Client potential intelligent evaluation method and system based on dynamic weight adjustment

Abstract

The invention discloses a customer potential intelligent evaluation method and system based on dynamic weight adjustment. The method comprises the following steps: collecting related data from an internal system and an external API (Application Program Interface), and carrying out cleaning and standardization processing to obtain a preprocessing result; calculating the score of each customer based on the preprocessing result according to a preset dimension index and a weight so as to determine the customer level; converting the client address in the preprocessing result into a coordinate form, and recommending nearby potential clients based on the current position of the user; historical data are analyzed to construct a transformation prediction model, and the business opportunity loss risk is monitored and predicted in real time; and client data change in the preprocessing result is monitored, and an update notification is sent through an instant messaging tool according to a predefined rule. By implementing the method provided by the invention, the defects of the existing system in the aspects of customer grading, data analysis visualization, business opportunity tracking, field service support and the like can be accurately solved.

Description

Technical Field

[0001] This invention relates to data analysis methods, and more specifically to a method and system for intelligent assessment of customer potential based on dynamic weight adjustment. Background Technology

[0002] Current mainstream customer relationship management (CRM) systems primarily focus on functions such as customer information management, sales process tracking, and basic data report generation. These systems can record detailed customer information, transaction history, and communication records, and provide refined management of the conversion process from potential business opportunities to sales through the sales funnel. In addition, they offer various reporting functions such as sales performance analysis, customer classification statistics, and regional distribution reports, and some support mobile applications, allowing users to access information and manage tasks anytime, anywhere.

[0003] However, despite their powerful features, these systems still have significant shortcomings in several areas. First, existing systems largely rely on static data to assess customer value, lacking dynamic evaluation of customer potential, strategic value, and cooperation possibilities. This makes it difficult to identify high-potential customers and provide accurate customer segmentation recommendations. Second, these systems have weak visualization capabilities for regional and business development analysis, typically only displaying data through tables or simple charts, lacking advanced features such as interactive maps or real-time sales amount distribution maps. Furthermore, traditional opportunity funnel analysis lacks dynamic linkage, mobile terminal functions are limited and lack contextual design, performance and finance modules are disconnected, and customer potential assessment models are overly simplistic, all of which greatly limit their effectiveness and practicality. These problems have spurred the demand for a new generation of CRM solutions.

[0004] Therefore, it is necessary to design a new method to accurately address the shortcomings of existing systems in areas such as customer segmentation, data analysis and visualization, business opportunity tracking, and field support. Summary of the Invention

[0005] The purpose of this invention is to overcome the shortcomings of the prior art and provide a method and system for intelligent assessment of customer potential based on dynamic weight adjustment.

[0006] To achieve the above objectives, the present invention adopts the following technical solution: a customer potential intelligent assessment method based on dynamic weight adjustment, comprising:

[0007] Relevant data is collected from internal systems and external APIs, and then cleaned and standardized to obtain preprocessed results;

[0008] Based on the preprocessing results, a score for each customer is calculated according to preset dimension indicators and weights to determine the customer level;

[0009] The customer addresses in the preprocessed results are converted into coordinates, and potential customers nearby are recommended based on the user's current location.

[0010] Analyze historical data to build conversion prediction models and monitor and predict the risk of lost business opportunities in real time;

[0011] Monitor changes in customer data in the preprocessing results and send update notifications via instant messaging tools according to predefined rules.

[0012] The further technical solution is as follows: the internal system includes CRM and financial system; the external APIs include Qixinbao and Gaode Map.

[0013] The further technical solution is as follows: The calculation of each customer's score based on the preprocessing results according to preset dimension indicators and weights to determine the customer level includes:

[0014] Each dimension is assigned a corresponding weight based on the preset dimension indicators;

[0015] Based on the preprocessing results and the weights, the scores for each dimension are calculated and summed to obtain the total score;

[0016] Based on the total score, a nine-square grid analysis method is applied to comprehensively consider potential scores and strategic importance to determine customer levels.

[0017] Its further technical solution is as follows: Based on the total score, a nine-square grid analysis method is applied to comprehensively consider potential scores and strategic importance to determine customer levels, including:

[0018] Based on the total score, customers are divided into different levels in a nine-square grid. A nine-square grid coordinate system is defined, and corresponding interaction logic is set to determine the customer level. The X-axis of the nine-square grid coordinate system represents the potential score, and the Y-axis represents the strategic importance.

[0019] The further technical solution is as follows: converting the customer address in the preprocessing result into coordinate form and recommending nearby potential customers based on the user's current location includes:

[0020] Geocoding technology is used to convert customer addresses in the preprocessed results into coordinates for storage. When business development personnel use mobile devices, they can obtain the current location and recommend nearby potential customers by comparing distances.

[0021] Its further technical solution is: the analysis of historical data to construct a conversion prediction model, and the real-time monitoring and prediction of the risk of lost business opportunities, including:

[0022] Extract and clean historical business opportunity flow data from the enterprise's business system to obtain historical data;

[0023] Based on the historical data, the key stages of the business opportunity funnel are identified, and all business opportunities are classified and managed accordingly.

[0024] Based on whether the business opportunity was successfully converted or lost and the dwell time, each business opportunity event is labeled with positive and negative samples to obtain a sample set;

[0025] Based on the sample set, the number of successful conversions of business opportunities at each stage is statistically analyzed, and the conversion rate is calculated and analyzed.

[0026] Based on the number of successful conversions and the conversion rate, a funnel chart is used to visually present the number of business opportunities and their conversion status at each stage.

[0027] By combining the current status of business opportunities with other influencing factors, a conversion prediction model is built using machine learning algorithms to predict future business opportunity trends and churn risk.

[0028] The further technical solution is as follows: The monitoring of changes in customer data within the preprocessing results, and the sending of update notifications via instant messaging tools according to predefined rules, includes:

[0029] Listen for customer data change events and define key trigger rules; when customer data changes, send instant notifications via Lark or other platforms.

[0030] This invention also provides a customer potential intelligent assessment system based on dynamic weight adjustment, comprising:

[0031] The acquisition unit is used to collect relevant data from internal systems and external APIs, and to clean and standardize the data to obtain preprocessed results.

[0032] The rating determination unit is used to calculate the score of each customer based on the preprocessing results according to preset dimension indicators and weights, so as to determine the customer rating.

[0033] The address processing unit is used to convert the customer address in the preprocessing result into coordinate form and recommend nearby potential customers based on the user's current location;

[0034] The prediction unit is used to analyze historical data to build conversion prediction models and monitor and predict the risk of lost business opportunities in real time.

[0035] The monitoring unit is used to monitor changes in customer data in the preprocessing results and send update notifications via instant messaging tools according to predefined rules.

[0036] The further technical solution is as follows: the level determination unit includes:

[0037] The weight assignment subunit is used to assign corresponding weights to each dimension based on preset dimension indicators;

[0038] The summarization subunit is used to calculate the score of each dimension based on the preprocessing results and the weights, and to summarize the total score.

[0039] The analysis subunit is used to apply the nine-square grid analysis method based on the total score, comprehensively considering potential score and strategic importance, to determine customer level.

[0040] The further technical solution is as follows: The analysis subunit is used to classify customers into different levels in the nine-square grid based on the total score, define the nine-square grid coordinate system, and set the corresponding interaction logic to determine the customer level, wherein the X-axis of the nine-square grid coordinate system represents the potential score; the Y-axis represents the strategic importance.

