Marketing personnel capability assessment method and system

Abstract

The invention discloses a marketing personnel capability assessment method and system, and relates to the technical field of computers. According to the method, after target data, namely predetermined data related to marketing personnel capability assessment, is obtained, assessment features for each marketing personnel are determined based on the target data, then the assessment features are input into a capability assessment model, and capability assessment results for each marketing personnel are obtained, the ability evaluation model is a two-stage model which is obtained based on historical target data training and is composed of an XGBoost model and a deep neural network model, the XGBoost model is used for carrying out importance sorting on the evaluation features, and the deep neural network model is used for evaluating the ability of each marketing personnel based on the sorted evaluation features. The problems that the ability evaluation of the marketing personnel based on a single index and the like is not scientific enough and the comprehensive ability level of the marketing personnel cannot be reflected truly and objectively in the prior art are solved.

Description

Technical Field

[0001] The present application relates to the field of computer technology, and more specifically, to a method and system for evaluating marketing personnel capabilities. Background Art

[0002] Current evaluation methods for marketers are overly simplistic. For example, they rely on static rankings based on a single metric (such as sales), subjective manual evaluations (using a supervisor's scoring system), and periodic paper-based exams to assess knowledge mastery. Without a scientific evaluation system, these methods fail to truly and objectively reflect a marketer's overall capabilities. Summary of the Invention

[0003] In view of this, the embodiments of the present application are committed to providing a marketing personnel capability assessment method and system to solve the problem that the existing marketing personnel capability assessment based on a single indicator is not scientific enough and cannot truly and objectively reflect the comprehensive capability level of the marketing personnel.

[0004] In a first aspect, the present invention provides a method for evaluating marketing personnel capabilities, comprising:

[0005] Acquiring target data, wherein the target data is predetermined data related to the marketing personnel capability assessment;

[0006] Determining evaluation characteristics for each marketing personnel based on the target data;

[0007] Inputting the evaluation characteristics into a capability evaluation model to obtain capability evaluation results for each of the marketing personnel;

[0008] The capability assessment model is a two-stage model consisting of an XGBoost model and a deep neural network model obtained through training based on historical target data; wherein, the XGBoost model is used to rank the importance of the assessment features, and the deep neural network model is used to assess the capability of each of the marketing personnel based on the ranked assessment features.

[0009] In one possible implementation, determining the evaluation characteristics for each marketing personnel based on the target data includes:

[0010] Sorting the target data in time sequence to obtain time sequence target data;

[0011] Analyzing the latest time series target data within a first preset time period based on an observation window that slides at a preset frequency to obtain a first eigenvalue;

[0012] Calculating an arithmetic mean of the first eigenvalue and a second eigenvalue, where the second eigenvalue is a eigenvalue obtained by analyzing the time series target data within a latest second preset time period based on the observation window; the second preset time period is greater than the first preset time period;

[0013] The arithmetic mean is used as the evaluation feature.

[0014] In a possible implementation, inputting the evaluation characteristics into a capability evaluation model to obtain capability evaluation results for each marketing personnel includes:

[0015] Inputting the evaluation features into the XGBoost model, and having the XGBoost model sort the importance of the evaluation features to obtain the sorted evaluation features;

[0016] The sorted evaluation features are input into the deep neural network model, and the deep neural network model performs capability evaluation on the marketing personnel based on the mapping relationship to obtain the capability evaluation result; the mapping relationship is the correspondence between the evaluation features and the capability dimension that characterizes the capability of the marketing personnel.

[0017] In one possible implementation, the XGBoost model ranks the evaluation features by importance, including:

[0018] Determining whether the evaluation feature satisfies any preset condition, the preset condition including: being greater than the mean of the evaluation feature and having a fluctuation less than a preset fluctuation range;

[0019] If satisfied, the weight of the assessment feature in the capability dimension is increased compared to other assessment features;

[0020] If not, the weight of the evaluation feature in the capability dimension is reduced compared to other evaluation features.

[0021] In one possible implementation, the XGBoost model ranks the evaluation features by importance, including:

[0022] determining whether the assessment feature has a preset required weight;

[0023] If yes, the weight value of the assessment feature within the capability dimension is determined according to the preset required weight.

