KR20030043678A - Sales forecasting apparatus and sales forecasting method - Google Patents

Abstract

PURPOSE: To provide a sales prediction apparatus capable of making appropriate sales prediction while avoiding complexity of processing without reference to existence of sales achievements. CONSTITUTION: A host computer 6 receives incoming data and outgoing data from a computer 2 in a distribution center performing intensively entering/ dispatching and storing. Store register devices 4 are provided to each store in whole country and sends sales data to the host computer 6. The host computer 6 places an order to a computer in a manufacturer 3, receives above described data to summarize at a predetermined time every day and communicates to a database server 9 to update a database 8. A sale prediction computer 10 is connected to the database server 9 through a LAN and makes the sales prediction based on the data stored into the database 8.

Description

Sales forecasting apparatus and sales forecasting method {SALES FORECASTING APPARATUS AND SALES FORECASTING METHOD}

The present invention relates to a system for predicting sales of goods and services.

In order to eliminate the excess and shortage of a product, and to sell efficiently, the ordering quantity is determined by carrying out a sales forecast.

For example, Japanese Patent Laid-Open No. 5-120314 discloses the following sales prediction system. This system predicts future sales based on past sales of the target product. The precision is predicted by considering all the sales of the product for the last three months, the sales of the last three months of the whole country, and the sales of January (three months) before and after the same month last year.

However, this system is based on the premise that the same merchandise is sold in the previous year, and there is a problem that prediction cannot be made for a merchandise having no sales record. For this reason, there existed a problem that it could not apply to the case where the fluctuation | variation of the handled product was large, and the industry where the input of a new product was active. In addition, since each product is predicted by referring to past sales results of each product, when the number of products to be handled is large, the processing becomes complicated and there is a problem that it is difficult to realize.

Japanese Patent Laid-Open No. Hei 8-278959 discloses a system for predicting sales of products having no sales results. In this system, merchandise is classified in detail by attributes such as color, pattern, size, and grade, and the sales record is recorded. For products without sales results, the products with the most similar attributes are selected from those already recorded, and the sales are predicted based on the sales results of the products. In addition, sales forecasting is performed in time series, and when there exists a difference between a forecast and a performance, the forecast is corrected.

According to this system, it is possible to make a sales forecast for a product having no sales record. However, in order to increase the accuracy of the prediction, it is necessary to record detailed attributes of goods, and as a result, there is a problem that a large number of attributes must be recorded and accumulated.

SUMMARY OF THE INVENTION An object of the present invention is to provide a sales prediction apparatus capable of solving the above-described problems and performing appropriate sales forecasting, with or without sales results, and while avoiding complicated processing.

1 shows an entire system.

2 is a diagram illustrating an outline of a business model.

3 is a diagram illustrating a hardware configuration of a host computer.

4 is a diagram illustrating a flowchart of data aggregation processing.

5 is a diagram showing sales inventory data.

6 is a diagram showing sales inventory data.

7 is a diagram showing a hardware configuration of a sales prediction computer.

8 is a diagram illustrating a flowchart of a sales index creating process.

9 shows sales performance data.

10 is a diagram illustrating a calculation process of a main index.

11 is a diagram illustrating an example of a main index.

It is a figure which shows the flowchart of a sales prediction and order quantity calculation process.

Fig. 13 is a diagram showing the contents of condition setting.

14 is a diagram illustrating a flowchart of the first sales prediction process.

15 is a diagram illustrating sales forecast for each product.

16 is a diagram illustrating a flowchart of a sales prediction process at the time of addition.

17 shows performance data and main index data.

18 is a diagram illustrating sales prediction and ordering data.

19 is a diagram schematically illustrating a method for calculating an order amount.

20 is a diagram illustrating a flowchart of a required order quantity calculation.

<Explanation of symbols for main parts of drawing>

2: logistics center computer

3: maker

4: store register device

6: host computer

9: database server

10: Sales Prediction Computer

The sales prediction apparatus of this invention acquires the sales result of each transaction object, and based on the said sales result, the transaction which groups each transaction object and makes a forecast using the sales timing of each transaction object as at least one indicator, Based on the sales time of the target, it is judged which group the trading target for making the prediction belongs to, and the trading target group of the corresponding trading target in the past corresponding time having similarity with the timing for making the prediction for the trading target making the prediction. Based on the sales results, the sales forecast of the transaction target is calculated.

The sales prediction apparatus of the present invention is configured to allow access to a recording unit that records the sales performance of each group to the trading subjects grouped by using the sales timing of each trading subject as at least one indicator, and the trading subjects that make predictions. On the basis of the selling time of, it is judged which group the trading target for making the prediction belongs to, and the sales of the trading target group in the past corresponding time having similarity with the timing of making the prediction for the trading target for making the prediction. It is characterized by acquiring the results from the recording unit and calculating the sales forecast of the transaction target based on the sales results.

That is, each transaction object is grouped according to the sales time, and it is judged which group the transaction object to predict belongs to, and the sales prediction is made based on the sales performance of the group. Since the trade objects are grouped according to the sales time, the classification of the trade objects is clear and the prediction with high precision can be performed. In addition, even when the number of objects to be handled increases, there is no fear that the classification increases dramatically and the processing becomes complicated.

