Price revision approval support system

Abstract

This system provides a price revision approval support system that streamlines the purchasing approval process, makes approval decisions more objective, and optimizes purchasing costs by integrating external market data and purchase history data to statistically evaluate the validity of proposed price revisions and automatically approve or counter-propose them based on the results. [Solution] The price revision approval support system 10 includes: an item extraction unit 12 that extracts items subject to price revision based on past purchase history; a price acquisition unit 14 that acquires price information from multiple external data sources for items subject to price revision; a statistical calculation unit 16 that calculates at least one of the deviation rate, standard deviation, and market index fluctuation rate based on the price information and reference price data; a price evaluation unit 18 that determines the appropriateness of the price revision according to predetermined judgment rules and generates an approval or disapproval result; and an automatic approval unit 20 that automatically registers approval information to the purchasing system based on the approval or disapproval result.

Description

Technical Field

[0001] The present invention relates to a system for assisting in approval judgment of price revisions in the purchasing operations of enterprises and organizations. In particular, by automatically evaluating the validity of price revisions using external market data and making approvals or counter-proposals based on the results, it relates to a price revision approval support system that aims to improve the efficiency of purchasing operations and optimize costs.

Background Art

[0002] Conventionally, in the purchasing department of an enterprise, for a price revision proposal presented by a supplier, the person in charge judged its validity based on past purchasing history and internal standards and carried out approvals or price reduction negotiations as necessary. Such an approval process often depends on the experience and subjectivity of the person in charge, has the problem that the judgment criteria are opaque, and also takes time for approval.

[0003] As a technology related to the present invention, for example, Patent Document 1 discloses a purchasing system constructed among a purchasing department, a purchasing request department, and a purchasing execution department. The purchasing request department 62 has a purchasing request means for inputting purchasing request information including personal information of the purchasing requester and a budget checking means for the purchased item. The purchasing execution department 64 has a purchasing request reception means and a purchasing order means for ordering the purchased item based on the purchasing order information including personal information of the purchasing execution person in charge. The purchasing department has a pre-registered means for the purchased item for registering pre-selected purchased item information in the purchased item list, a pre-registered means for the price of the purchased item for registering the price information of the purchased item negotiated with the supplier in the price list, a purchasing request information storage means for storing the purchasing request information, and a purchasing performance database update means 7 for sequentially updating and accumulating purchasing performance data by the purchasing order means.

Prior Art Documents

Patent Documents

[0004]

Patent Document 1

Summary of the Invention

[0005] Traditional purchasing systems allowed for price comparisons based on a single catalog or internal database, but they could not evaluate proposed price revisions in conjunction with external market data such as the Corporate Goods Price Index (CGPI) or online price comparison sites. As a result, disadvantages sometimes arose, such as price revisions being approved despite falling market prices, or opportunities for price reduction negotiations being lost.

[0006] This invention has been made in view of the above problems, and aims to streamline the purchasing approval process, make approval decisions more objective, and optimize purchasing costs by statistically evaluating the validity of price revision proposals by integrally utilizing external market data and purchase history data, and by automatically approving or counter-proposing based on the results. [Means for solving the problem]

[0007] The price revision approval support system according to the present invention is a price revision approval support system that automatically evaluates the appropriateness of price revisions using external market data, comprising: an item extraction unit that extracts items subject to price revision based on past purchase history, and for the items subject to price revision, from a plurality of external data sources market A price acquisition unit that acquires price information, (a) the rate of deviation based on the difference between the proposed revised price stated in the price revision proposal and the market price based on the market price information, (b) the standard deviation of the price distribution based on the market price information, and (c) the market index fluctuation rate based on a predetermined index included in the external market data, respectively. A statistical calculation unit calculates the results, and using the calculated deviation rate, standard deviation, and market index fluctuation rate By determining whether the proposed price revision falls within a statistically acceptable range, A price evaluation unit statistically assesses the validity of the proposed price revision and determines whether it is approved, conditionally approved, or rejected; and an automatic approval unit automatically registers the approval result to the purchasing system if the determination result is approved. Only if the aforementioned determination result is not approved, the items subject to price revision. The price evaluation unit comprises a notification unit that automatically generates alternative product suggestions or price reduction suggestions and notifies the purchasing manager's terminal, and the price evaluation unit determines the deviation rate to be predetermined deviation rate Approval will be denied if the threshold is exceeded, and the market index fluctuation rate When a negative value is shown, the above corresponds to the disapproval. Generate alternative product suggestions or price reduction suggestions. death,targets for comparison External data source If the number of companies is three or more, outliers are excluded based on statistical significance testing. one Company only statistically significant outliers This indicates In some cases teeth, The one Only the company will be subject to counter-proposals or rejections, based on the aforementioned standard deviation. The range is the average value ± a predetermined coefficient multiple. Statistical tolerance as use By doing so, Optical characteristics Depending on the price fluctuation characteristics of each It is characterized by making approval decisions.