[0024] In a possible implementation, the deep neural network model performs capability assessment on the marketing personnel based on the mapping relationship, including:

[0025] determining a combination of features present in the evaluation features;

[0026] Based on the evaluation characteristics and / or the combination of characteristics, the marketing personnel are evaluated for their capabilities.

[0027] In one possible embodiment, the capability assessment results include a real-time capability ranking, a multi-dimensional capability radar chart generated based on the scores corresponding to the multiple capability dimensions, and the score gap between the current marketer and the benchmark marketer on the capability dimensions, where the benchmark marketer is a marketer whose score belongs to the top preset percentage in the industry.

[0028] In a possible implementation, the capability assessment model is dynamically updated; a condition for the capability assessment model to be dynamically updated is that the number of new orders placed by a timing or marketing personnel reaches a preset threshold.

[0029] In a possible implementation, obtaining target data includes:

[0030] Obtain preselected data at a preset frequency;

[0031] The preselected data is cleaned to obtain the target data.

[0032] In a second aspect, the present invention provides a marketing personnel capability assessment system, comprising:

[0033] A collection unit, configured to obtain target data, wherein the target data is predetermined data related to the marketing personnel capability assessment;

[0034] A first processing unit, configured to determine evaluation characteristics for each marketing personnel based on the target data;

[0035] A second processing unit is configured to input the evaluation characteristics into a capability evaluation model to obtain capability evaluation results for each of the marketing personnel;

[0036] The capability assessment model is a two-stage model consisting of an XGBoost model and a deep neural network model obtained through training based on historical target data; wherein, the XGBoost model is used to rank the importance of the assessment features, and the deep neural network model is used to assess the capability of each of the marketing personnel based on the ranked assessment features.

[0037] According to the marketing personnel capability assessment method provided by the present invention, after first obtaining target data (i.e., pre-determined data related to the marketing personnel capability assessment), assessment features for each marketing personnel are determined based on the target data. Then, the assessment features are input into a capability assessment model to obtain capability assessment results for each marketing personnel. The capability assessment model is a two-stage model composed of an XGBoost model and a deep neural network model trained based on historical target data. The XGBoost model is used to rank the assessment features by importance, and the deep neural network model is used to assess the capability of each marketing personnel based on the ranked assessment features. That is, by inputting the assessment features for each marketing personnel determined based on the target data into the two-stage model, a multi-dimensional evaluation of the marketing personnel is achieved, thereby improving the accuracy and scientific nature of the evaluation. BRIEF DESCRIPTION OF THE DRAWINGS

[0038] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are merely embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on the provided drawings without any creative work.

[0039] Figure 1 Shown is a flow chart of a marketing personnel capability assessment method provided by an embodiment of the present invention;

[0040] Figure 2 The figure shows an example of a segmentation chart for marketing personnel capability dimensions provided by an embodiment of the present invention;

[0041] Figure 3 FIG. 1 shows an example of an evaluation result display diagram provided by an embodiment of the present invention;

[0042] Figure 4 An example process diagram for implementing the marketing personnel capability assessment method provided by an embodiment of the present invention;

[0043] Figure 5 FIG2 is a structural diagram of a marketing personnel capability evaluation system provided by an embodiment of the present invention;

[0044] Figure 6 FIG2 is a schematic structural diagram of an electronic device provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0045] Unless otherwise defined, technical or scientific terms used in the embodiments of this specification should have the same ordinary meaning as those understood by persons of ordinary skill in the art to which this specification pertains. The terms "first," "second," and similar terms used in the embodiments of this specification do not denote any order, quantity, or importance, but are provided solely to avoid confusion between constituent elements.

[0046] Unless the context requires otherwise, throughout this specification, the term "plurality" means "at least two," and "including" is to be interpreted as open and inclusive, meaning "including, but not limited to." Throughout this specification, the terms "one embodiment," "some embodiments," "exemplary embodiments," "example," "specific example," or "some examples" are intended to indicate that a particular feature, structure, material, or characteristic associated with the embodiment or example is included in at least one embodiment or example of this specification. The schematic representations of these terms do not necessarily refer to the same embodiment or example.