The sales prediction apparatus of the present invention is characterized in that the prediction time is every predetermined unit time period, and the sales prediction is calculated for a plurality of unit time periods for each of the predetermined unit time periods. Therefore, time series sales prediction can be performed.

The sales prediction apparatus of the present invention is characterized in that the sales prediction calculates at least a relative sales ratio in each unit period. Therefore, at least the trend of the sales trend can be predicted in time series.

The sales prediction apparatus of this invention is based on the relative sales ratio with respect to the elapsed unit time from which the sales performance was already obtained among the some unit time, and the relative sales ratio with respect to the unlapsed unit time to perform a sales forecast, It is characterized by performing a sales forecast about an elapsed unit time.

Therefore, based on the relative sales ratio calculated in time series, the future sales can be predicted in consideration of the performance in the progress unit period.

The sales prediction apparatus of the present invention is characterized by using the immediately preceding elapsed unit period as the elapsed unit period. Therefore, the sales prediction which appropriately reflected the tendency of the sales performance can be performed.

The sales prediction device of the present invention is characterized by using a plurality of unit periods as an elapsed unit period. Therefore, even if there are fluctuations in the sales performance due to unexpected special factors, stable sales prediction can be performed without being affected by this.

In addition, by using a plurality of unit periods immediately before, stable sales prediction without special factors can be performed while reflecting the tendency of the sales performance as much as possible.

The sales prediction apparatus of the present invention estimates the presence or absence of a stock shortage based on the arrival and the sales on the sales performance at an elapsed unit time, and uses the corrected sales results by correcting the sales results when it is determined that there is a stock shortage. It is characterized by performing a sales forecast.

Therefore, accurate sales prediction can be performed in consideration of the fluctuation in the sales performance caused by the lack of inventory.

The sales prediction device of the present invention is characterized by performing a sales forecast using sales results at a plurality of corresponding times. Therefore, the prediction can be made based on more accurate sales results.

The sales predicting apparatus of the present invention is characterized in that the sales forecasting is performed in consideration of the sales results before or after the corresponding time or both. Therefore, the sales prediction can be made in consideration of the sales results of other groups whose sales time is before or after the sales time of the group to which the subject of the transaction for which the prediction is performed belongs, or both. In other words, stable sales prediction can be performed without being affected by the gap in time series.

The sales predicting apparatus of the present invention is based on the overall sales schedule of the entire plural time periods for the trading target and the change in the sales performance at the corresponding time of the trading target group to which the trading target belongs to the corresponding trading at each unit time. It is characterized by calculating the sales forecast of the target.

Therefore, time-series sales prediction in each unit period can be performed based on the overall sales schedule.

The sales prediction device of the present invention is characterized by calculating a required delivery amount of the transaction target based on the calculated sales prediction, and outputting delivery order data.

Therefore, based on the predicted sales amount, the required delivery amount can be calculated.

The sales prediction apparatus of the present invention is characterized in that the predetermined unit time is set shorter than the difference period required from ordering to delivery of the transaction target. Therefore, it is possible to determine the required payment amount while considering the arrival schedule and sales performance of the ordered transaction target.

In the sales prediction apparatus of the present invention, the calculation of the required delivery amount calculates the required delivery amount at a future unit time considering the difference period, and the required delivery amount at a future unit time is performed based on the sales forecast. It is characterized by.

Therefore, the required delivery amount can be determined with good accuracy while considering the difference period.

The sales forecasting apparatus of the present invention, in the required delivery amount in the future unit time, estimates the predicted inventory amount at the corresponding unit time so as to be equal to the total of the forecast sales for the predetermined unit time before the unit time or to have a predetermined relationship. And determining the delivery amount based on the predicted inventory amount.

Therefore, based on the sales forecast, the appropriate stock amount can be determined and the required delivery amount can be calculated. In addition, according to the frequency of the prediction or the precision of the prediction, the appropriate delivery amount can be determined by adjusting whether the inventory has the same amount of inventory (or a predetermined relationship) as the forecast sales for the previous unit time.

The sales prediction device of the present invention is characterized in that the sales target fluctuates depending on seasonal factors. Therefore, the sales forecasting with high precision can be performed by grouping by the sales time.

The sales prediction apparatus of the present invention is characterized in that the object of trade is classified according to its attributes, and is grouped by a combination of the attribute and the selling time.

Therefore, more appropriate grouping can be performed, and high-precision sales prediction can be performed.

The sales predicting apparatus of the present invention is characterized in that a selling time used for grouping is a selling start time. By grouping according to the sales time, it becomes possible to perform sales forecasting at the beginning of sales.

According to the sales prediction method of the present invention, the sales results of each trading target are acquired, and based on the sales timing of the trading target to make a prediction, it is determined which trading target to which the trading target to perform the prediction corresponds to and the prediction is made. It is characterized by calculating the sales forecast of the transaction target based on the sales performance of the transaction target at a past time having similarity with the timing of the prediction for the transaction target to be performed.