[0008] Furthermore, in the price revision approval support system according to the present invention, it is preferable that the external data source includes at least one of a corporate goods price index database, an online price comparison site, an electronic catalog server, and a company-specific catalog.

[0010] Furthermore, in the price revision approval support system according to the present invention, it is preferable that the price acquisition unit acquires the price information from the external data source via a cXML, OCI, or CSV interface.

[0011] Furthermore, in the price revision approval support system according to the present invention, it is preferable that the statistical calculation unit stores price history for a predetermined period in the past as time-series data and predicts future prices using a learning model.

[0012] Furthermore, in the price revision approval support system according to the present invention, the price evaluation unit performs scoring on a plurality of judgment items, Approval is granted if the total score obtained by summing the scores is equal to or greater than the upper threshold; rejection or counter-proposal is made if the total score is less than the lower threshold; and conditional approval is granted if the total score is less than the upper threshold but equal to or greater than the lower threshold. It is preferable.

[0013] Furthermore, in the price revision approval support system according to the present invention, if a comparison based on multiple external data sources shows a statistically significant deviation in only one company, it is preferable to target only that one company for a counter-proposal or rejection. [Effects of the Invention]

[0014] According to the present invention, by combining external market data and in-company purchase history data to statistically evaluate a price revision plan, an objective approval judgment that does not depend on the experience and subjectivity of the person in charge becomes possible. In addition, by automatically approving a price revision plan with high validity, labor saving and speedup of the approval work can be achieved.

[0015] Furthermore, when the market index is decreasing or low price information of substitute products is obtained, a counterproposal can be notified, so that the opportunity for price negotiation is not missed, contributing to the reduction of purchase costs. As a result, the efficiency of the entire purchasing business, the optimization of prices, and the cost reduction effect can be realized at a high level.

Brief Description of the Drawings

[0016] [Figure 1] It is a block diagram showing the overall configuration of a price revision approval support system according to an embodiment of the present invention. [Figure 2] It is a flowchart showing the processing procedure of a price evaluation unit in a price revision approval support system according to an embodiment of the present invention.

Modes for Carrying Out the Invention

[0017] Hereinafter, embodiments of the present invention will be described in detail with reference to the accompanying drawings. In the following, the same elements in all the drawings are denoted by the same reference numerals, and duplicate explanations are omitted. In addition, in the description in the text, the reference numerals described previously will be used as necessary.

[0018] As used herein, "external market data" refers to information on market prices and price indices obtained from external institutions or external services, which is different from the internal purchase history data of the enterprise or organization. Specific examples include Corporate Goods Price Index (CGPI) database, price information on online price comparison sites, product price information obtained from electronic catalog servers, listed price information in corporate-specific catalogs, foreign exchange rate information, raw material price statistics, and other public or private price information services. The corporate-specific catalog refers to a login-type dedicated price catalog set by each e-commerce operator for each contracting enterprise, and the price specific to the company is displayed when the user logs in. The format is similar to a general web catalog, but the displayed price is based on the dedicated conditions for each contracting enterprise and is different from the standard price that is publicly available.