[0047] The following will be combined with the drawings in the embodiments of this specification to clearly and completely describe the technical solutions in the embodiments of this specification. Obviously, the embodiments described are only part of the embodiments of this specification, not all of the embodiments. Based on the embodiments in this specification, all other embodiments obtained by ordinary technicians in this field without making any creative efforts are within the scope of protection of this specification.

[0048] Understandably, the current methods commonly used to assess marketer capabilities all have certain flaws. For example, static rankings based on a single metric, such as sales, fail to consider other factors like customer feedback and number of visits, resulting in biased assessments. Manual evaluations by superiors, based solely on their subjective opinions, can lead to unscientific and inaccurate results. Knowledge assessments based on periodic exams also lack scientific validity because they fail to consider marketers' actual performance.

[0049] Based on this, an embodiment of the present invention provides a method for evaluating the capabilities of marketing personnel by inputting evaluation features determined based on predetermined data related to the evaluation of marketing personnel capabilities into a two-stage model composed of an XGBoost (Extreme Gradient Boosting) model and a deep neural network model, thereby improving the accuracy and scientificity of the evaluation.

[0050] The marketing personnel capability assessment method provided by the present invention is executed on an electronic device, which can be a smart terminal device such as a laptop, a personal computer, and a tablet computer connected to an enterprise CRM (customer relationship management system), an OLS (online school) system, etc. Based on the above content, see Figure 1 , Figure 1: is a flow chart of a marketing personnel capability assessment method provided by an embodiment of the present invention. The process of the marketing personnel capability assessment method provided by this embodiment may include:

[0051] S100: Acquire target data.

[0052] The target data is predetermined data related to the assessment of marketing personnel's capabilities.

[0053] Specifically, target data may include order data, visit data, customer complaint data, learning data, etc.

[0054] More specifically, order data may include: marketing personnel number, customer number, order status (new, approved, canceled, invalidated), number of devices, order amount, order creation time, affiliated business unit, affiliated agent, etc.; visit data may include: marketing personnel number, customer number, visit form status (new, approved, canceled, invalidated), actual visit time, visit form creation time, etc.; customer complaint data may include: marketing personnel number, customer number, complaint form status (new, approved, canceled, invalidated), complaint form creation time, complaint response time, satisfaction score, etc.; and learning data may include: marketing personnel number, learning task number, learning task status (not started, in progress, passed, failed), planned start time, planned completion time, actual completion time, learning task test score, etc.

[0055] It should be noted that, in order to facilitate understanding of the ability level of each marketing personnel, the target data may also include marketing personnel data. Further, the marketing personnel data may include: marketing personnel number, marketing personnel name, affiliated business unit, affiliated agency, etc.

[0056] Furthermore, target data can be obtained through enterprise CRM, OLS systems, etc.

[0057] S110. Determine evaluation characteristics for each marketing personnel based on the target data.

[0058] It's understandable that the number of completed orders, the total value of completed orders, the number of customer visits, the number of complaints received, and academic performance are all related to a marketer's abilities. Based on this, evaluation features can be defined. Once the target data is obtained, these evaluation features can then be further determined based on the target data.

[0059] In some possible embodiments, the evaluation characteristics may be unified into the dimensions of the business unit, the agency, and the marketing personnel, and may include:

[0060] Cumulative number of orders completed by marketing personnel, cumulative amount of orders completed by marketing personnel, cumulative number of customers completed by marketing personnel, average number of orders completed by marketing personnel (cumulative number of orders completed / cumulative number of customers completed), average unit price per customer completed by marketing personnel (cumulative amount of orders completed / cumulative number of customers completed), cumulative number of visits by marketing personnel, cumulative number of customers visited by marketing personnel, average number of visits by marketing personnel (cumulative number of visits / number of days in the observation window), average number of customer visits by marketing personnel (cumulative number of visits / cumulative number of customers visited), cumulative number of complaints received by marketing personnel, cumulative number of customers complained about by marketing personnel, average number of complaints received by marketing personnel (cumulative number of complaints received / number of days in the observation window), average number of customer complaints received by marketing personnel (cumulative number of complaints received / cumulative number of customers complained about ), cumulative response time of marketing personnel (sum of complaint ticket response time minus complaint ticket creation time), average complaint response time of marketing personnel (cumulative response time / cumulative number of complaints), cumulative satisfaction score of marketing personnel, average complaint satisfaction score of marketing personnel (cumulative satisfaction score / cumulative number of complaints), cumulative number of learning tasks of marketing personnel, number of tasks completed on time by marketing personnel, rate of task completion on time by marketing personnel (number of tasks completed on time / cumulative number of learning tasks), cumulative learning performance of marketing personnel, average learning performance of marketing personnel (cumulative learning performance / cumulative number of learning tasks), cumulative learning completion time of marketing personnel, and average learning completion time of marketing personnel (cumulative learning completion time / cumulative number of learning tasks) (excluding unfinished learning tasks).