[0041] The advantages of this invention compared to existing technologies are as follows: By collecting data from internal systems and external APIs and performing cleaning and standardization, this invention first ensures data quality and consistency, laying a solid foundation for subsequent analysis. Based on the preprocessed data, customer scores are calculated according to set dimensional indicators and weights, achieving accurate customer segmentation. Converting customer addresses into geographic coordinates not only facilitates the discovery of business opportunities based on geographical location but also recommends nearby potential customers based on the user's current location, enhancing the effectiveness of field support. Constructing a conversion prediction model to analyze historical data and monitor business opportunity status in real time enables early warning of churn risks and optimizes the business opportunity tracking process. Furthermore, the function of monitoring changes in customer data and sending update notifications via instant messaging tools ensures timely information delivery, greatly improving work efficiency and response speed, thus comprehensively solving the shortcomings of existing systems in customer segmentation, data analysis visualization, business opportunity tracking, and field support.

[0042] The present invention will be further described below with reference to the accompanying drawings and specific embodiments. Attached Figure Description

[0043] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the following description of the embodiments will be briefly introduced. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0044] Figure 1 This is a schematic diagram illustrating an application scenario of the intelligent customer potential assessment method based on dynamic weight adjustment provided in this embodiment of the invention.

[0045] Figure 2 A flowchart illustrating the intelligent customer potential assessment method based on dynamic weight adjustment provided in this embodiment of the invention;

[0046] Figure 3A schematic diagram of a sub-process of the intelligent customer potential assessment method based on dynamic weight adjustment provided in an embodiment of the present invention;

[0047] Figure 4 A schematic diagram of a sub-process of the intelligent customer potential assessment method based on dynamic weight adjustment provided in an embodiment of the present invention;

[0048] Figure 5 A schematic block diagram of a customer potential intelligent assessment system based on dynamic weight adjustment provided in an embodiment of the present invention;

[0049] Figure 6 A schematic block diagram of a computer device provided for an embodiment of the present invention. Detailed Implementation

[0050] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0051] It should be understood that, when used in this specification and the appended claims, the terms "comprising" and "including" indicate the presence of the described features, integrals, steps, operations, elements and / or components, but do not exclude the presence or addition of one or more other features, integrals, steps, operations, elements, components and / or collections thereof.

[0052] It should also be understood that the terminology used in this specification is for the purpose of describing particular embodiments only and is not intended to limit the invention. As used in this specification and the appended claims, the singular forms “a,” “an,” and “the” are intended to include the plural forms unless the context clearly indicates otherwise.

[0053] It should also be further understood that the term "and / or" as used in this specification and the appended claims refers to any combination of one or more of the associated listed items and all possible combinations, and includes such combinations.

[0054] Please see Figure 1 and Figure 2 , Figure 1 This is a schematic diagram illustrating an application scenario of the intelligent customer potential assessment method based on dynamic weight adjustment provided in this embodiment of the invention. Figure 2This is a schematic flowchart illustrating the intelligent customer potential assessment method based on dynamic weight adjustment provided in this embodiment of the invention. This intelligent customer potential assessment method based on dynamic weight adjustment is applied in a server. The server interacts with the terminal, collects and processes data from internal systems and external APIs, calculates customer scores based on preset dimension indicators and dynamic weights to determine the customer level, and uses geocoding technology and the user's current location to recommend nearby potential customers, thus solving the shortcomings of existing systems in customer segmentation. Simultaneously, it constructs a conversion prediction model by analyzing historical data, monitors and predicts the risk of lost opportunities in real time, and achieves precise tracking and management of the opportunity process. Furthermore, it refines customer classification by considering potential scores and strategic importance using a nine-square grid analysis method, enhancing the visualization effect of data analysis. Finally, the function of monitoring changes in customer data and sending update notifications via instant messaging tools strengthens support for field staff, effectively improving work efficiency and response speed. This comprehensive solution significantly improves the existing system functions in customer segmentation, data analysis visualization, opportunity tracking, and field support.

[0055] Figure 2 This is a flowchart illustrating the intelligent customer potential assessment method based on dynamic weight adjustment provided in an embodiment of the present invention. Figure 2 As shown, the method includes the following steps S110 to S150.

[0056] S110. Collect relevant data from internal systems and external APIs, and perform cleaning and standardization to obtain preprocessed results.

[0057] In this embodiment, the preprocessing result refers to the structured data set after data collection, cleaning, and standardization.

[0058] The internal systems include CRM and financial systems; the external APIs include Qixinbao and Gaode Maps.

[0059] Internal system data collection includes:

[0060] CRM systems are primarily responsible for collecting basic customer information, business opportunity status, project progress, and past cooperation records. This information is crucial for assessing customer potential because it provides direct evidence of business interactions.

[0061] Financial system: Used to obtain information about customer transactions, including contract amounts and payment status. This helps assess a customer's financial health and payment ability, and is one of the important dimensions for calculating customer value.

[0062] CTMS (Clinical Trial Management System): Provides detailed progress information on projects, such as trial phases and research centers, which is crucial for understanding clients' activities in the pharmaceutical R&D field.

[0063] Salesforce: Real-time synchronization of business opportunity information, including inquiry records and bidding status, helps track potential sales opportunities.

[0064] External API data collection includes:

[0065] Qixinbao API: This API retrieves a company's business registration information, such as registered address, financing history, and affiliated companies, through a RESTful interface. This is extremely useful for supplementing and improving a client's company background information, especially when dealing with newly contacted or previously uncooperative clients.

[0066] Amap API: Primarily used for geocoding services, converting customer-provided registration addresses into latitude and longitude coordinates to facilitate subsequent location-related analyses, such as recommending nearby potential customers.

[0067] Data cleaning and standardization

[0068] The collected data is often disorganized and inconsistently formatted, therefore it needs to undergo a series of cleaning and standardization steps:

[0069] Operations such as removing duplicates, correcting erroneous data, and filling in missing values ​​ensure the accuracy and completeness of the dataset.

[0070] Integrate data from different sources, such as by using ID mapping technology to match internal and external data based on key fields like customer taxpayer identification number or full name, to build a unified customer identifier.

[0071] The address information can be converted into geographic coordinates using the Amap API, which facilitates spatial analysis.

[0072] A unified data model is used to standardize the storage of all processed information, enabling data from different sources to be accessed and analyzed within the same framework.

[0073] Ultimately, the preprocessing result obtained through the above process is a structured and standardized dataset that not only contains comprehensive customer information but also provides a solid foundation for subsequent intelligent analysis. This process is a crucial first step in the entire customer potential intelligent assessment method based on dynamic weight adjustment, directly impacting the functional implementation of subsequent modules.

[0074] S120. Based on the preprocessing results, calculate the score of each customer according to the preset dimension indicators and weights to determine the customer level.

[0075] In this embodiment, customer level refers to the intelligent customer potential assessment model technology based on dynamic weight adjustment of multi-source data fusion. It integrates internal and external data sources to build a complete 360-degree view of customers and calculates the score of each customer according to preset dimension indicators and weights to determine their level.

[0076] In one embodiment, please refer to Figure 3 The above-mentioned step S120 may include steps S121 to S123.

[0077] S121. Assign corresponding weights to each dimension according to the preset dimension indicators.

[0078] In this embodiment, firstly, it is necessary to define and assign corresponding weights to each dimension. These dimensions typically include, but are not limited to, the following aspects:

[0079] Financial information, such as: customer type, listing status, financing status, etc.;

[0080] Research and development strategies include: drug type, disease area, and application strategy;

[0081] Cooperation history and potential, such as past satisfaction, corporate win rate, Tiger Investment status, etc.;

[0082] Budget adequacy, for example: total number of pipelines.