[0019] FIG. 1 is a block diagram showing the overall configuration of a price revision approval support system 10 according to the present invention. The price revision approval support system 10 is constructed on a purchase management server and includes an item extraction unit 12, a price acquisition unit 14, a statistical calculation unit 16, a price evaluation unit 18, an automatic approval unit 20, and a storage unit 22. These are configured as program modules and can be accessed from the terminals of a plurality of users 4 via a network 2.

[0020] In the present invention, the "user 4" refers to a person or organization that operates or supervises the price revision approval process in the purchase system. Specifically, it includes a purchasing staff, a procurement responsible person, or a manager with approval authority, etc. The user 4 can confirm the content of the price revision application via the system screen and execute actions such as approval, conditional approval, counter-proposal, and rejection as necessary.

[0021] The item extraction unit 12 refers to the purchase history for the past 12 months from the storage unit 22, which is the database of the purchasing system, and identifies items subject to price revisions from suppliers. Extraction conditions include (a) the price has been maintained for a certain period in the past, (b) a price revision flag has been sent from the supplier, and (c) the contract expiration date is approaching. The extraction results are listed along with information such as the item code, supplier code, current price, and last revision date.

[0022] The price acquisition unit 14 acquires price information for identified items subject to price revision from multiple external data sources. These external data sources include at least one of the following: a Corporate Goods Price Index (CGPI) database, an online price comparison site, and an electronic catalog server.

[0023] The price acquisition unit 14 acquires price information for identified price revision target items from multiple external data sources using CXML, OCI, or CSV interfaces, either in real time or in periodic batch processing. The acquired data is normalized to unit prices, taking into account exchange rates and transportation costs.

[0024] In this embodiment, "automatic determination" refers to a process that classifies the validity of a proposed price revision as approved, conditionally approved, or rejected based on external market data and internal purchase history data, using predetermined determination rules or a learning model, without requiring human intervention. This enables fully automated processing from application to final approval, thereby accelerating and objectifying approval decisions.

[0025] Data acquisition is performed in real time or in scheduled batches using CXML, OCI, or CSV interfaces. Acquired data is normalized to unit prices, taking into account exchange rates and transportation costs.

[0026] The statistical calculation unit 16 calculates the following values ​​based on the acquired price information and reference price data. (a) Deviation rate = (Revised price - Market average price) / Market average price × 100[%] (b) Standard deviation = Square root of the variance of the historical price data (c) Market index volatility = (Current index - Base period index) / Base period index × 100[%]

[0027] Furthermore, in this embodiment, price history for the past 12 months or more is stored as time-series data in the storage unit 22, and this data is used to predict future market price trends. Time-series analysis models (ARIMA model, exponential smoothing, etc.) and machine learning models (LSTM: Long Short-Term Memory network, gradient boosting model, random forest regression, etc.) can be used for prediction. These models learn from explanatory variables such as past price fluctuation patterns, market indices, exchange rates, and raw material prices as input, and output predicted prices or price fluctuation rates for the next 1 to 6 months. Moreover, even when multiple external catalogs (company-specific catalogs) are provided in different formats and styles, the data structure can be standardized using the generation AI, and they can be automatically matched and integrated based on common keys such as product names and model numbers. This makes it easy to perform cross-searches and comparisons between multiple catalogs, which were previously impossible due to format differences, and can be used as foundational data for validity judgments in the price evaluation unit 18.

[0028] The prediction results are used in the price evaluation unit 18 to determine the discrepancy between future market price trends and the proposed revised price. For example, if future prices are predicted to be in a downward trend, the approval score can be reduced or priority can be given to creating a counter-proposal. This allows for approval decisions that consider medium-term price fluctuation risks, rather than simply comparing current prices.