That is, each transaction object is classified according to the sales time, and it is judged which existing transaction object the prediction target transaction corresponds to, and the sales prediction is performed based on the sales performance of the transaction object. Since the trade object is classified according to the sales time, the classification of the trade object is clear and the prediction with high precision can be performed. In addition, even when the number of objects to be handled increases, there is no fear that the classification increases dramatically and the processing becomes complicated.

In addition, in this specification, invention of water can be grasped | ascertained by invention of a method, and invention of a method can be grasped as invention of water. The invention also holds as a recording medium or a program product in which a program for realizing a predetermined function in a computer is recorded.

"Trade object" means the goods and services which predict a sale in this invention. In the embodiment, the apparel product corresponds to this.

"Sales forecast" refers to predicting unknown sales, and attaches whether it is past or future. It is a concept that includes not only the prediction of sales of goods but also the demand forecasting such as electricity consumption. The concept includes not only the case of predicting the absolute amount and the absolute number of sales, but also the case of predicting the tendency (relative ratio) of the sales, such as the index described in the examples.

"Sales results" means the quantity actually sold.

"Grouping based on sales time" refers to grouping of trading objects based on the sales time of the trading object, and according to not only the sales start time such as the start date of sale, the week and year, the day of the sale start day, and the end time of the sale. The concept involves grouping.

The "recording medium on which a program is recorded" refers to a recording medium such as a ROM, a RAM, a flexible disk, a CD-ROM, a memory card, or a hard disk on which a program is recorded. The concept also includes communication media such as a telephone line and a carrier path. The concept includes not only a recording medium such as a hard disk which is directly connected to a CPU and directly executes a recorded program, but also a recording medium such as a CD-ROM that records a program to be executed after being installed once in a hard disk or the like.

A "program" is a concept that includes not only a program that can be executed directly by the CPU, but also a program in a source format, a program that has been compressed, an encrypted program, and the like.

<Example>

1. Overall view of the system

Fig. 1 shows a system configuration when the sales prediction system according to the embodiment of the present invention is applied to the management of apparel goods. In this embodiment, the sales prediction computer 10 constitutes a sales prediction device. The host computer 6 receives the receipt data and the shipment data from the computer 2 of the distribution center which concentrates on the incoming and outgoing and storage of goods. In addition, the store register device 4 is arranged in each store of the whole country, and sends sales data to the host computer 6. In addition, the host computer 6 places an order for the maker computer 3.

Although data is transferred to the host computer 6 in this embodiment, the data may be sent via a recording medium such as a flexible disk. The data printed out on paper or the like may be input to the host computer 6. The data output from the host computer 6 may also be printed out on a recording medium or paper.

The host computer 6 receives these data, aggregates the data at a predetermined time every day, and communicates with the database server to update the database 8. The sales prediction computer 10 is connected to the database server 9 by a LAN and performs the sales prediction process based on the data accumulated in the database server 8.

2. Overview of the system

FIG. 2 schematically shows the flow of goods and the flow of information when the system of FIG. 1 is applied for sales prediction order management. ① The headquarters 18 recorded sales results prior to the previous year. The headquarters 18 performs the sales prediction based on the sales schedule (or target) determined for each product and the past sales trend of the group to which the product belongs. (2) The headquarters 18 orders the manufacturer 19 for the first time based on the calculated sales forecast. (3) The manufacturer 19 produces a product based on this ordering instruction and delivers it to the distribution center 12.

(4) The distribution center 12 that has received the goods is shipped to each of the stores 14 and 16 according to the instructions of the headquarters 18. From the distribution center 12, the data of arrival and shipment are transmitted to the headquarters 18. ⑤ Each store 14 or 16 sells the acquired product. Sales data is transmitted from the stores 14 and 16 to the headquarters 18.

(6) The headquarters 18 groups the products on the basis of the sale start time, and records data on the sales trend for each group. The headquarters 18 predicts the sales of future products based on the sales data (elapsed sales results) of each product and the past sales trend of the group to which each product belongs. (7) The headquarters 18 places additional orders for the maker 19 based on the calculated sales forecast. Thereafter, ③ to ⑦ are repeatedly executed.

As described above, by performing a high-precision sales forecast and placing an order based on this, a system with less occurrence of excess inventory and loss of sales opportunity can be constructed.

3. Host computer

3 shows a hardware configuration of the host computer 6. The host computer 6 includes a CPU 20, a hard disk 22, a display 24, a communication unit 26, a memory 28, a keyboard / mouse 30, a CD-ROM drive 32, a printer ( 34) and a flexible disk drive 36.

The communication unit 26 is for communicating with the distribution center computer 2, the store register device 4, the maker computer 3, and the sales prediction computer 10. In the hard disk 22, an operating system and a program for data aggregation processing are recorded. In addition, a database 8 is constructed on the hard disk 22. In the database 8, aggregated sales inventory data, commodity master data indicating the price of the commodity, and the like are recorded.