[0083] The specific metrics for each dimension will be selected and adjusted according to business needs, and each metric will be assigned a weight to reflect its importance in the overall evaluation system.

[0084] S122. Calculate the scores for each dimension based on the preprocessing results and the weights, and sum them up to obtain the total score.

[0085] In this embodiment, each dimension is then scored based on the preprocessing results and the aforementioned weights. This process includes:

[0086] Data obtained from financial systems, project management systems, CRM systems, and external APIs (such as Qixinbao and Gaode Maps) is cleaned and transformed before being used as input. Based on the indicators and corresponding weights of each dimension, scores are calculated for each dimension. For example, for the "financial situation" dimension, factors such as customer type and listing status can be considered to give a specific score. All dimensions' scores are then summed to obtain a total score.

[0087] S123. Based on the total score, the nine-square grid analysis method is applied to comprehensively consider the potential score and strategic importance to determine the customer level.

[0088] In this embodiment, customers are divided into different levels in a nine-square grid based on the total score. A nine-square grid coordinate system is defined, and corresponding interaction logic is set to determine the customer level. The X-axis of the nine-square grid coordinate system represents the potential score, and the Y-axis represents the strategic importance.

[0089] Finally, the nine-square grid analysis method is applied based on the total score. This method mainly considers two dimensions:

[0090] X-axis: Potential score (range 0-100), reflecting the potential value of the client's future development.

[0091] Y-axis: Strategic importance (range 1-5), reflecting the client's importance to the company's long-term strategic goals.

[0092] In this coordinate system, customers are divided into different levels (such as A1, A2, B1, etc.). For example, if a customer has a potential score of 80 and a strategic importance score of 4, then that customer might be classified as level "A2". Furthermore, interactive logic can be set up, such as clicking on a specific level area to view a list of customers at that level or generate a detailed potential customer report.

[0093] This systematic approach not only allows for the quantitative assessment of each client's potential and strategic value but also provides businesses with scientific decision-making support, optimizing resource allocation and enhancing market competitiveness. Furthermore, considering market changes and adjustments to company strategy, this model supports dynamic parameter adjustments to ensure that assessment results always reflect reality.

[0094] S130. Convert the customer address in the preprocessing result into coordinate form, and recommend nearby potential customers based on the user's current location.

[0095] In this embodiment, geocoding technology is used to convert the customer addresses in the preprocessed results into coordinates for storage. When business development personnel use mobile devices, they can obtain the current location and recommend nearby potential customers by comparing distances.

[0096] In this embodiment, in order to realize the intelligent recommendation function based on geolocation, geocoding technology is used to convert the customer address in the preprocessing result into coordinate form for storage, and to recommend potential customers nearby based on the user's current location.

[0097] First, customer registration addresses obtained from various data sources (such as CRM systems, Qixinbao, etc.) need to be geocoded. This process includes:

[0098] Address resolution: Using the Amap API or other similar services, we convert customers' text address information into precise latitude and longitude coordinates.

[0099] Coordinate storage: The obtained latitude and longitude coordinates are stored in the MySQL database in the form of POINT(lon,lat) for subsequent queries and calculations.

[0100] When a Business Development (BD) professional opens a mobile application (such as the Lark mini-program), the system automatically retrieves their current location information. Then, it calculates and recommends nearby potential customers using the following methods:

[0101] Obtaining the current location (longitude and latitude) of BD personnel can typically be achieved using the GPS function of a mobile device.

[0102] Use the RedisGEO module to efficiently calculate the straight-line distance between business development personnel and all known customers. RedisGEO supports quickly finding points within a specified radius, making it ideal for this type of application.

[0103] Based on the distances calculated above, the system can sort and recommend the most recent potential clients to business development personnel according to certain rules (such as industry relevance, potential score, etc.). Specific operations include:

[0104] The app displays all high-potential clients within a 5-kilometer radius who are not currently partners with the company, and prioritizes them based on specific criteria (such as industry fit, funding status, etc.).

[0105] For existing clients, competitor clients, and high-potential unreached clients that are specifically marked, different icons or colors can be used to distinguish them on the map for easy identification.

[0106] In addition, additional functions can be integrated, such as clicking on a customer tag to view detailed information, including but not limited to contact information, business opportunity status, and historical interaction records, thereby helping business development personnel to carry out business development work more effectively.

[0107] This service solution, combining geocoding technology and real-time location, not only significantly improves the work efficiency of business development (BD) personnel but also facilitates the implementation of precision marketing strategies, bringing enterprises greater market competitiveness. Furthermore, with data accumulation and technological optimization, the system is expected to further enhance its personalized recommendation capabilities, providing a service experience that better meets actual needs.

[0108] S140. Analyze historical data to build a conversion prediction model, and monitor and predict the risk of lost business opportunities in real time.

[0109] In this embodiment, the conversion prediction model refers to the analysis of historical business opportunity flow data through machine learning algorithms to identify key factors affecting the successful conversion of business opportunities, and to use these factors to predict future business opportunity trends and potential churn risks.

[0110] The risk of lost business opportunities refers to the risk that a business opportunity may fail to be converted into actual business due to various reasons (such as unreasonable pricing or competitor advantages) during the process from initial contact to final transaction. This risk can occur at any stage of the business opportunity funnel, including inquiry and quotation, bidding, and contract signing.

[0111] In one embodiment, please refer to Figure 4 The above-mentioned step S140 may include steps S141 to S146.

[0112] S141. Extract and clean historical business opportunity flow data from the enterprise business system to obtain historical data.

[0113] In this embodiment, historical data refers to records of all business opportunities' status changes over a past period, including but not limited to the source of the opportunity, the person responsible for business development, the duration of each stage, and whether a transaction was ultimately completed. The data cleaning process aims to remove duplicates, correct erroneous data, and fill in missing values ​​to ensure the accuracy of subsequent analysis.

[0114] S142. Based on the historical data, identify the key stages of the business opportunity funnel and classify and manage all business opportunities accordingly.

[0115] In this embodiment, business opportunities are categorized according to their development stages, such as potential opportunities, inquiry and quotation, bidding, pending contract, and winning the deal. This helps to more accurately understand the characteristics of different stages and their transition patterns.

[0116] S143. Based on whether the business opportunity was successfully converted or lost and the dwell time, label each business opportunity event with positive and negative samples to obtain a sample set.

[0117] In this embodiment, positive samples represent business opportunities that successfully complete a certain stage and move to the next stage; negative samples represent business opportunities that stagnate at a certain stage or are ultimately lost. Dwell time is also an important consideration; prolonged dwell time at a certain stage often indicates a higher risk of churn.

[0118] S144. Based on the sample set, count the number of successful conversions of business opportunities at each stage, and calculate and analyze the conversion rate.

[0119] In this embodiment, the number of successful conversions refers to the number of business opportunities that can smoothly transition from the current stage to the next stage.

[0120] Conversion rate is an important indicator for measuring the efficiency of a stage, and it is usually calculated by dividing the number of successful conversions by the total number of business opportunities in that stage.

[0121] S145. Based on the number of successful conversions and the conversion rate, a funnel chart is used to visually present the number of business opportunities and their conversion status at each stage.

[0122] In this embodiment, the funnel chart is a very effective visualization tool that can clearly show the changes in the number of business opportunities and conversion efficiency at each stage, helping to identify bottlenecks.

[0123] S146. Combining the current status of business opportunities with other influencing factors, use machine learning algorithms to build a conversion prediction model to predict future business opportunity trends and churn risk.