[0029] The price evaluation unit 18 has the function of determining the appropriateness of the price revision based on predetermined judgment rules and deciding whether to approve or disapprove it as a result. This judgment is performed by scoring according to the judgment flow shown in Figure 2. An example of each judgment item and scoring conditions is as follows. (1) Attachment of valid evidence data (S2): +1 point for YES (2) No extreme discrepancies in external comparisons (S4): +1 point for YES (3) No change compared to past price trends (S6): +1 point for YES (4) No abnormalities found in comparison with past trends (S8): +1 point for YES (5) No replacement product with the same specifications available (S10): +1 point for YES (6) Price reasonableness in comparison with market index (S12): +1 point for YES (7) Consistency with external e-commerce prices (e.g., Amazon) (S14): +1 point for YES (8) The price difference from the previous price is within the specified range (S16): +1 point for YES

[0030] The judgment rule in this embodiment is characterized by rejecting the proposed price revision if the deviation rate calculated by the statistical calculation unit 16 exceeds a predetermined threshold, and further notifying purchasing managers and suppliers of alternative product information or price reduction proposals if the market index fluctuation rate is negative and falls below a predetermined threshold, i.e., the market index is declining. In addition, the validity check is based in principle on a comparison of catalog prices from three or more companies, and if the number of companies to compare is insufficient, it is supplemented with past price trends and price trends of major e-commerce sites. Furthermore, if only one company shows a large deviation, only that company is subject to rejection or counter-proposal, and the remaining data is moved to the next process. In addition, the acceptable range can be flexibly set according to the characteristics of the product based on a statistical standard using standard deviation (e.g., sigma 2).

[0031] Figure 2 is a flowchart showing the processing procedure of the price evaluation unit 18 in the price revision approval support system 10 according to the present invention. The details of each step will be described below.

[0032] (S2: Determination of the number of catalogs to compare) First, we check whether the target item is listed in multiple company-specific catalogs. Validity checks are generally conducted by comparing prices across three or more catalogs; if there are two or fewer, it is determined that the price discrepancy has not been adequately assessed. Furthermore, even if multiple catalogs exist, the target item may only be listed in one. For example, validity is confirmed by comparing the prices of specified bearings across each catalog.

[0033] (S4: Check catalog price fluctuations) Next, we compare the catalog prices of multiple companies (e.g., four companies) to check if there is a deviation of more than ±10% from the average or lowest price of other companies. The calculation of the deviation rate is limited to cases where comparisons of three or more companies are possible. If only one company's price is significantly different, only that company's data will be subject to a counter-proposal or rejection, and the data of the remaining three companies will proceed to the next process. For example, if the price of one company is statistically significantly different from that of other companies (e.g., exceeding 2 standard deviations), only that company will be subject to a counter-proposal or rejection, and processing will continue based on the catalog prices of the remaining three companies.

[0034] (S6: Comparison of past price trends when catalog comparison is not possible) If a catalog comparison is not possible with fewer than three companies, price trends are referenced based on the company's internal purchasing history for the past 12 months to check recent price revision history and price fluctuation patterns. If the proposed revised price is within the standard deviation range of past trends (e.g., sigma 2 or less), it is evaluated as a stable price, a score is added, and it is decided that it can be treated as a catalog item. On the other hand, if it significantly exceeds the standard deviation (greater than sigma 2), the item is treated as a custom item or a single-company exclusive item, considered an item requiring individual negotiation, and subject to return or counter-proposal to reduce the overall approval processing burden.

[0035] (S8: Price trend confirmation) When comparing catalogs from three or more companies, the system compares past price rankings with the trends in revised prices to check for any unusual ranking fluctuations. For example, if a company has consistently been the lowest-priced for the past four months but suddenly jumps to the highest price or a higher ranking, this will be considered unusual behavior and subject to a counter-proposal or rejection. On the other hand, if the rankings of multiple companies are constantly changing, indicating a normal competitive relationship without any discernible pattern, the system will be deemed normal and a score will be added. This allows for flexible approval decisions that take into account ranking fluctuation trends, not just simple numerical price comparisons.

[0036] (S10: Substitute existence determination) Next, we refer to electronic catalogs and supplier lists to confirm the existence of alternative products with the same specifications and equivalent quality. If an alternative product exists and is inexpensive, it is listed as a potential alternative and its value influences the approval decision. If no alternative product exists, a score is added.

[0037] (S12: Market index comparison) The Corporate Goods Price Index (CGPI) and related market indices will be obtained and their consistency with the proposed revised price will be verified. For example, if the CGPI has recently fallen by 2% but the proposed revised price has risen by 5%, the discrepancy will be highlighted and the approval criteria will be tightened.