[0061] Furthermore, in an optional embodiment, based on the target data, determining evaluation characteristics for each marketing personnel includes:

[0062] Sort the target data in time series to obtain time series target data;

[0063] Analyze the latest time series target data within a first preset time period based on an observation window that slides at a preset frequency to obtain a first eigenvalue;

[0064] Compute the arithmetic mean of the first and second eigenvalues.

[0065] The second characteristic value is a characteristic value obtained by analyzing the latest time series target data within a second preset time length based on the observation window; the second preset time length is greater than the first preset time length.

[0066] The arithmetic mean is used as the evaluation feature.

[0067] Specifically, after sorting the target data in time series to obtain time series target data, assuming the window size is set to 30 days, the sliding interval (i.e., frequency) is set to 1 day, and a 60-day moving window is used to observe daily changes, when using the method provided in the above embodiment to determine the evaluation characteristics for each marketer, the observation window can be fixed to analyze the target data of the most recent 30 days each time, and then the first characteristic value within the window can be calculated each time the observation window moves, for example: the cumulative number of completed orders within 30 days, the average number of completed orders per customer within 30 days, etc. The evaluation characteristics are then obtained by calculating the arithmetic mean of the second characteristic values ​​of the most recent 60 days, i.e., the characteristic values ​​obtained each time the observation window moves.

[0068] More specifically, in some possible embodiments, taking the calculation of the evaluation feature of a marketing personnel's cumulative number of completed orders as an example, the cumulative number of completed orders for the marketing personnel in the past 30 days will be obtained daily over a rolling 60-day period and the average value will be calculated and returned. In actual application, the program logic may be:

[0069] #pseudocode

[0070] Input: Sales order data for the last 90 days [data[1~90]]

[0071] process:

[0072] 1. Set variable I = 0

[0073] 2. While I is less than 60:

[0074] 3. Get the number of sales orders in the last 30 days: data[60-I~90-I]

[0075] 4. Calculate the number of orders in data[60-I~90-I] based on the dimensions of the business unit, agent, and marketing representative, and get the result R[60-I]

[0076] 5. Add R[60-I] to the set ARR{} to obtain ARR{R[60-I]}

[0077] 6. I=I+1

[0078] 7. Until I = 60, get ARR {1 ~ 60}

[0079] 8. Traverse the 60 data in ARR{1~60} and calculate the average to get the final result

[0080] Furthermore, obtaining the capability assessment results for each marketing personnel through the capability assessment model may, in an optional embodiment, include:

[0081] The evaluation features are input into the XGBoost model, and the XGBoost model ranks the evaluation features by importance to obtain the ranked evaluation features;

[0082] The sorted evaluation features are input into the deep neural network model, and the deep neural network model evaluates the capabilities of the marketers based on the mapping relationship to obtain the capability evaluation results.

[0083] The mapping relationship is the correspondence between the evaluation characteristics and the capability dimensions that characterize the capabilities of marketers.