[0124] By analyzing the data above and employing appropriate machine learning algorithms (such as decision trees, random forests, and neural networks), a predictive model can be built to predict whether new business opportunities will be successfully converted or whether there is a risk of churn. This not only helps to take proactive measures to reduce the probability of churn but also guides the effective allocation of resources.

[0125] Finally, throughout the process, S150 emphasized the importance of monitoring changes in customer data and sending update notifications via instant messaging tools to ensure that relevant personnel can respond to market dynamics in a timely manner and optimize business strategies.

[0126] In this embodiment, relevant data on business opportunity flow is extracted from the company's business system, including but not limited to the dwell time of business opportunities at each stage and whether they have been successfully converted. This raw data is then cleaned and processed to ensure its quality and consistency, preparing it for subsequent analysis.

[0127] Identify the five key stages of the business opportunity funnel: potential opportunity, inquiry and offer, bidding, contract pending, and final win, and categorize opportunities according to these stages. This helps to clearly understand where each opportunity stands in its lifecycle and its possible development path.

[0128] The labels needed to build a training model based on the actual flow of business opportunities:

[0129] Positive sample: indicates that the business opportunity has successfully moved from the previous stage to the next stage (such as smoothly moving from the inquiry and quotation stage to the bidding stage).

[0130] Negative samples: These indicate that business opportunities were lost at a certain stage (such as going directly from inquiry and quotation to lost order), or that the business remained at a certain stage for more than 30 days without any progress.

[0131] For each stage of a business opportunity, count the number of successful conversions and calculate the corresponding conversion rate. This step is crucial for understanding how business opportunities flow between different stages and is the foundation for predicting future business opportunity trends.

[0132] Funnel charts visually represent the changes in the number of opportunities and their conversion rates at each stage. This visualization method helps teams quickly identify bottlenecks in the opportunity process and develop targeted strategies for optimization.

[0133] By combining current business opportunity order conversion rates with other influencing factors (such as market dynamics and competitor activities), a predictive model is built using machine learning algorithms. This model aims to anticipate the risk of business opportunity loss and thus take proactive measures to mitigate that risk.

[0134] Offline training: The prediction model is updated weekly to ensure that it can reflect the latest market conditions and business changes in a timely manner.

[0135] Online inference: Get real-time predictions of lead conversions by calling PyTorch Serving services in real time, enabling sales teams to make faster and smarter decisions.

[0136] By following the steps above, a comprehensive opportunity funnel conversion prediction system can be effectively built. This system can not only help identify potential opportunities lost, but also guide companies to optimize their sales strategies and improve overall sales efficiency.

[0137] S150. Monitor changes in customer data in the preprocessing results and send update notifications via instant messaging tools according to predefined rules.

[0138] In this embodiment, customer data change events are monitored, and key triggering rules are defined; when customer data changes, an instant notification is sent via Lark or other platforms.

[0139] In this embodiment, to ensure that relevant personnel can promptly obtain information on changes in key customer data and respond quickly accordingly, the system is designed with a message push mechanism based on real-time data monitoring and instant messaging tools. The following is a detailed explanation of this mechanism:

[0140] Establish an efficient data change monitoring mechanism to monitor changes in customer data within the data warehouse in real time. This typically involves the application of database triggers or message queue technologies to capture these changes immediately when they occur.

[0141] Determine which types of data changes should trigger notifications. For example, when a customer's potential rating changes (e.g., from B2 to A1), project status is updated, new business opportunities emerge, or the status of existing business opportunities changes.

[0142] Based on business needs, define what data changes should trigger notifications. For example, notifications should only be issued when a customer's potential rating is upgraded from a low level to a high level, or when a high-value business opportunity is about to be lost.

[0143] Different data changes may have different levels of urgency, so different notification priorities need to be set. This ensures that the most important information is processed first.

[0144] Once the notification to be sent is determined, the system will call the instant messaging tool's API to send the message. The example mentioned here uses Lark's ` / message / card / send` API to send a message card containing deep links, allowing the recipient to directly jump to the relevant details page to view specific information.

[0145] Each notification should include a clear summary of the information and necessary operational guidance. For example, a message such as, "Customer [Customer Name]'s potential rating has been upgraded from B2 to A1. Please follow up," not only informs the user of the change but also suggests the actions to be taken next.

[0146] All sent notifications should be logged, including the notification type, the customer ID involved, the recipient, and the notification status (read / unread). This can be achieved by maintaining a MySQL table (such as message_log).

[0147] To enable users to easily manage and configure their message subscriptions, the system should provide a corresponding UI interface on the front end. Users can use this interface to view historical notifications, adjust notification preferences, and so on. Considering the large-scale user scenario, efficiently storing each user's subscription preferences becomes a challenge. Here, it is recommended to use Redis Bitmap to store the message on / off configuration between the business development team (BD) and clients, which saves space and improves query efficiency.

[0148] Through the above steps, the S150 not only enables real-time monitoring and intelligent alerts for changes in customer data, but also ensures the effectiveness and relevance of information delivery, thereby helping business teams better grasp market dynamics, optimize resource allocation, and improve work efficiency.

[0149] In this embodiment, the method is particularly applicable to the pharmaceutical R&D service industry. This technology integrates internal and external data sources to provide a full-chain solution from customer analysis and business opportunity mining to intelligent push notifications. It adopts a "PC + mobile" dual-terminal collaborative architecture to achieve real-time synchronization and intelligent analysis of customer data.

[0150] First, customer profiling analysis is used. The group profile provides statistical information such as the total number of customers, the distribution of signed contracts (map / BD group), and the conversion rate of the opportunity funnel. The individual profile allows you to query detailed information such as the potential rating (e.g., A1 / B2), opportunity status, project progress, and financing history of a single customer.

[0151] Search for customers on mobile devices and view their geographic location and potential customers within a 5-kilometer radius; analyze sales amount and win rate by region or business development group, and support time trend comparison; perform end-to-end conversion analysis from potential opportunities to winning deals to help identify bottlenecks.

[0152] BDs are ranked TOP10 / Bottom10 based on order amount and number of CDA agreements; global data can be viewed, while regular BDs can only view the ranking of their own group.

[0153] Push key information such as changes in customer level and pipeline changes, and click to go directly to the details page; perform a fuzzy search for customers (including leads that have not been entered), and view nearby business opportunities in map mode.

[0154] It supports exporting customer details, potential assessment forms, project data, etc., to Excel / PDF format; and directly links from charts or tables to related pages (such as customer profiles - potential assessment).

[0155] The system adopts a distributed deployment architecture, which is divided into the following layers:

[0156] Data acquisition layer: including internal data sources (CRM system, project management system, financial system) and external data sources (Qixinbao enterprise information, Gaode Map API, etc.).

[0157] Data processing layer: includes data cleaning module, data fusion module, geocoding module, etc.

[0158] Data analysis layer: includes customer profiling analysis engine, business opportunity funnel analysis module, potential assessment model, etc.

[0159] Application service layer: includes PC application services, mobile application services, push notification services, etc.

[0160] Customer master data comes from the CRM system and is updated in real time; transaction records come from the financial system and are updated in batches on T+1 days; project data comes from the CTMS system and is updated almost in real time (every 5 minutes); business opportunity information comes from Salesforce and is updated in real time.

[0161] Qixinbao API is used for completing business registration information, with a QPS limit of 100 times / second; Gaode Map WebService is used for address coordinate conversion, with a QPS limit of 500 times / day; financing data is collected through web crawlers and is mainly used for historical financing analysis.

