specific guidance on reducing waste (e.g., within a certain percentage), instructions to reduce shifts / labor costs, and requests to change the sales floor layout. Reports and proposals from the field include complaints such as "Product X is not selling well" or "POP is ineffective," reports on results after shelf rearrangements, and sharing of reasons for stockouts and ordering problems. Collaboration history with headquarters includes email / chat / Slack exchanges, the contents of weekly, monthly, and daily reports, and the contents of meeting materials and reports.

[0020] Sales promotion content 240 includes, for example, advertising (creative (digital, in-store), media, target, schedule, ROI, live commerce planning), and sales floor promotion (standard, end, selling price, points, coupons, overall store layout). Creative generation (digital and in-store) includes visual expression (generation of product image variations (composition, color, texture), generation of illustrations / photo materials that match the brand's worldview, output of layout design proposals for in-store POP, banners, and flyers), catch phrases and wording (target-specific catch phrases (e.g., for young people, wealthy people, etc.), proposals for product benefit appeals, short phrases for posting on social media (e.g., 140 characters or less)), and video and audio advertising elements (scenario composition (storyboard), narration script, AI-generated proposals for in-store announcements).

[0021] Media selection generation includes online media (selection of advertising distribution platforms (e.g., Meta, Google, TikTok), a list of recommended channels for each target, and suggestions for cost-effective advertising destinations (by purpose)), offline media (utilization of the four mass media, effective media by region such as in-store and transit advertising, plans linked to exhibitions and events, and analysis of the degree of match between local TV and print media). Target setting generation includes persona generation (generation of detailed personas based on age, gender, occupation, values, etc., extraction of new personas based on analysis of existing customers, and association with customer behavior models (purchase funnels)), target extraction methods (clustering analysis from existing customer data, extraction of topics of interest from social media / search data, extraction of local targets based on geographic information and trade area data), and segment analysis (demand forecasts by region and age, interest analysis based on behavioral history and social media trends, and suggestions for similar audiences).

[0022] In addition, advertising schedule generation includes campaign calendars (automatic generation of schedules in line with product launches, distribution plans optimized for days of the week, seasons, and events, and creative rotation design including measures to prevent advertising fatigue), automatic optimization scenarios (automatic adjustment proposals for time-zone distribution based on response rates, proposals for mid-term review timing during the flight period, and output of on / off switching criteria), while ROI setting generation includes KPI model proposals (weighting proposals for impressions / clicks / conversions, simulation of ROAS (ad spend vs. sales) target values, generation of advertising evaluation criteria from an LTV perspective), and budget allocation models (return on investment predictions by channel, dynamic reallocation scenarios for highly effective measures, and cost allocation models for CPA optimization).Live commerce plan generation includes script generation (structure, performer selection, product explanation), timing judgment, and target setting.

[0023] In addition, the product composition for standard shelves includes product selection by category (top-selling products, trending products, regional / seasonal limited items), shelf allocation optimization (automatic calculation of face count, zoning (by price range, by target)), and inventory / order linkage (linked optimization with ordering points, replacement proposals for stagnant products). The product composition for end shelves includes theme design (seasonal campaigns, events (Halloween, Valentine's Day, etc.), food scene proposals (breakfast, parties, quick cooking)), related product set proposals (cross MD (linking different categories), pairing proposals based on concurrent sales results), and sales floor area optimization (shelf number allocation based on sales efficiency, display presentation (POP, fixtures)). The overall store layout includes shelf arrangement and flow design.

[0024] Sales price setting generation includes price proposals based on competitor comparisons (utilizing online price monitoring data and linking with data from local competitor stores), price elasticity analysis (estimating response rates from past sales promotion results and proposing price ranges to maximize profits), and automatic generation of discount timing (dynamic price changes based on expiration dates and inventory status, optimal price setting for different time periods (night sales, etc.)).

[0025] Points are awarded through the selection of eligible products (new products / repeater promotion targets, ranking by sales contribution), optimization of the award rate (setting return rates based on gross profit margins and purchase frequency, linked to campaign budget allocation), and membership rank linkage (special points for VIPs, special returns for new / dormant customers). Coupons include personalized coupon generation (product suggestions + coupons based on purchase history, different content depending on store visit frequency), and coupon condition design (conditions based on basket amount, time-limited / day-limited coupons, store-limited coupons).

[0026] The model generation means 130 trains the machine learning model 260 on a training data set 250, which is a combination of the product for sale 210, the feature 220 based on the management-related data, and the sales promotion content 240. The model generation means 130 thereby generates the machine learning model 260 that outputs the sales promotion content 240 when the product for sale 210 and the feature 220 based on the management-related data are input. Note that the learning algorithm used by the model generation means 130 is not particularly limited.

[0027] The first content generation means 140 inputs one product for sale 210 and one feature 220 based on business-related data into the trained machine learning model 260 (including a form in which the trained machine learning model 260 references an external database) generated by the model generation means 130, and generates one sales promotion content 240. Note that the external database is not particularly limited.