The program for data aggregation processing is installed in the hard disk 22 from the CD-ROM 38 or the like through the CD-ROM drive 32. 4 is a flowchart of a program for data aggregation processing.

In step S1, the CPU 20 stores the sales data (how many products were sold) from the store register device 4 of each store, and the incoming and outgoing data from the distribution center computer 2 (how many products are received). The number of items shipped and the ordering data (how many items were ordered) from the sales prediction computer 10 are received and accumulated in the database 8 of the database server 9. This process is repeatedly performed until a predetermined time is reached (step S2).

When the predetermined time comes, the received data is aggregated (step S3). Usually, this aggregation process is often performed at night when no sales data or incoming and outgoing data is generated. In the data aggregation process, the sales data from each store is aggregated for each product, and recorded on the hard disk 22 as the number of sales on the day. Similarly, the goods receipt data and the shipment data are collected for each product and recorded in the database 8 of the database server 9. As a result, daily sales, goods receipt, and shipping data are accumulated.

In addition, for each product, the total of sales of the week (the total of sales from Monday), the total of the number of purchases so far, the total of the number of sales, the number of stocks, etc. are also calculated. These data are recorded in the database 8 of the database server 9 (step S4).

5 shows an example of recorded data. As shown in Fig. 5, each product is specified according to a combination of a brand and an article number. In this embodiment, a case where a plurality of products are included in one brand and a plurality of brands are present will be described.

5 shows an example of the sales inventory data accumulated in this manner. The number of sales for each week is the number of sales for each week from Monday to Sunday. In FIG. 5, as for the product of article number "62421", the first stock price is "0",. It is shown that the 30th week is 31, the 40th week is 40, and the 32nd week is the number of sales of 48. The cumulative sales of the week (here 32) shall be confirmed on Sunday night.

The cumulative number of purchases is the cumulative number of receipts for the product. The cumulative number of sales is the cumulative number of sales so far. The number of inventories is the inventory calculated by the cumulative number of purchases-number of purchases-adjustments. This stock may be updated daily. The number of adjustments refers to the number of decreases in inventory due to factors other than sales, and is, for example, the number of damaged products, lost products, and the like. Data of these products is also transmitted from each store.

The starting sale was the first record of a non-zero share (ie, the first sale).

As described above, not only the sales inventory data for this year but also the sales performance data before the previous year are recorded in the database 8. 6 shows an example of sales performance data of last year. In addition, although not shown, the sales performance data of the year before last year are similarly recorded.

4. Hardware configuration of the sales forecasting computer

7 shows the hardware configuration of the sales prediction computer. The basic configuration is the same as for the host computer. However, a program for predicting sales is recorded on the hard disk 52. In this embodiment, at the beginning of the year (or at the end of the previous year), a main index (sales index) for predicting sales is calculated in advance and recorded on the hard disk 52.

5. Output treatment of stock index

8 is a flowchart of a main index calculation program executed at the beginning of the year. First, the CPU 50 accesses the database 8 and extracts data of products having a large number of sales from the sales performance data of last year (step S11). In this embodiment, when merchandise is arranged in ascending order, the merchandise having a cumulative total sales ratio of 70% or more is selected. Thus, by extracting and using the sales performance data of the product with many sales numbers, it is possible to grasp the exact sales trend. 9A shows an example of sales performance data of the extracted last year.

Next, the products of this sales performance data are grouped by item and start of sale stock (step S12). Here, an item means product groups, such as a skirt, a jacket, and pants. For example, in the data of FIG. 9A, data of only the skirt of item number &quot; 01 &quot; is shown, but similar data exists for other items such as a jacket and pants. In addition, the start-of-sale week refers to the week in which the commodity began to be sold, that is, the first week in which one or more sales were made.

In the data of FIG. 9A, it is common that the sale starting stock of the goods of the first row and the second row is "01" (that is, the first week). Therefore, the goods of the 1st line and the 2nd line are integrated as one group, and the sales results are totaled. The data obtained by such grouping is shown in FIG. 9B. In this way, the results of last year grouped by item and by sale starting stock are obtained.

Next, the CPU 50 calculates the average Savg (Item, Ws, Wf) with respect to the sales performance data of each week, together with the group sales data including one week before and after the sale start week (step S13). That is, the average of sales performance is computed by following formula (1).

Here, Item represents the item number, and Ws represents the sale starting stock. That is, the group is specified according to Item and Ws. Wf is the number of weeks when the sale start week is set as "the first week." Savg (Item, Ws, Wf) is an average of the sales performance in the week Wf in the group specified by Item and Ws. S (X, Y, Z) is the sales result in week Z of the group whose sale starting stock of item X is Y.

The average sales performance which included before and after the week in FIG. 10 is shown. In FIG. 10, only the data up to the seventh week is shown. However, when the sales results exist even after the eighth week, these data are also recorded. In this way, by averaging the sales results before and after the week, even if there are fluctuations in the sales results due to sudden factors, the influence can be reduced by averaging them.