[0162] The customer potential assessment model enables quantitative evaluation and dynamic grading of customer value; it covers data from multiple dimensions such as finance, R&D, and cooperation history; it collects customer data from various dimensions, assigns different weights, automatically calculates a comprehensive score, and classifies customers into different levels.

[0163] The system uses the Gaode Map API for coordinate parsing, RedisGEO for distance calculation, and Flink for real-time event processing. It converts customer registration addresses into latitude and longitude coordinates, locates business development personnel in real time, and recommends all customers within a 5-kilometer radius.

[0164] The details are shown in Tables 1 and 2.

[0165] Table 1. Internal Data Sources

[0166]

[0167] Table 2. External Data Sources

[0168] Data categories Interface type Main uses QPS Limit Qixinbao API RESTful Business registration information completion 100 times / second Gaode Map WebService Address coordinate conversion 500 times / day Financing data crawler collection Historical Financing Analysis -

[0169] The test environment is as follows:

[0170] Server: 172.18.17.146;

[0171] Components: MySQL 8.0.24, Redis 6.2.7, Nginx 1.26.1;

[0172] Application: Front-end (Vue.js) port 81, back-end (Java) port 8081;

[0173] The production environment is as follows:

[0174] Database server: 172.22.36.146 (MySQL 8.0.24, Redis 6.2.13);

[0175] Application server: 172.22.36.145 (Nginx 1.17.9, front-end and back-end applications);

[0176] Security configuration: Database password protection, Redis access control.

[0177] The entire process of transformation from potential opportunities to winning deals is analyzed; offline training updates the model weekly, and online inference calls PyTorchServing in real time.

[0178] Multi-source data integration and processing technologies ensure data consistency across systems; ETL technology is used to integrate internal data, and external data sources are accessed through API interfaces to establish a unified customer data model.

[0179] Intelligent alerts based on dynamic customer changes support real-time push notifications on Lark mobile devices, with user-configurable push rule management; a customer data change monitoring mechanism is established, key trigger rules are defined, and instant notifications are sent through Lark or other platforms.

[0180] In this embodiment, the data sources for achieving quantitative assessment and dynamic grading of customer value are shown in Table 3.

[0181] Table 3. Data Sources

[0182]

[0183] Specifically, customer data is collected from various dimensions such as finance, R&D, and cooperation history; each dimension is assigned a different weight (e.g., R&D investment accounts for 40%, and financial health accounts for 30%); a comprehensive score (0-100 points) is automatically calculated; and customers are divided into nine levels (A1 / A2 / B1, etc.) according to the score.

[0184] The nine-square grid analysis is shown in Table 4, with coordinates defined as follows:

[0185] X-axis: Potential score (0-100);

[0186] Y-axis: Strategic importance (levels 1-5);

[0187] Table 4. Nine-square grid

[0188]

[0189] Interaction logic:

[0190] Click cell A1 - Displays the list of customers at that level;

[0191] Right-click menu - Generate potential customer report.

[0192] The model quantifies customer potential and cooperation levels using a series of specific parameters and indicators, thereby achieving accurate customer segmentation. In the potential component, the model primarily considers financial situation (50% weight) and R&D strategy (50% weight). Regarding financial situation, scoring is based on customer type, listing status, financing status, and transaction details; R&D strategy is scored based on drug type, disease area, and application strategy. For example, MNC-type customers can receive 9 points, while projects using MRCT application strategies or involving novel drug types such as ADCs also receive higher scores.

[0193] For the cooperation level component, the model considers past satisfaction, cooperation potential, and budget adequacy. Past satisfaction is scored based on the amount and frequency of cooperation, while the company's win rate and Tiger's investment situation play a crucial role in assessing the cooperation potential. Budget adequacy is evaluated based on the total number of pipelines. These factors collectively determine the client's cooperation level score.

[0194] Customer tiering is based on a combination of potential and cooperation scores. Potential scores are categorized into A (7-9 points), B (5-7 points), and C (3-5 points), while cooperation scores are categorized into 1 (7-9 points), 2 (5-7 points), and 3 (3-5 points). A nine-square grid of tiers combines potential and cooperation scores to create unique customer tier labels such as A1 and B2, visually reflecting the customer's value proposition. Furthermore, the system supports dynamic adjustment of model parameters based on the company's key business development strategies, ensuring the accuracy and adaptability of the assessment.

[0195] Using a pharmaceutical client as an example, this demonstrates how to calculate their potential and cooperation scores. This client, being a multi-market company (MNC), already marketed, and possessing a robust pipeline, received a high potential score of 7.2, classifying them as Grade A. Regarding cooperation, thanks to high satisfaction and a sufficient budget, they achieved a score of 7.95, placing them at Grade 1. Therefore, the client was ultimately rated A1, indicating not only significant potential but also extremely high cooperation, making them a highly valuable partner. This detailed and quantitative assessment method provides companies with a scientific basis for decision-making, helping to improve the efficiency and accuracy of client management.

[0196] The location-based real-time business opportunity radar technology aims to achieve intelligent recommendations through the physical space dimension. Its core technology stack includes the Amap API for coordinate parsing, RedisGEO for distance calculation, and Flink for real-time event processing. The workflow first converts customer addresses into geographic coordinates and stores them in MySQL. When business development (BD) personnel open the Lark mini-program, the system obtains their current location and displays all customers within a 5-kilometer radius, sorted by industry relevance, while also highlighting existing clients, competitor clients, and high-potential unreached clients.

[0197] To predict the risk of lost business opportunities, a conversion prediction algorithm based on the opportunity funnel was proposed. This algorithm defines five stages from potential opportunity to winning the deal, and updates the model weekly through offline training, while online inference relies on real-time calls to PyTorchServing. Label construction considers both positive and negative samples to reflect successful conversions or churn at different stages. The entire process begins with data extraction from the business system, followed by cleaning and processing to statistically analyze the conversion rates at each stage, and finally, a visual funnel chart to present the predicted business opportunity trends.

[0198] Multi-source data integration and processing technologies ensure data consistency across systems, encompassing internal data such as customer basic information and order data from Tigermed's business system, as well as external data such as corporate credit information from Qixinbao, geolocation services from Gaode Maps, and identity authentication and push notification functions from Lark. ETL technology and API interfaces are used to achieve efficient integration of internal and external data and establish a unified customer data model. Specific steps include data extraction, cleaning and standardization, supplementing with external information, ID mapping, and storage in a data warehouse and cache.

[0199] Finally, real-time push notification technology allows for intelligent alerts based on dynamic changes in customer behavior, supports real-time push notifications on Lark mobile devices, and enables users to customize push rules. This technology implements a mechanism for monitoring changes in data warehouse data, sends notifications based on rules triggered by key events, and utilizes Lark's message card API for real-time message delivery. Furthermore, a comprehensive push control scheme has been designed, including message storage, front-end subscription status updates, and the use of Redis Bitmap to manage message on / off configurations on the user side. These technologies together constitute a complete and efficient customer relationship management and opportunity mining solution.

[0200] This embodiment integrates radar charts, word clouds, and a nine-square grid evaluation model to dynamically analyze customer potential, cooperation level, and financial health. It supports visualized sales data by province, BD group, and time dimension, with clicks linking to detailed pages. It tracks opportunity conversion rates in real time, supports penetrating queries, and quickly identifies problem areas. It recommends nearby customers based on geographic location and provides real-time message alerts to meet the needs of field BD operations. It combines the BD leaderboard with accounts receivable collection rate analysis to provide a comprehensive view of business health. It generates a comprehensive rating through multi-dimensional scoring, supporting precise resource allocation.