[0038] (S14: Price trend matching) For items for which the Corporate Goods Price Index (CGPI) does not exist, we will check if price trends can be obtained from major e-commerce sites (e.g., Amazon, Kakaku.com, Rakuten Market, etc.). If available, we will use this price trend data to compare with the proposed revised price. If the difference is within ±2%, it will be considered appropriate and a score will be added; if it is 10% or more higher, it will be considered a candidate for rejection. On the other hand, if neither the CGPI nor e-commerce sites exist, the item will be classified as a custom item or similar subject to individual negotiation and processed separately from regular catalog items.

[0039] (S16: Market trend difference confirmation) The appropriateness of the proposed revised price will be confirmed based on a comparison of price trends with market data (such as CGPI and Amazon). The degree of price fluctuation varies depending on the item and time period; some items may hardly fluctuate, while others may fluctuate significantly. For example, even items that have traditionally been stable, such as rice, may experience sharp price fluctuations recently. In such cases, the criteria for judgment will be flexibly set according to the magnitude of the fluctuation. Specifically, the standard deviation of the market index over the past year will be calculated, and the allowable deviation will be set at ±3% if the fluctuation range is small (e.g., standard deviation is less than 1%), ±5% if the fluctuation range is moderate, and ±8-10% if the fluctuation range is large. This will enable appropriate approval decisions according to the characteristics of each item and avoid unnecessary rejections and counter-proposals.

[0040] The automated approval unit 20 has the function of automatically registering approval information to the purchasing system based on the judgment result (approval or rejection) by the price evaluation unit 18. If the total score is 6 points or higher, the automated approval unit 20 automatically registers the approval information to the purchasing system and sends an approval notification to the supplier. If the total score is 4 to 5 points, it is a conditional approval and automatically generates negotiation policies and conditional prices. If the total score is 3 points or lower, it issues a counter-proposal notification or a return process and presents alternative product information and recommended prices to the purchasing officer.

[0041] The memory unit 22 stores all data used for the determination (external price information, calculated values, scores, and determination results) as a history. This history is used for future model retraining and determination rule revisions, contributing to the continuous improvement of system accuracy.

[0042] Next, the effects of the price revision approval support system 10 configured as described above will be explained. With this embodiment configured as described above, it becomes possible to objectively and quickly make price revision approval decisions, which were previously dependent on human judgment and time-consuming, based on external market data and statistical analysis. In addition, the counter-proposal function ensures that opportunities for price negotiation are not missed, leading to a reduction in purchasing costs.

[0043] Furthermore, the price revision approval support system 10 of this embodiment can be continuously improved even after implementation. The judgment results, negotiation results, and actual order price data stored in the memory unit 22 are used to reconstruct the judgment rules and prediction models in the statistical calculation unit 16 and the price evaluation unit 18. This makes it possible to maintain flexible approval criteria that respond immediately to changes in the market environment and transaction conditions.

[0044] Furthermore, the historical data stored in the memory unit 22 can be used for purchasing strategy planning and creating reports for management. For example, the reduction effect from price negotiations over a certain period, the amount of cost reduction from adopting substitute products, and the average reduction in approval processing time can be aggregated and visualized as KPIs (Key Performance Indicators). This allows for a quantitative evaluation of the effects of system implementation and is effective as a basis for management decisions.

[0045] Furthermore, this system 10 can seamlessly integrate with purchasing systems and ERP (Enterprise Resource Planning) systems. Price approval results can be directly reflected in the order processing via CXML or OCI interfaces, eliminating the need for re-entering data after approval and preventing human input errors. In addition, in the case of non-approval or counter-proposals, feedback messages are automatically sent to suppliers via the EDI communication function to encourage prompt price renegotiation.

[0046] This configuration allows the purchasing department to significantly reduce the time required for approval decisions and increase the transparency of approval criteria. Furthermore, by utilizing external market data, the system automatically references the latest information to make decisions, eliminating the need for personnel to constantly monitor market trends, thus ensuring consistent approval quality that is not affected by differences in experience or knowledge.