In the present embodiment, the sales prediction of the product is performed based on the sales start stock as described below. Therefore, it is preferable that all the shares be the starting shares of any one product group. However, for an item for which a sufficient number of sales performance data are not obtained, data for a specific week may not be obtained for the week where the sale starts. For example, as shown in Figs. 9A and 9B, no commodity having the fourth week as the starting stock exists. However, the data can be obtained as shown in Fig. 10A by calculating the average sales results including the group including the starting and ending weeks. The total of Savg (Item, Ws, Wf) of all the states is recorded as T (Item, Ws).

As described above, after calculating the average sales performance as shown in FIG. 10A, the main index (sales index) is calculated for each group (step S14). The calculation of the main index Index (Item, Ws, Wf) is performed by the following equation.

An example of the index Index (Item, Ws, Wf) calculated by the above formula is shown in FIG. 10B.

In addition, the CPU 50 weights and averages the main index of two years ago and the main index of three years ago calculated in the same manner as described above with the main index of the previous year (step S15). In this embodiment, the average value is calculated by giving a weight of 5: 3: 2 to the main indexes of the previous year, two years ago, and three years ago, respectively. In FIG. 11, the main index which carried out the weighted average of the main index for three years is shown. The CPU 50 writes this weighted average main index into the hard disk 52.

By weighting the average of the three-year stock index, it is possible to create an index that reflects recent sales trends while eliminating fluctuations in sales due to special factors.

As described above, the weighted average main index for performing the sales prediction is recorded on the hard disk 52 of the sales prediction computer 10.

6. Sales forecast and order volume calculation processing

6.1 Condition Setting

12 shows a flowchart of the sales prediction order amount calculation processing program recorded on the hard disk 52 of the sales prediction computer 10.

First, the CPU 50 displays setting screens of various conditions on the display 54 to prompt input (step S21). 13 shows a screen for condition setting. In this screen, the lead time (period required from ordering to receipt) can be entered as the number of weeks for each product.

`` K '' is a product using knit material, `` C '' is a product using mainly cut and used material, `` J '' is a product using material used in trainers, etc., `` Goods '' is a product such as shoes, muffler, hat It is shown. In the example of FIG. 12, a lead time of 4 weeks for "K", a lead time of 3 weeks for "C", a lead time of 3 weeks for "J", and a lead time of 3 weeks for "Goods" are input. It is. These lead times are used when determining the order amount.

In addition, in the column of a defective article correction, it is possible to select whether it is "valid" or "invalid". Here, the lack of product correction means a situation in which the product cannot be sold due to the lack of goods at the head when calculating the sales results of pre-weeks for the sales forecast, and corrects the sales results. If it is set to "valid", the defective article correction is performed. If it is set to "invalid", the defective article correction is not performed.

In addition, the column of "expected stock weeks" is a numerical value which shows how many weeks worth of stocks are equivalent. In Fig. 12, the inventory amount is determined so as to have a stock corresponding to the sales for four weeks.

Each of the pieces of information described above is recorded on the hard disk 52 and used in the sales prediction order amount determination process.

6. 2 sales forecast

After the conditions are set as described above, the sales are predicted (step S22). 14 and 15 show flowcharts of the sales prediction process. In the present embodiment, the sales prediction (first sales prediction) in the state where there is no sales performance for the product and the sales prediction (additional sales prediction) after obtaining the sales performance for the product are processed separately. The flowchart of a first time sales prediction is shown in FIG. 14, and the flowchart of a further sales forecast is shown in FIG.

In the first sales prediction, the CPU 50 first acquires a predetermined sales schedule for this year from the database 8 for the product for which the sales prediction is made (step S31). This sales schedule is determined based on the sales target of the brand, the characteristics of the product, and the like, and is recorded in the database 8.

Next, the sale start date of the said product is acquired from the database 8. This scheduled sale start date is also determined in advance based on the characteristics of the product and recorded. Based on the scheduled sale start date and the item classification of the product, it is determined which group the product belongs to, and the weighted average main index of the group is obtained from the hard disk 52 (step S32).

Subsequently, the CPU 50 predicts each week's sales SP (good, Wf) based on the sales schedule of the product and the weighted average main index (step S33).

Here, SP (good, Wf) is the sales expected quantity of the Wf week from the scheduled start sale of the goods good. ST is the full sale of the product for this period, and Index (Item, Ws, Wf) is the weighted average share index. Item Item is the item classification to which the corresponding product good belongs, and is going to start selling Ws is the starting to be sold of the corresponding product good. That is, the total sales schedule is distributed to each state according to the weighted average share index, and the sales forecast for each state is performed.

The CPU 50 records this sales prediction in the hard disk 52 and outputs it in a format such as print out or data (step S34).

These processes are performed for all the products targeted for sales forecasting. Usually, this process is performed from time to time after determining the scheduled number of sales of each product. As a result, sales prediction data for each product as shown in FIG. 15 is recorded on the hard disk 52. In FIG. 15, the SKU is a code for specifying an individual product.

The sales forecast of each product calculated in this way is recalculated based on the sales performance when there is a sales performance on the products. The processing is shown in the flowchart of FIG. The sales prediction process in Fig. 16 is usually executed on the first day of each week (Monday in this embodiment).