[0201] This embodiment integrates multi-source data such as customer basic information, business opportunity data, and order records to form a unified 360-degree view of customers; it automatically calculates customer ratings based on radar charts to assist business development (BD) personnel in screening high-value customers; it visualizes customer conversion rates to identify bottlenecks; it displays the distribution of order amounts by province using heat maps to assist in regional market strategy formulation; it intuitively extracts customer characteristics through keywords to improve cognitive efficiency; it categorizes customers by rating to quickly identify high-potential and high-risk customers; BD personnel can query customer information and receive message reminders at any time via mobile devices to improve field work efficiency; it pushes key events such as changes in customer pipelines and rating adjustments in real time to avoid missing business opportunities; it has a built-in Lark spreadsheet link to collect user suggestions and continuously optimize the system experience; senior management focuses on macro indicators, while ordinary BD personnel focus on practical data such as personal rankings and business opportunity conversions; it publicly ranks orders such as order amount and CDA agreement number to stimulate team competitiveness; business opportunity funnel analysis reduces ineffective follow-ups and improves order conversion rates; the collection rate ranking function prompts BD personnel to proactively collect debts and reduce bad debt rates; and it intervenes in low-satisfaction customers in advance through potential ratings and message reminders.

[0202] The method in this embodiment utilizes radar charts to comprehensively assess multiple aspects of a client, such as financial health, potential for cooperation, drug type, and R&D strategy, and automatically generates classification labels such as A1 / B2 through algorithms and data analysis. This method not only replaces traditional subjective rating methods but also allows for dynamic adjustment of model parameters based on the company's key business development strategies.

[0203] Users can click on any rating in the nine-square grid (e.g., A1) to jump to a detailed analysis page, where they can view the distribution of similar customers or specific evaluation criteria in real time, forming a closed-loop process of "classification-analysis-decision".

[0204] Traditional CRM systems rely on human experience to judge customer value, while the method in this embodiment achieves dynamic quantification of customer value through a data-driven potential assessment model, thereby improving the accuracy and efficiency of the assessment.

[0205] On mobile maps, not only can the location of target customers be displayed, but also potential customers within a 5-kilometer radius (such as competitors or related companies in the industry chain) can be intelligently recommended, helping business development personnel to expand the offline market more effectively; it supports searching for and locating customers not yet recorded in the system (such as "Tencent Technology"), automatically marking them as unreachable and quickly adding them to the business opportunity pool.

[0206] Traditional customer management systems lack location-based association functions, while the method in this embodiment combines location-based services (LBS) with CRM, achieving seamless integration of online data and offline scenarios, and improving the efficiency and accuracy of business opportunity discovery.

[0207] When significant events occur, such as changes in a client's pipeline or rating adjustments, the system will automatically push messages to the Business Development (BD) personnel responsible for that client and provide a direct jump to the client profile page, ensuring timely information delivery and reducing delays. BD personnel can disable message notifications for non-core clients as needed to avoid information overload, while improving their focus on and response speed to changes in high-value clients.

[0208] Traditional systems require manual checks on customers' latest activities periodically, while the method in this embodiment uses an automated message push mechanism to ensure that business development (BD) can pay attention to important customer changes in a timely manner, greatly improving work efficiency and response speed.

[0209] This embodiment addresses the problems of static nature, poor industry adaptability, and lack of data validation inherent in traditional customer assessment models by dynamically adjusting model parameters and weights, utilizing LBS technology for scenario-based opportunity mining, and employing an event-triggered real-time message push mechanism. This system not only upgrades from static management to dynamic intelligent operation but also deeply integrates business intelligence (BI), customer relationship management (CRM), and geographic information technology. By driving decision-making with data and linking business with specific scenarios, it significantly improves the accuracy of customer value identification, the timeliness of opportunity delivery, and the collaborative efficiency of the business development team. These improvements effectively solve pain points in traditional customer management such as reliance on manual processes, delayed responses, and lack of geographic dimension correlation, providing enterprises with an intelligent, scenario-based, full-lifecycle customer management solution with high commercial value and broad industry application potential.

[0210] The aforementioned intelligent customer potential assessment method based on dynamic weight adjustment first ensures data quality and consistency by collecting, cleaning, and standardizing data from internal systems and external APIs, laying a solid foundation for subsequent analysis. Based on the preprocessed data, customer scores are calculated according to set dimensional indicators and weights, achieving accurate customer segmentation. Converting customer addresses to geographic coordinates not only facilitates geographical opportunity mining but also recommends nearby potential customers based on the user's current location, enhancing the effectiveness of field support. A conversion prediction model is built to analyze historical data and monitor opportunity status in real time, enabling early warning of churn risk and optimizing opportunity tracking processes. Furthermore, the function of monitoring customer data changes and sending update notifications via instant messaging ensures timely information delivery, greatly improving work efficiency and response speed, thus comprehensively addressing the shortcomings of existing systems in customer segmentation, data analysis visualization, opportunity tracking, and field support.

[0211] Figure 5 This is a schematic block diagram of a customer potential intelligent assessment system 300 based on dynamic weight adjustment provided in an embodiment of the present invention. Figure 5As shown, corresponding to the above-described intelligent customer potential assessment method based on dynamic weight adjustment, the present invention also provides an intelligent customer potential assessment system 300 based on dynamic weight adjustment. This intelligent customer potential assessment system 300 based on dynamic weight adjustment includes a unit for executing the above-described intelligent customer potential assessment method based on dynamic weight adjustment, and the system can be configured in a server. Specifically, please refer to... Figure 5 The customer potential intelligent assessment system 300 based on dynamic weight adjustment includes an acquisition unit 301, a level determination unit 302, an address processing unit 303, a prediction unit 304, and a monitoring unit 305.

[0212] The acquisition unit 301 is used to collect relevant data from internal systems and external APIs, and perform cleaning and standardization processing to obtain preprocessed results; the grade determination unit 302 is used to calculate the score of each customer based on the preprocessed results according to preset dimension indicators and weights to determine the customer grade; the address processing unit 303 is used to convert the customer address in the preprocessed results into coordinate form, and recommend nearby potential customers based on the user's current location; the prediction unit 304 is used to analyze historical data to build a conversion prediction model, and monitor and predict the risk of lost business opportunities in real time; the monitoring unit 305 is used to monitor changes in customer data in the preprocessed results and send update notifications through instant messaging tools according to predefined rules.

[0213] In one embodiment, the level determination unit 302 includes:

[0214] The weighting subunit is used to assign corresponding weights to each dimension according to preset dimension indicators; the summarization subunit is used to calculate the scores of each dimension based on the preprocessing results and the weights, and summarize them to obtain the total score; the analysis subunit is used to apply the nine-square grid analysis method based on the total score, and comprehensively consider the potential score and strategic importance to determine the customer level.

[0215] In one embodiment, the analysis subunit is used to classify customers into different levels in a nine-square grid based on the total score, define a nine-square grid coordinate system, and set corresponding interaction logic to determine the customer level, wherein the X-axis of the nine-square grid coordinate system represents the potential score; and the Y-axis represents the strategic importance.

[0216] In one embodiment, the address processing unit 303 is used to convert the customer address in the preprocessing result into coordinate form using geocoding technology and store it. When business development personnel use mobile devices, they can obtain the current location and recommend nearby potential customers by comparing distances.