[0084] Specifically, the mapping relationship is pre-established. In some possible embodiments, the corresponding relationship between the evaluation characteristics and the capability dimensions representing the capabilities of the marketing personnel can be established as shown in Table 1 below:

[0085] Table 1. Correlation mapping table between assessment characteristics and capability dimensions

[0086] Derived Features Ability dimension attribution Cumulative number of completed orders Business capabilities Average number of customer orders Business capabilities Average transaction price per customer Business capabilities Average number of visits Business capabilities Average number of customer visits Business capabilities Average number of complaints Customer relationship capabilities Average number of customer complaints Customer relationship capabilities Average complaint response time Customer relationship capabilities Average complaint satisfaction score Customer relationship capabilities On-time completion rate Professional capabilities Average academic performance Professional capabilities Average time to complete study Professional capabilities

[0087] In this embodiment, by ranking the evaluation features by importance based on the XGBoost model before evaluating the capabilities of marketing personnel based on the mapping relationship based on the deep neural network model, the capability evaluation of marketing personnel is made more scientific, thereby improving the accuracy of the evaluation results.

[0088] In an optional embodiment, the XGBoost model ranks the evaluation features by importance, including:

[0089] Determine whether the evaluation characteristic meets any preset conditions, including: being greater than the mean of the evaluation characteristic and having a fluctuation less than a preset fluctuation range;

[0090] If satisfied, the weight of the assessment feature in the capability dimension is increased compared to other assessment features;

[0091] If not, the weight of the assessment feature in the capability dimension will be reduced compared to other assessment features.

[0092] It is understandable that if a certain evaluation characteristic of a marketer, for example, the cumulative number of completed orders is more than 50% higher than the average, it means that the marketer is significantly better than other marketers in terms of order volume. In other words, this evaluation characteristic can clearly distinguish between excellent and ordinary employees. Increasing the weight of this evaluation characteristic in the business capability dimension compared to other evaluation characteristics will make it easier to distinguish the ability differences between different marketers in the business capability dimension.

[0093] Furthermore, if the value of a certain evaluation feature has been low before and suddenly becomes very high, such as the average transaction unit price, then the fluctuation of this evaluation feature is too large. At this time, reducing the weight of this evaluation feature in the business capability dimension compared to other evaluation features will avoid the impact of the evaluation feature with large fluctuations on the accuracy of the evaluation results.

[0094] It should be noted that different users may attach different importance to different dimensions of marketers' capabilities. For example, some users pay more attention to marketers' business capabilities, while some users pay more attention to marketers' professional capabilities.

[0095] Based on this, in an optional embodiment, the XGBoost model ranks the evaluation features by importance, including:

[0096] Determine whether the assessment characteristics have the preset required weights;

[0097] If so, the weight value of the assessment characteristic within the capability dimension is determined according to the preset required weight.

[0098] In this embodiment, when an evaluation feature has a preset required weight, the weight value of the evaluation feature in the capability dimension is determined according to the preset required weight, so that the user can flexibly set the weight of each evaluation feature according to his or her own needs, thereby obtaining an evaluation result that better meets his or her own requirements.

[0099] In some possible embodiments, users can set their own minimum values. For example, when calculating the weight of a marketing staff's professional ability, it is manually set that the "average learning performance" must account for 50% of the professional ability weight. Then, the "on-time task completion rate" and "average learning completion time" account for 50% in total. The program logic for dynamically adjusting the weights through the XGBoost model can be:

[0100]

[0101]

[0102] weights=calculate_weights(self_data,peer_data)

[0103] print(f"""

[0104] Final weight distribution:

[0105] Average study score: {weights['avg_study_score']: .1%}

[0106] On-time completion rate: {weights['task_completion_rate']: .1%}

[0107] Learning duration: {weights['avg_duration']: .1%}

[0108] """)

[0109] In an optional embodiment, the deep neural network model evaluates the capabilities of marketing personnel based on the mapping relationship, including:

[0110] Determine the combination of features present in the assessment characteristics;

[0111] Conduct a competency assessment of marketing personnel based on assessment characteristics and / or combinations of characteristics.

[0112] Specifically, before evaluating marketing personnel's capabilities based on a deep neural network (DNN) model, the DNN model should be trained first. The specific training process mainly includes the following steps:

[0113] 1. Data cleaning: Clean the evaluation features into a standard format.