The CPU 50 first acquires performance data of each product from the database 8 (step S41). Moreover, a digestibility rate etc. are calculated based on the acquired data. 17A shows an example of acquired performance data. These data receive data from each store or distribution center, are aggregated by a server computer, and recorded in the database 8.

"Part number", "SZ", and "COL" have shown the product number, size, and color of a product, respectively, and can identify individual products by these. By combining these three, the above-mentioned SKU is obtained. In the "material name", the above-mentioned divisions of "K", "C", "J" and "goods" are recorded. The sale start week of the product is recorded in the sale start week. "Relative" is the standard selling price at the store.

In the "5 weeks ago" column, the number of sales for the week before 5 weeks is recorded. "4 weeks ago", "3 weeks ago", "last week", and "weekend" are the number of sales of 4 weeks ago, 3 weeks ago, the previous week, and the previous week, respectively. "Accumulation of purchase" is the accumulation of the number of purchases to the previous week for the said product. "Sum of sales" is the total of the number of sales until the previous week for the said product. Weekend inventory is inventory at the end of the weekend. "This week's arrival" has already been ordered and is the quantity scheduled for arrival this week. Similarly, "introduction in stock", "receipt after three weeks" and "receipt after four weeks" are the stocks scheduled to be received four weeks later, after the inner week, three weeks, respectively.

The CPU 50 calculates the ratio (digestion rate) of the number of purchases based on the accumulated accumulation amount and the accumulated accumulation amount, and records it in the column of "cumulative digestion rate". In addition, based on the number of purchases of the last week and the number of purchases of the last week, how many ratios were sold (weekly digestibility) among the number of purchases was calculated and recorded in the column of "Weekly digestibility". In the present embodiment, the weekly stocks are added to the weekend stock, and the weekly stocks are subtracted here to obtain the weekly digestibility.

Next, the CPU 50 determines whether the defective article correction set in FIG. 13 is valid (step S42). If it is not valid, defect correction is not performed, and the flow proceeds to step S45. If the defective article correction is effectively set, step S43 or below is executed to execute the defective article correction.

At any one store, the product may be unexpectedly sold and a situation (shortage) that could not be sold due to lack of inventory may occur. Since the sales performance in the event of such a situation does not reflect the original performance state, if the sales prediction is made using this, there is a possibility of making an incorrect prediction. Therefore, in such a case, the sales performance used for sales prediction is corrected. This is referred to as defect correction in this embodiment.

The CPU 50 first estimates whether there is such a shortage in step S43. In this embodiment, the presence or absence of a defective article is estimated based on the cumulative digestibility and the weekly digestibility. That is, when the cumulative digestibility exceeds 60% and the weekly digestibility exceeds 60%, it is determined that a defect is generated at any store.

If it is determined that no shortage occurs, the flow advances to step S45. If it is determined that a shortage is occurring, the sales performance of the previous week is corrected according to the following equation (step S44).

Thereby, the selling of the electric pole of the 1st line goods of FIG. 17A is corrected to 386.4 (refer to the column of defective article correction sales). In addition, the selling of the electric pole of the goods of the 2nd line and the 3rd line which did not satisfy the said conditions is not corrected, and becomes 277 and 241 streets.

Next, the CPU 50 calculates the sales prediction value SG (good, n) after n weeks in accordance with the following equation for each product.

Here, SR (good) is the sales performance of the previous week (after correction if there is a defective article correction). Index (Item, Ws, Wc-1) is the main index of the previous item Wc-1 of the commodity item and the starting group group Ws. Wc means this week and Wc + n means week after n weeks. In this embodiment, this week is 39th week. In addition, n is changed from 0 to 6, so that sales forecasts for each week from this week to six weeks later are made. This is done for each product.

In addition, an example of the main index (3-year weighted average) used for the prediction is shown in Fig. 17B. The calculated sales forecast for each week is shown in FIG. 18. Sales forecast from "39 weeks" (this week) to "45 weeks" is calculated.

6. 3 need order quantity output

As described above, after initially or additionally calculating the sales forecast for each week, the CPU 50 calculates the required ordering amount (step S23 in FIG. 12).

19, the concept for calculating the required ordering amount in a present Example is shown typically. Items ordered this week (39 weeks) will be available 42 weeks after the lead time LT (controllable leading stock). This lead time LT is set in FIG.

In the present embodiment, when the week (42 weeks) after the lead time LT of the corresponding product starts from 39 (week 39), the order amount of the week is determined so as to have a stock consistent with the forecast sales for the next 4 weeks. . The number of weeks that you have in stock that matches your forecasted sales for the next several weeks is called the expected number of stocks. That is, in this example, the predicted stock number of weeks was set to "4". In other words, on the basis of the sales forecast, stocks of an amount capable of covering sales up to four weeks later are secured.

20 is a flowchart of this required order quantity calculation program. The CPU 50 sums the sales prediction value for the product from this week (39 weeks) to the controllable leading week (42 weeks), and makes it "purchase prediction 1" (step S51). For example, as shown in Fig. 18, the sales forecast of 39 to 41 shares is added to the first-line product, and "773" is obtained and recorded.