[0217] In one embodiment, the prediction unit 304 includes:

[0218] The system comprises the following sub-units: Extraction Sub-unit, which extracts and cleans historical opportunity flow data from the enterprise's business system to obtain historical data; Classification Sub-unit, which identifies key stages of the opportunity funnel based on the historical data and classifies all opportunities accordingly; Tagging Sub-unit, which labels each opportunity event with positive and negative samples based on whether the opportunity was successfully converted or lost and the dwell time, to obtain a sample set; Statistics Sub-unit, which calculates and analyzes the conversion rate based on the sample set to count the number of successful conversions at each stage; Presentation Sub-unit, which uses a funnel chart to visually present the number of opportunities and their conversion status at each stage based on the number of successful conversions and the conversion rate; and Prediction Sub-unit, which combines the current status of opportunities with other influencing factors and uses machine learning algorithms to build a conversion prediction model to predict future opportunity trends and churn risks.

[0219] In one embodiment, the listening unit 305 is used to listen for customer data change events and define key triggering rules; when customer data changes, it sends an instant notification through Lark or other platforms.

[0220] It should be noted that those skilled in the art can clearly understand that the specific implementation process of the above-mentioned customer potential intelligent assessment system 300 based on dynamic weight adjustment and each unit can be referred to the corresponding description in the foregoing method embodiments. For the sake of convenience and brevity, it will not be repeated here.

[0221] The aforementioned intelligent customer potential assessment system 300 based on dynamic weight adjustment can be implemented as a computer program, which can, for example... Figure 6 It runs on the computer device shown.

[0222] Please see Figure 6 , Figure 6 This is a schematic block diagram of a computer device provided in an embodiment of this application. The computer device 500 can be a server, wherein the server can be a standalone server or a server cluster composed of multiple servers.

[0223] See Figure 6 The computer device 500 includes a processor 502, a memory, and a network interface 505 connected via a system bus 501. The memory may include a non-volatile storage medium 503 and internal memory 504.

[0224] The non-volatile storage medium 503 may store an operating system 5031 and a computer program 5032. The computer program 5032 includes program instructions that, when executed, cause the processor 502 to perform a customer potential intelligent assessment method based on dynamic weight adjustment.

[0225] The processor 502 provides computing and control capabilities to support the operation of the entire computer device 500.

[0226] The internal memory 504 provides an environment for the operation of the computer program 5032 in the non-volatile storage medium 503. When the computer program 5032 is executed by the processor 502, the processor 502 can execute a customer potential intelligent assessment method based on dynamic weight adjustment.

[0227] This network interface 505 is used for network communication with other devices. Those skilled in the art will understand that... Figure 6 The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the computer device 500 to which the present application is applied. The specific computer device 500 may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0228] The processor 502 is used to run a computer program 5032 stored in the memory to perform the following steps:

[0229] Relevant data is collected from internal systems and external APIs, and cleaned and standardized to obtain preprocessed results. Based on the preprocessed results, scores are calculated for each customer according to preset dimension indicators and weights to determine customer level. Customer addresses in the preprocessed results are converted into coordinate form, and potential customers nearby are recommended based on the user's current location. Historical data is analyzed to build a conversion prediction model, and the risk of lost business opportunities is monitored and predicted in real time. Changes in customer data in the preprocessed results are monitored, and update notifications are sent via instant messaging tools according to predefined rules.

[0230] The internal systems include CRM and financial systems; the external APIs include Qixinbao and Gaode Maps.

[0231] In one embodiment, when the processor 502 calculates the score of each customer based on the preprocessing result according to preset dimension indicators and weights to determine the customer level, it specifically implements the following steps:

[0232] Each dimension is assigned a corresponding weight according to the preset dimension indicators; the score of each dimension is calculated based on the preprocessing results and the weights, and the total score is obtained by summing them; the nine-square grid analysis method is applied based on the total score to comprehensively consider the potential score and strategic importance in order to determine the customer level.

[0233] In one embodiment, when implementing the step of applying the nine-square grid analysis method based on the total score to determine the customer level by comprehensively considering potential score and strategic importance, the processor 502 specifically implements the following steps:

[0234] Based on the total score, customers are divided into different levels in a nine-square grid. A nine-square grid coordinate system is defined, and corresponding interaction logic is set to determine the customer level. The X-axis of the nine-square grid coordinate system represents the potential score, and the Y-axis represents the strategic importance.

[0235] In one embodiment, when the processor 502 implements the step of converting the customer address in the preprocessing result into coordinate form and recommending nearby potential customers based on the user's current location, it specifically implements the following steps:

[0236] Geocoding technology is used to convert customer addresses in the preprocessed results into coordinates for storage. When business development personnel use mobile devices, they can obtain the current location and recommend nearby potential customers by comparing distances.

[0237] In one embodiment, when the processor 502 implements the step of analyzing historical data to build a conversion prediction model and monitor and predict the risk of lost business opportunities in real time, it specifically implements the following steps:

[0238] Historical opportunity flow data is extracted and cleaned from the enterprise's business system to obtain historical data. Based on this historical data, the key stages of the opportunity funnel are identified, and all opportunities are classified and managed accordingly. Based on whether an opportunity is successfully converted or lost and its dwell time, positive and negative samples are labeled for each opportunity event to obtain a sample set. Based on the sample set, the number of successful conversions of opportunities at each stage is counted, and the conversion rate is calculated and analyzed. Based on the number of successful conversions and the conversion rate, a funnel chart is used to visually present the number of opportunities and their conversion status at each stage. Combining the current status of opportunities and other influencing factors, a conversion prediction model is constructed using machine learning algorithms to predict future opportunity trends and churn risks.

[0239] In one embodiment, when the processor 502 implements the step of monitoring changes in customer data in the preprocessing result and sending an update notification via an instant messaging tool according to predefined rules, it specifically implements the following steps:

[0240] Listen for customer data change events and define key trigger rules; when customer data changes, send instant notifications via Lark or other platforms.

[0241] It should be understood that in the embodiments of this application, the processor 502 may be a central processing unit (CPU), or it may be other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor may be a microprocessor or any conventional processor.

[0242] It will be understood by those skilled in the art that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program includes program instructions and can be stored in a storage medium, which is a computer-readable storage medium. The program instructions are executed by at least one processor in the computer system to implement the process steps of the embodiments of the above methods.

[0243] Therefore, the present invention also provides a storage medium. This storage medium can be a computer-readable storage medium. The storage medium stores a computer program, wherein when executed by a processor, the computer program causes the processor to perform the following steps:

[0244] Relevant data is collected from internal systems and external APIs, and cleaned and standardized to obtain preprocessed results. Based on the preprocessed results, scores are calculated for each customer according to preset dimension indicators and weights to determine customer level. Customer addresses in the preprocessed results are converted into coordinate form, and potential customers nearby are recommended based on the user's current location. Historical data is analyzed to build a conversion prediction model, and the risk of lost business opportunities is monitored and predicted in real time. Changes in customer data in the preprocessed results are monitored, and update notifications are sent via instant messaging tools according to predefined rules.

[0245] The internal systems include CRM and financial systems; the external APIs include Qixinbao and Gaode Maps.

[0246] In one embodiment, when the processor executes the computer program to implement the step of calculating the score of each customer based on the preprocessing result according to preset dimension indicators and weights to determine the customer level, the processor specifically implements the following steps:

[0247] Each dimension is assigned a corresponding weight according to the preset dimension indicators; the score of each dimension is calculated based on the preprocessing results and the weights, and the total score is obtained by summing them; the nine-square grid analysis method is applied based on the total score to comprehensively consider the potential score and strategic importance in order to determine the customer level.

[0248] In one embodiment, when the processor executes the computer program to implement the step of applying the nine-box analysis method based on the total score to determine the customer level by comprehensively considering potential score and strategic importance, the processor specifically implements the following steps:

[0249] Based on the total score, customers are divided into different levels in a nine-square grid. A nine-square grid coordinate system is defined, and corresponding interaction logic is set to determine the customer level. The X-axis of the nine-square grid coordinate system represents the potential score, and the Y-axis represents the strategic importance.