[0028] The content implementation means 150 performs processing to implement one sales promotion content 240 generated by the first content generation means 140 in a sales channel 270, such as a physical store, online sales, mail order, or other sales channel for one sales target product 210. The sales channel 270 refers to a general medium or means by which a product is provided or sold, including, but not limited to, a physical store, an e-commerce site, mail order, catalog sales, television shopping, live commerce, sales via mobile apps, sales via SNS, etc. The evaluation information acquisition means 160 acquires the sales performance 280 for one product 210 for sale from the management information system 290 . The result evaluation means 170 compares the sales performance 280 with a predetermined target value 300 to determine whether the target has been achieved.

[0029] The model re-learning means 180 re-learns the trained machine learning model 260 by training the training data set 250 corrected by data acquired after processing by the model generation means 130 after generating one sales promotion content 240. The processing start condition of the model re-learning means 180 is not limited to when the outcome evaluation means 170 determines that the goal has not been achieved, but is not limited to when the training data set 250 is corrected after processing by the first content generation means 140 and after processing by the model generation means 130. This is to allow the second content generation means 190, which will be described later, to generate another sales promotion content 240 different from the one sales promotion content 240. The learning algorithm used by the model re-learning means 180 is not limited to any particular condition.

[0030] The second content generation means 190 performs the following process when the outcome evaluation means 170 determines that the goal has not been achieved. The second content generation means 190 inputs one product for sale 210 and one feature 220 based on management-related data into the trained machine learning model 260 after re-training (including a form in which the trained machine learning model 260 references an external database), and generates another sales promotion content 240. It should be noted that even if the input to the trained machine learning model 260 is the same between the first content generation means 140 and the second content generation means 190, the output of the machine learning model 260 will be different. Note that the external database is not particularly limited.

[0031] Based on the above operating principle, the device 100 makes it easy to obtain promotional content 240 that matches the sales strategy of the product 210, and even if that does not lead to the expected results, it makes it easy to obtain other promotional content 240. (Hardware configuration of the sales promotion support device according to this embodiment)

[0032] An example of the hardware configuration of the present device 100 will be described using Fig. 3. Fig. 3 is a diagram showing an example of the hardware configuration of the present device 100. As shown in Fig. 3, the present device 100 has a CPU (Central Processing Unit) 510, a ROM (Read-Only Memory) 520, a RAM (Random Access Memory) 530, an auxiliary storage device 540, a communication I / F 550, an input device 560, a display device 570, and a storage medium I / F 580.

[0033] CPU 510 is a device that executes programs stored in ROM 520, performs arithmetic processing on data loaded into RAM 530 in accordance with program instructions, and controls the entire device 100. ROM 520 stores programs and data to be executed by CPU 510. When CPU 510 executes a program stored in ROM 520, the programs and data to be executed are loaded into RAM 530, and RAM 530 temporarily holds the arithmetic data during the calculation.

[0034] The auxiliary storage device 540 is a device that stores the OS (Operating System), which is basic software, the application program according to this embodiment, and other related data. The auxiliary storage device 540 is, for example, a hard disk drive (HDD) or flash memory, and includes the model storage means 110 and the training data storage means 120.

[0035] The communication I / F 550 is an interface for connecting to a communication network 310 such as a wired or wireless LAN (Local Area Network) or the Internet, and for transmitting and receiving data to and from another device 290 that provides a communication function.

[0036] The input device 560 is a device such as a keyboard for inputting data to the device 100. The display device (output device) 570 is a device formed of an LCD (Liquid Crystal Display) or the like, and functions as a user interface when the user uses the functions of the device 100 or when making various settings. The storage medium I / F 580 is an interface for sending and receiving data to and from a storage medium 590 such as a CD-ROM, DVD-ROM, or USB memory.

[0037] Each of the means included in device 100 may be realized by CPU 510 executing a program corresponding to each of the means stored in ROM 520 or auxiliary storage device 540. Each of the means included in device 100 may also be realized by hardware that performs the processing associated with the means. Alternatively, device 100 may be caused to execute the program by loading the program according to the present invention from an external server device via communication I / F 550 or from storage medium 590 via storage medium I / F 580. (Example of processing by the sales promotion support device according to this embodiment) An example of the flow of processing by the device 100 will be described with reference to Fig. 4. Fig. 4 is a flowchart showing the flow of processing by the device 100.

[0038] In S10, the model generation means 130 causes the machine learning model 260 to learn the training data set 250, which is a combination of the product for sale 210, the feature 220 based on the management-related data, and the sales promotion content 240. The model generation means 130 thereby generates the machine learning model 260 that outputs the sales promotion content 240 when the product for sale 210 and the feature 220 based on the management-related data are input.

[0039] At S20, the first content generation means 140 inputs one product for sale 210 and one feature 220 based on management-related data into the machine learning model 260 trained at S10 (including a form in which the trained machine learning model 260 references an external database), and generates one sales promotion content 240.

[0040] In S30, the content implementation means 150 performs processing to realize the implementation of one sales promotion content 240 generated in S20 in a sales channel 270 such as a physical store, online sales, mail order sales, or other sales channel for one sales product 210.