Next, the arrival planned number of the said period is totaled (step S52). That is, with respect to the said product, although already ordered, the quantity not yet received is totaled, and it is set as "preparation 1". In addition, this can be obtained based on the order record for each product and the record of the receipt recorded in the database 8. In FIG. 18, "preparation 1" is calculated and recorded as "1100".

Next, the number of inventories at the time when the controllable leading stock (42 weeks) starts is calculated as "stock 1" (step S53). This "stock 1" can be calculated by adding "preparation 1" to "last weekend stock" and reducing "sales forecast 1".

Next, the sales forecast of the estimated number of stock weeks from the controllable leading stock (42 weeks) is totaled (step S54). That is, the sales forecast for 42 to 45 weeks is summed up. Subsequently, "Preparation 1" is subtracted from "the total preparation stock", and "Preparation 2" is calculated (step S55). Here, the "preparation stock total" means the quantity currently ready (the state which was ordered but is not received) with respect to the said product. That is, "preparation 2" shows the start of the arrival schedule after the controllable leading stock.

If the "stock 1" is larger than 0, the CPU 50 calculates the "ordering frame" by the following formula (steps S56 and S57).

In addition, when "stock 1" is 0 or less, a "ordering frame" is calculated by the following formula (steps S56 and S58).

If the "ordering frame" obtained as a result of the said calculation is larger than 1, let this order number be this "ordering frame" (step S59, S60). If the "ordering frame" is equal to or less than 0, no ordering is performed this time (steps S59 and S61).

18 shows an example of the calculated "ordering frame" and "ordering number". The first row and second row products do not have this order. The third line shows that 270 orders must be placed.

The calculated order amount and the like are output to the database, recorded, and printed out. In addition, it is transmitted to the maker 3 via the host computer 6. The maker 3 receives this and produces it (refer FIG. 1).

In addition, although "ordering frame" = "ordering number" is mentioned above, you may make a human order determine the ordering number with an "ordering frame" as an upper limit.

In addition, although the ordering instruction is transmitted to the maker 3 online, the ordering instruction may be sent to a recording medium or a printout sheet.

7. Other Examples

(1) In the above embodiment, the apparel product has been described as an example, but the present invention can also be applied to products that are effective in performing management from production to sale, for example, industrial products and processed food products. In addition, it is also applicable to the order management for the goods whose management of purchase is important, for example, the goods placed in a convenience store.

(2) In the above embodiment, the order amount is calculated based on the sales forecast, but the production amount, the purchased amount, the acquired amount and the like may be determined based on the sales forecast.

(3) Although the sales prediction number is calculated in the above embodiment, the sales amount, the sales amount and the like may be calculated.

(4) In the above embodiment, the first order of the product is calculated based on the sales forecast, but the first order is determined by another method (for example, breakdown from the overall plan), and the additional order amount May be determined based only on the sales forecast.

According to the present invention as described above, appropriate sales prediction can be performed regardless of the sales result and avoiding the complexity of the processing.

Claims (36)