[0250] In one embodiment, when the processor executes the computer program to convert the customer address in the preprocessed result into coordinate form and recommend nearby potential customers based on the user's current location, it specifically implements the following steps:

[0251] Geocoding technology is used to convert customer addresses in the preprocessed results into coordinates for storage. When business development personnel use mobile devices, they can obtain the current location and recommend nearby potential customers by comparing distances.

[0252] In one embodiment, when the processor executes the computer program to implement the step of analyzing historical data to build a conversion prediction model and monitor and predict the risk of lost business opportunities in real time, it specifically implements the following steps:

[0253] Historical opportunity flow data is extracted and cleaned from the enterprise's business system to obtain historical data. Based on this historical data, the key stages of the opportunity funnel are identified, and all opportunities are classified and managed accordingly. Based on whether an opportunity is successfully converted or lost and its dwell time, positive and negative samples are labeled for each opportunity event to obtain a sample set. Based on the sample set, the number of successful conversions of opportunities at each stage is counted, and the conversion rate is calculated and analyzed. Based on the number of successful conversions and the conversion rate, a funnel chart is used to visually present the number of opportunities and their conversion status at each stage. Combining the current status of opportunities and other influencing factors, a conversion prediction model is constructed using machine learning algorithms to predict future opportunity trends and churn risks.

[0254] In one embodiment, when the processor executes the computer program to monitor changes in customer data in the preprocessing result and sends an update notification via an instant messaging tool according to predefined rules, the processor specifically implements the following steps:

[0255] Listen for customer data change events and define key trigger rules; when customer data changes, send instant notifications via Lark or other platforms.

[0256] The storage medium can be any computer-readable storage medium capable of storing program code, such as a USB flash drive, portable hard drive, read-only memory (ROM), magnetic disk, or optical disk.

[0257] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of both. To clearly illustrate the interchangeability of hardware and software, the components and steps of the various examples have been generally described in terms of functionality in the foregoing description. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementations should not be considered beyond the scope of this invention.

[0258] In the embodiments provided by this invention, it should be understood that the disclosed systems and methods can be implemented in other ways. For example, the system embodiments described above are merely illustrative. For example, the division of each unit is only a logical functional division, and there may be other division methods in actual implementation. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed.

[0259] The steps in the method of this invention can be adjusted, merged, or reduced in order according to actual needs. The units in the system of this invention can be merged, divided, or reduced according to actual needs. Furthermore, the functional units in the various embodiments of this invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit.

[0260] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a storage medium. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, a terminal, or a network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention.

[0261] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any person skilled in the art can easily conceive of various equivalent modifications or substitutions within the technical scope disclosed in the present invention, and these modifications or substitutions should all be covered within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A method for intelligent assessment of customer potential based on dynamic weight adjustment, characterized in that, The method comprises the following steps: Collecting relevant data from internal systems and external APIs, and performing cleaning and standardization processing to obtain pre-processing results; Based on the pre-processing results, the score of each customer is calculated according to the preset dimension index and weight to determine the customer level; Convert the customer address in the pre-processing results into coordinate form, and recommend potential customers nearby based on the current location of the user; Analyze historical data to build a conversion prediction model, and monitor and predict the risk of business opportunity loss in real time; Listen to the changes of customer data in the pre-processing results, and send update notifications through instant messaging tools according to predefined rules.

2. The method for intelligent evaluation of customer potential based on dynamic weight adjustment according to claim 1, characterized in that, The internal system includes CRM, financial system; The external API includes Qixinbao and Gaode map.

3. The method for intelligent evaluation of customer potential based on dynamic weight adjustment according to claim 2, characterized in that, The method comprises the following steps: According to the preset dimension index, each dimension is given a corresponding weight; Based on the pre-processing results, the score of each dimension is calculated based on the weight, and the total score is obtained by summarizing; Based on the total score, the nine-square analysis method is applied, and the potential score and strategic importance are comprehensively considered to determine the customer level.

4. The method for intelligent evaluation of customer potential based on dynamic weight adjustment according to claim 3, characterized in that, The method comprises the following steps: Based on the total score, the customer is divided into different levels in the nine-square grid, the nine-square coordinate system is defined, and the corresponding interaction logic is set to determine the customer level, wherein the X-axis of the nine-square coordinate system represents the potential score; The Y-axis represents the strategic importance.

5. The method for intelligent evaluation of customer potential based on dynamic weight adjustment according to claim 1, characterized in that, The method comprises the following steps: Use geocoding technology to convert the customer address in the pre-processing results into coordinate form for storage. When the BD personnel use a mobile device, the current location is obtained and the nearby potential customers are recommended by comparing the distance.

6. The method for intelligent evaluation of customer potential based on dynamic weight adjustment according to claim 1, characterized in that, The method comprises the following steps: Extract and clean historical business opportunity flow data from enterprise business systems to obtain historical data; Based on the historical data, the key stages of the business opportunity funnel are determined, and all business opportunities are classified and managed accordingly; Based on whether the business opportunity is successfully converted or lost and the residence time, mark positive and negative samples for each business opportunity event to obtain a sample set; Based on the sample set, the number of successful conversions of each stage of business opportunity is calculated and analyzed to calculate and analyze the conversion rate; Based on the number of successful conversions and the conversion rate, use the funnel chart to intuitively present the number of business opportunities at each stage and their conversion situation; Combine the current state of the business opportunity and other influencing factors, and use machine learning algorithms to build a conversion prediction model to predict future business opportunity trends and loss risks.

7. The method for intelligent evaluation of customer potential based on dynamic weight adjustment according to claim 1, characterized in that, The method comprises the following steps: Listen to the customer data change event, define the key trigger rule; When the customer data changes, send instant notifications through Feishu or other platforms.

8. A system for intelligent assessment of customer potential based on dynamic weight adjustment, characterized by, The method comprises the following steps: An acquisition unit is configured to collect relevant data from internal systems and external APIs, and perform cleaning and standardization processing to obtain pre-processing results; The grade determining unit is configured to calculate scores of the customers according to preset dimension indexes and weights based on the preprocessing result, so as to determine customer grades. The address processing unit is configured to convert customer addresses in the preprocessing result into coordinate forms, and recommend nearby potential customers based on a current location of a user. The prediction unit is configured to analyze historical data to construct a conversion prediction model, and monitor and predict a business opportunity loss risk in real time. The monitoring unit is configured to monitor changes in customer data in the preprocessing result, and send an update notification through an instant messaging tool according to a predefined rule.

9. The system for intelligent assessment of customer potential based on dynamic weight adjustment as claimed in claim 8 wherein, The grade determining unit includes: The weight assignment subunit is configured to assign corresponding weights to each dimension according to preset dimension indexes. The summary subunit is configured to calculate dimension scores based on the preprocessing result and the weights, and summarize the total scores. The analysis subunit is configured to apply a nine-square grid analysis method based on the total scores, comprehensively consider potential scores and strategic importance, and determine customer grades.

10. The system for intelligent assessment of customer potential based on dynamic weight adjustment as claimed in claim 8 wherein, The analysis subunit is configured to divide the customers into different grades in the nine-square grid based on the total scores, define a nine-square grid coordinate system, and set corresponding interaction logics, so as to determine customer grades. The X-axis of the nine-square grid coordinate system represents potential scores, and the Y-axis represents strategic importance.

Images