[0041] In S40, the evaluation information acquisition means 160 acquires the sales performance 280 for one product 210 for sale from the management information system 290, and the result evaluation means 170 compares the sales performance 280 with a predetermined target value 300 to determine whether the target has been achieved.

[0042] Here, after generating one sales promotion content 240, the model re-learning means 180 re-learns the trained machine learning model 260 by training the training dataset 250 that has been corrected using data acquired after processing by the model generation means 130. Note that the processing by the model re-learning means 180 is a processing system separate from the processing system in S10 to S50.

[0043] The processing start condition of the model re-learning means 180 is not limited to when the outcome evaluation means 170 determines that the goal has not been achieved, but is after the processing of S20 and S10, and after the training data set 250 has been modified. This is to allow the second content generation means 190 in S50 to generate another sales promotion content 240 different from the one sales promotion content 240.

[0044] In S50, the second content generation means 190 performs the following process if it is determined in S40 that the goal has not been achieved. The second content generation means 190 inputs one product for sale 210 and one feature 220 based on management-related data into the machine learning model 260 (including a form in which the trained machine learning model 260 references an external database) that has been retrained by the model retraining means 180, and generates another sales promotion content 240. It should be noted that even if the input to the machine learning model 260 is the same in S20 and S50, the output of the machine learning model 260 is different.

[0045] By performing the above-described processing, the device 100 makes it easier to obtain promotional content 240 that matches the sales strategy of the product 210, and even if that content does not lead to the expected results, it makes it easy to obtain other promotional content 240.

[0046] Although the embodiments of the present invention have been described in detail above, the present invention is not limited to such specific embodiments, and various modifications and variations are possible within the scope of the gist of the present invention as set forth in the claims. [Explanation of symbols]

[0047] 100 Sales promotion support equipment 110 Model storage means 120 Learning data storage means 130 Model Generation Method 140 First content generation means 150 Content Practice Methods 160 Evaluation information acquisition means 170 Outcome Evaluation Instruments 180 Model Retraining Method 190 Second content generation means 210 Products for Sale 220 Features based on business-related data 230 Purchasing Data 240 Promotional Content 250 training datasets 260 Machine Learning Models 270 sales channels 280 sales results 290 Management Information Systems 300 target value 310 Communication Network 510 CPU 520 ROM 530 RAM 540 Auxiliary storage 550 Communication Interface 560 Input Device 570 Output Device 580 Storage Media Interface 590 Storage medium

Claims

1. a first content generation means for inputting one of the target products for sale and one of the feature quantities into a trained machine learning model (including a form in which the trained machine learning model references an external database) generated by a model generation means that generates a machine learning model that outputs the promotional content when the target products for sale and the feature quantities are input by training a training dataset that is a combination of the target products for sale, feature quantities based on management-related data, and promotional content, and for generating one of the promotional content; a content implementation means for performing processing to realize the implementation of the sales promotion content in a sales channel including at least one of a physical store, online sales, mail order sales, catalog sales, television shopping, live commerce, sales via a mobile app, and sales via SNS for the one product for sale; evaluation information acquisition means for acquiring sales performance information for the one product to be sold from a management information system; A result evaluation means for comparing the sales performance with a predetermined target value and determining whether the target has been achieved; a model retraining means for retraining the trained machine learning model by training the training dataset corrected by data acquired after processing by the model generation means after generating the one sales promotion content; If the goal is not achieved, a second content generation means inputs the one product for sale and the one feature into the trained machine learning model after the re-training (including a form in which the trained machine learning model references an external database) and generates other sales promotion content; A sales promotion support device characterized by having a model storage means that stores parameters of the trained machine learning model.

2. A sales promotion support method carried out by a computer, comprising: a step of inputting one of the target products for sale and one of the feature quantities into the trained machine learning model (including a form in which the trained machine learning model references an external database) generated by a model generation means that generates a machine learning model that outputs the promotional content when the target products for sale and the feature quantities are input by a first content generation means that trains a training dataset that is a combination of the target products for sale, feature quantities based on management-related data, and promotional content, and generating one of the promotional content; a step in which a content implementation means performs processing to realize implementation of the one sales promotion content in a sales channel including at least one of a physical store, online sales, mail order sales, catalog sales, television shopping, live commerce, sales via a mobile app, and sales via SNS for the one sales target product; an evaluation information acquisition means acquiring sales performance information for the one product to be sold from a management information system; a step in which an achievement evaluation means compares the sales performance with a predetermined target value and determines whether the target has been achieved; a step in which a model retraining means retrains the trained machine learning model by training the training dataset corrected by data acquired after processing by the model generation means after generating the one sales promotion content; A sales promotion support method characterized by including a step in which, if the goal is not achieved, a second content generation means inputs the one product for sale and the one feature into the trained machine learning model after the re-training (including a form in which the trained machine learning model references an external database) and generates other sales promotion content.

3. A sales promotion support program for causing a computer to execute the method according to claim 2.

Images