[0114] For example, all values ​​and percentages are converted to a scale of 0-100. For the cumulative number of completed orders, average number of completed customer orders, average average transaction price per customer, average number of visits, average number of customer visits, average number of complaints received, average number of customer complaints received, average complaint response time, and average learning completion time, scores are calculated linearly based on the data of other employees during the same period. For the average complaint satisfaction score and average learning performance, since the results are within the range of 0-100, the results are used directly. For the on-time task completion rate, since the results are within the range of 0-1, the results are directly multiplied by 100.

[0115] 2. Layered learning:

[0116] First level: Analyze feature combinations, for example: high cumulative number of completed orders + low average number of customer complaints.

[0117] The second layer: output scores for different capability dimensions, such as business capability, customer relationship capability, and professional capability with a score of 0-100.

[0118] 3. Model training: Use historical target data to repeatedly adjust parameters until you can accurately distinguish between known excellent and average employees.

[0119] For example, first input all target data from the past three years and the currently calculated evaluation features and weight ratios, and then the model training calculation outputs results similar to the following:

[0120] Business ability: 85 points (10 points higher than the team average);

[0121] Customer relationship ability: 70 points (15 points lower than the team average);

[0122] Professional ability: 91 points (5 points higher than the team average).

[0123] The specific program logic is as follows:

[0124] import numpy as np

[0125] import pandas as pd

[0126] from sklearn.preprocessing import MinMaxScaler

[0127] from tensorflow.keras.models import Model

[0128] from tensorflow.keras.layers import Dense,Input

[0129]

[0130]

[0131] # Initialize the model

[0132] input_dim=12 #12 features

[0133] ability_model=EmployeeAbilityModel(input_dim)

[0134] #Simulate historical training data (actually real data should be used)

[0135] #Assume there are 1000 historical samples

[0136] np.random.seed(42)

[0137] X_train = np.random.rand(1000,12)*100#Simulate the standardized data

[0138] y_train = np.random.rand(1000,3)*100#Simulate the scores of three ability dimensions

[0139] #Model training

[0140] print("Start training deep neural network...")

[0141] ability_model.train(X_train,y_train,epochs=200)

[0142] #Current employee data processing

[0143] current_employee={

[0144] #...(same as original input data)

[0145] }

[0146] #Data preprocessing

[0147] processed_data=ability_model.preprocess_data(current_data)

[0148] #Prediction ability score

[0149] ability_scores=ability_model.predict(processed_data)[0]

[0150] #Generate result description

[0151] team_averages = {'Business Capability': 75, 'Customer Relationship Capability': 85, 'Professional Capability': 86} dimension_names = ['Business Capability', 'Customer Relationship Capability', 'Professional Capability'] result = {}

[0152] for i,dimin enumerate(dimension_names):

[0153] score=int(round(ability_scores[i]))

[0154] diff=score-team_averages[dim]

[0155] if diff>0:

[0156] comment=f"Higher than the team average {diff} score"

[0157]

[0158] In this embodiment, by determining the feature combinations present in the evaluation features and then performing capability evaluation on the marketing personnel based on the evaluation features and / or feature combinations, the accuracy and scientificity of the evaluation are further improved.

[0159] It is understandable that by reasonably setting the content of the evaluation results displayed to marketers and company managers, it can help company managers accurately identify polar marketers, help marketers see their own shortcomings in ability, and by comparing with other marketers, find their own gaps, set benchmarks, promote improvements, and improve the overall capabilities of the company's marketing team.

[0160] Based on this, in an optional embodiment, the capability assessment results include real-time capability rankings, a multi-dimensional capability radar chart generated based on scores corresponding to multiple capability dimensions, and the score gap between the current marketer and the benchmark marketer in the capability dimensions.

[0161] Among them, benchmark marketers are marketers whose scores belong to the top preset percentage in the industry.

[0162] Specifically, the evaluation results presented to company managers and marketing personnel can be based on Figure 2 The three dimensions shown are:

[0163] The first dimension: business capabilities (cumulative number of completed orders, average number of completed orders per customer, average transaction price per customer, average number of visits, average number of customer visits);

[0164] The second dimension: professional ability (rate of completing tasks on time, average academic performance, average time to complete learning);

[0165] The third dimension: customer relationship capabilities (average number of complaints received, average number of customer complaints received, average complaint response market, and average complaint satisfaction score).