In the sales prediction apparatus which predicts the future sales based on the past sales results, Acquire sales results for each transaction, On the basis of the sales results, grouping each trading subject with at least one index of selling time of each trading subject, Based on the sales time of the transaction target to make a prediction, it is determined to which group the transaction target to make a prediction belongs to, A sales prediction device, characterized in that the sales target of the trading target is calculated based on the sales performance of the trading target group in the past corresponding time having similarity with the timing of the prediction with respect to the trading target for prediction. In the sales prediction apparatus which predicts the future sales based on the past sales results, It is configured to have access to a record record of sales performance of each group for the grouped transactions with at least one indicator of the sales timing of each transaction. Based on the sales time of the transaction target to make a prediction, it is determined to which group the transaction target to make a prediction belongs to, With respect to the transaction target to be predicted, the sales results of the transaction target group in the past corresponding time having similarity with the timing of the prediction are acquired from the recording unit, and based on the sales performance, the sales forecast of the transaction target is calculated. A sales prediction device, characterized by calculating. The method of claim 1, The time for performing the prediction is every predetermined unit time, and the sales prediction device is calculated for a plurality of unit time for each of the predetermined unit time. The method of claim 2, The time for performing the prediction is every predetermined unit time, and the sales prediction device is calculated for a plurality of unit time for each of the predetermined unit time. The method of claim 3, The sales prediction apparatus calculates a relative sales ratio in at least each unit period. The method of claim 4, wherein The sales prediction apparatus calculates a relative sales ratio in at least each unit period. The method of claim 5, Estimation of sales for an undeveloped unit time, based on the relative sales ratio and the past sales performance for an elapsed unit time for which the sales performance has already been obtained among the plurality of unit time periods, and the relative sales ratio for an undeveloped unit time for which sales are expected. A sales prediction device, characterized in that for performing. The method of claim 6, Estimation of sales for an undeveloped unit time, based on the relative sales ratio and the past sales performance for an elapsed unit time for which the sales performance has already been obtained among the plurality of unit time periods, and the relative sales ratio for an undeveloped unit time for which sales are expected. A sales prediction device, characterized in that for performing. The method of claim 7, wherein A sales forecasting device, characterized in that the previous elapsed unit period is used as the elapsed unit period. The method of claim 8, A sales forecasting device, characterized in that the previous elapsed unit period is used as the elapsed unit period. The method of claim 7, wherein And a plurality of unit periods are used as the elapsed unit periods. The method of claim 8, And a plurality of unit periods are used as the elapsed unit periods. The method according to any one of claims 7 to 12, With respect to the sales results of the transitional unit period, the sales forecast is estimated based on the arrival and the sales, and when the inventors judge that there is a lack of inventory, the sales sales are corrected by using the revised sales results after correcting the sales results. Sales prediction device to assume. The method according to any one of claims 3 to 12, And the predetermined unit time is a time, a time zone, a day, a week, a month, or a year. The method of claim 1, And the corresponding time period is a time period before the previous year, a month before, or a day before the day corresponding to the time when the prediction is made. The method of claim 2, And the corresponding time period is a time period before the previous year, a month before, or a day before the day corresponding to the time when the prediction is made. The method of claim 1, The similarity of the corresponding time includes the commonality of day, month, week, day, time or climate. The method of claim 2, The similarity of the corresponding time includes the commonality of day, month, week, day, time or climate. The method according to any one of claims 15 to 18, A sales prediction device, wherein sales sales are predicted using sales results in a plurality of corresponding times. The method according to any one of claims 1 to 12, A sales forecasting device, characterized in that the sales forecasting is performed in consideration of the sales results of other groups whose sales time is before or after the sales time of the group to which the prediction target transaction belongs, or both. The method according to any one of claims 1 to 12, Computing the sales forecast of the corresponding trading target at each unit time based on the overall sales schedule of the entire time period for the trading target and the change in the sales performance at the corresponding timing of the trading target group to which the trading target belongs. A sales prediction device, characterized in that. The method of claim 1, Based on the calculated sales forecast, the required delivery amount of the transaction target is calculated, and delivery order data is output. The method of claim 2, Based on the calculated sales forecast, the required delivery amount of the transaction target is calculated, and delivery order data is output. The method of claim 22, And said predetermined unit time period is set shorter than a difference period required from ordering to delivery of the transaction target. The method of claim 23, wherein And said predetermined unit time period is set shorter than a difference period required from ordering to delivery of the transaction target. The method of claim 24, The calculation of the required delivery amount calculates the required delivery amount in a future unit time in consideration of the difference period, and the required delivery amount in a future unit time is performed based on the sales forecast. The method of claim 25, The calculation of the required delivery amount calculates the required delivery amount in a future unit time in consideration of the difference period, and the required delivery amount in a future unit time is performed based on the sales forecast. The method of claim 26 or 27, In the required delivery amount in the future unit time, the predicted inventory amount in the unit time is determined to be equal to or equal to the total of the forecast sales of the predetermined number of unit time prior to the unit time, or in a predetermined relationship, and the prediction is made. The sales prediction device, characterized in that for determining the delivery amount based on the inventory amount. The method according to any one of claims 1 to 12, The sales target device, characterized in that the sales fluctuations due to seasonal factors. The method according to any one of claims 1 to 12, The transaction targets are classified according to their attributes, and are grouped by a combination of the attributes and the selling time. The method according to any one of claims 1 to 12, The sales timing used for the grouping is a sales start timing. In the sales prediction program for realizing a sales prediction apparatus using a computer, Acquire sales results for each transaction, On the basis of the sales results, grouping each trading subject with at least one index of selling time of each trading subject, Based on the sales time of the transaction target to make a prediction, it is determined to which group the transaction target to make a prediction belongs to, A program for causing a computer to execute a process for calculating a sales forecast of the trading target based on the sales performance of the trading target group at a past correspondence time having similarity with the timing of the prediction for the trading target to make a prediction. . The recording medium which recorded the program of Claim 32. In the sales prediction program for realizing a sales prediction apparatus using a computer, Based on the sales time of the transaction target to make a prediction, it is determined to which group the transaction target to make a prediction belongs to, Accessing a recorder which records the sales performance of each group, for the grouped transactions with at least one indicator of the sales timing of each transaction; With respect to the transaction target to be predicted, the sales performance of the corresponding trading target group in the past corresponding time having similarity with the timing of the prediction is acquired from the recording unit, and based on the sales performance, the sales prediction of the trading target is performed. A program for causing a computer to execute a calculating process. The recording medium which recorded the program of Claim 34. In the sales prediction method which predicts future sales based on the past sales results using a computer, Acquire sales results for each transaction, On the basis of the sales time of the trading target to make a prediction, it is determined which existing trading target corresponds to the trading target to make the prediction, A sales forecasting method for a trading target to be predicted, wherein the sales forecast of the trading target is calculated based on the sales performance of the trading target at a past time having similarity with the timing of the prediction.