[0047] Embodiments of the present invention can be customized according to the size and type of company. For example, for manufacturing companies, raw material price indices and exchange rate information can be added as external data sources, and for construction companies, material price statistics and public bidding price information can be linked. Furthermore, for companies with overseas bases, it can be applied to global procurement operations by providing a multilingual price evaluation screen and a price comparison function in local currency.

[0048] Furthermore, in the future, it will be possible to introduce AI-based reinforcement learning in the statistical calculation unit 16 and the price evaluation unit 18, and automatically update the optimal judgment rules based on feedback from approval results and negotiation results. Thus, the price revision approval support system 10 of the present invention is not merely a judgment support tool, but functions as a platform that supports the advancement of purchasing operations and strategic decision-making. In addition, by using generation AI to absorb format differences in external catalog data and integrating and organizing product specifications and price information using natural language processing, it becomes possible to utilize previously fragmented catalogs specific to each company as an integrated database. This mechanism makes it possible to centrally collect diverse data with different company sizes and trading partners, and to continuously improve the accuracy of price revision approval judgments. [Explanation of Symbols]

[0049] 2 network, 4 users, 10 price revision approval support system, 12 item extraction unit, 14 price acquisition unit, 16 statistical calculation unit, 18 price evaluation unit, 20 automatic approval unit, 22 storage unit.

Claims

1. A price revision approval support system that automatically evaluates the appropriateness of price revisions using external market data, An item extraction unit that extracts items subject to price revision based on past purchase history, The price acquisition unit acquires market price information from multiple external data sources for the aforementioned price revision target items, (a) The rate of deviation based on the difference between the proposed revised price stated in the price revision proposal and the market price based on the market price information mentioned above. (b) Standard deviation of the price distribution based on the market price information, (c) Market index fluctuation rate based on a predetermined index included in the external market data, A statistical calculation unit that calculates each of the following, A price evaluation unit that statistically evaluates the validity of the proposed price revision by determining whether the proposed price revision falls within a statistically acceptable range using the calculated deviation rate, standard deviation, and market index fluctuation rate, and determines whether to approve, conditionally approve, or reject the proposal. An automatic approval unit that automatically registers the approval result to the purchasing system if the judgment result is approval, If the aforementioned determination result is not approved, the notification unit automatically generates a proposal for an alternative product or a proposal for a price reduction for the item subject to price revision and notifies the purchasing officer's terminal. Equipped with, The price evaluation unit rejects the proposal if the deviation rate exceeds a predetermined deviation rate threshold, generates a substitute product proposal or price reduction proposal in response to the rejection if the market index fluctuation rate shows a negative value, excludes outliers based on statistical significance testing if there are three or more companies in the external data source used for comparison, targets only that company for a counter-proposal or rollback if only that company shows a statistically significant outlier, and uses the range of the mean ± predetermined coefficient multiple based on the standard deviation as the statistically acceptable range, thereby making approval decisions according to the price fluctuation characteristics of each product characteristic.

2. In the price revision approval support system described in claim 1, The price revision approval support system is characterized in that the external data source includes at least one of a corporate goods price index database, an online price comparison site, an electronic catalog server, and a company-specific catalog.

3. In the price revision approval support system described in claim 1, The price acquisition unit is characterized by acquiring the market price information from the external data source via a cXML, OCI, or CSV interface.

4. In the price revision approval support system described in claim 1, The aforementioned statistical calculation unit is a price revision approval support system characterized by accumulating price history for a predetermined period in the past as time-series data and predicting future prices using a learning model.

5. In the price revision approval support system described in claim 1, The price evaluation unit scores multiple evaluation items, and if the total score obtained by summing these scores is equal to or greater than the upper threshold, the price revision approval is granted; if the total score is less than the lower threshold, the price revision is rejected or a counter-proposal is made; and if the total score is less than the upper threshold and equal to or greater than the lower threshold, the price revision approval is granted conditionally.

6. In the price revision approval support system described in claim 1, A price revision approval support system characterized in that, in comparisons based on multiple external data sources, if only one company shows a statistically significant deviation, only that company will be subject to a counter-proposal or rejection.

Images