[0166] Then, based on the above three dimensions, the evaluation results are displayed as follows: Figure 3 As shown, the scores marked in the outer circle are the ability dimensions of benchmark marketers, and the scores marked in the inner circle are the ability dimensions of the marketer currently being evaluated. It can be seen that the evaluation results include both a multi-dimensional ability radar chart for individual capabilities and the score gap with the ability dimensions of benchmark marketers.

[0167] Furthermore, textual displays can be used to help marketers better understand their own capabilities, such as displaying "Your professional capabilities are 23% below the benchmark value." Furthermore, when marketers' capabilities fluctuate significantly, alerts can be sent to marketers and company managers, facilitating self-improvement for marketers and management for company managers.

[0168] It is understandable that ensuring the accuracy of the capability model directly affects the accuracy of the assessment results. Based on this, in an optional embodiment, the capability assessment model is dynamically updated; and the conditions for dynamic updating of the capability assessment model are: the number of new orders received by the marketing personnel reaches a preset threshold or the timing.

[0169] In this embodiment, the capability assessment model is updated by setting a timing or when the number of new orders from marketing personnel reaches a preset threshold. For example, the capability assessment model is updated at 0:00 a.m. every day or when the number of new orders from marketing personnel increases by 10. This ensures the accuracy of the capability assessment model and thus the accuracy of the capability assessment results of marketing personnel.

[0170] It is understandable that by reasonably setting the target data acquisition frequency, real-time identification of high-performing personnel / those who need improvement can be achieved.

[0171] Based on this, in an optional embodiment, obtaining target data includes:

[0172] Obtain preselected data at a preset frequency;

[0173] Perform data cleaning on the pre-selected data to obtain the target data.

[0174] Specifically, by performing data cleaning on the pre-selected data, the accuracy and authenticity of the data can be guaranteed, thereby improving the accuracy of the evaluation results.

[0175] In some possible embodiments, data cleansing for order data may include:

[0176] Data filtering: Eliminate documents with order status of New, Canceled, and Voided, and exclude data with empty customer numbers or marketing staff numbers;

[0177] Outlier processing: If the number of devices or order amount is negative, the absolute value is taken.

[0178] Data cleaning for visit data may include:

[0179] Data filtering: Eliminate documents with visit status of New, Canceled, and Voided, and exclude data with empty customer numbers or marketing staff numbers;

[0180] Abnormal value processing: If the actual visit time is empty or not in date format, it is replaced with the visit order creation time;

[0181] Data cleansing for customer complaint data may include:

[0182] Data filtering: Eliminate documents with the complaint status of New, Canceled, and Voided, and exclude data with empty customer numbers or marketing staff numbers;

[0183] Outlier processing: Data with complaint response time that is empty or not in date format is replaced with the average complaint response time of the marketing representative.

[0184] Data cleaning for learning data may include:

[0185] Data filtering: Eliminate data with empty marketing personnel numbers;

[0186] Outlier processing: All data with empty or less than 0 learning task test scores are replaced with 0.

[0187] Data cleansing for marketer data may include:

[0188] Data filtering: Eliminate data with empty marketing personnel numbers;

[0189] Outlier processing: If the data of the business unit or agent is empty, the data of the business unit and agent in the latest sales order of the associated marketing personnel will be used to supplement it.

[0190] The following is an example to illustrate how to implement the marketing personnel capability assessment method provided in the above embodiment. Figure 4 As shown, it includes the following stages:

[0191] 1. Data preparation stage:

[0192] By configuring the API interface, data on orders, visits, customer complaints, learning, etc. in the past three years can be collected.

[0193] 2. Data processing stage:

[0194] Data cleaning: Perform data cleaning according to data cleaning rules (data filtering, outlier processing, etc.) to obtain "clean" data that can reflect the actual situation.

[0195] Evaluate feature calculations.

[0196] 3. Model training phase:

[0197] Initialize the baseline model using historical data.

[0198] Set the dynamic update trigger conditions (executed at 0:30 am every day or when the number of marketing orders increases by 10).

[0199] 4. Application stage:

[0200] Marketers view real-time capability rankings and three-dimensional capability radar charts on the data visualization platform every day.

[0201] Managers receive early warning reports from polar personnel.

[0202] As can be seen, the above embodiment provides a dynamic, multi-dimensional marketing personnel capability assessment method. By integrating multi-source heterogeneous data to construct a capability assessment model, it enables real-time identification of high-performing and improvement-needed personnel, generates individual capability maps encompassing multiple core capability dimensions, and provides quantitative gap analysis based on industry benchmarks. This helps company managers accurately identify marketers at both ends of the spectrum, helping marketers clearly understand their own weaknesses. By comparing themselves with other marketers, they can identify their own gaps, set benchmarks, and promote improvement, ultimately enhancing the overall capabilities of the company's marketing team.

[0203] The following is an introduction to the marketing personnel capability assessment system provided in an embodiment of the present invention. The marketing personnel capability assessment system described below can be considered as a module architecture for implementing the marketing personnel capability assessment method provided in an embodiment of the present invention; the following description can be cross-referenced with the above.

[0204] Optional, see Figure 5 , Figure 5 This is a structural block diagram of a marketing personnel capability evaluation system provided by an embodiment of the present invention. The system may include:

[0205] The acquisition unit 10 is used to obtain target data, which is predetermined data related to the marketing personnel capability assessment;

[0206] A first processing unit 20 is configured to determine evaluation characteristics for each marketing personnel based on the target data;

[0207] The second processing unit 30 is used to input the evaluation characteristics into the capability evaluation model to obtain the capability evaluation results for each marketing personnel;

[0208] The capability assessment model is a two-stage model consisting of an XGBoost model and a deep neural network model trained based on historical target data. The XGBoost model is used to rank the importance of assessment features, and the deep neural network model is used to evaluate the capabilities of each marketer based on the ranked assessment features.

[0209] Optionally, the first processing unit 20 is specifically configured to:

[0210] Sort the target data in time series to obtain time series target data;

[0211] Analyze the latest time series target data within a first preset time period based on an observation window that slides at a preset frequency to obtain a first eigenvalue;

[0212] Calculating the arithmetic mean of the first eigenvalue and the second eigenvalue, where the second eigenvalue is a eigenvalue obtained by analyzing the latest time series target data within a second preset time period based on the observation window; the second preset time period is greater than the first preset time period;

[0213] The arithmetic mean is used as the evaluation feature.

[0214] Optionally, the second processing unit 30 is specifically configured to:

[0215] The evaluation features are input into the XGBoost model, and the XGBoost model ranks the evaluation features by importance to obtain the ranked evaluation features;

[0216] The sorted evaluation features are input into the deep neural network model, and the deep neural network model evaluates the capabilities of the marketers based on the mapping relationship to obtain the capability evaluation results; the mapping relationship is the correspondence between the evaluation features and the capability dimensions that characterize the capabilities of the marketers.

[0217] Optionally, the second processing unit 30 is more specifically configured to:

[0218] Determine whether the evaluation characteristic meets any preset conditions, including: being greater than the mean of the evaluation characteristic and having a fluctuation less than a preset fluctuation range;

[0219] If satisfied, the weight of the assessment feature in the capability dimension is increased compared to other assessment features;

[0220] If not, the weight of the assessment feature in the capability dimension will be reduced compared to other assessment features.

[0221] Optionally, the second processing unit 30 is more specifically configured to:

[0222] Determine whether the assessment characteristics have the preset required weights;

[0223] If so, the weight value of the assessment characteristic within the capability dimension is determined according to the preset required weight.

[0224] Optionally, a deep neural network model can be used to assess marketers’ capabilities based on the mapping relationships, including:

[0225] Determine the combination of features present in the assessment characteristics;

[0226] Conduct a competency assessment of marketing personnel based on assessment characteristics and / or combinations of characteristics.

[0227] Optionally, the capability assessment results include a real-time capability ranking, a multi-dimensional capability radar chart generated based on the scores corresponding to multiple capability dimensions, and the score gap between the current marketer and a benchmark marketer in the capability dimension, where the benchmark marketer is a marketer whose score falls within the top preset percentile in the industry.

[0228] Optionally, the capability assessment model is dynamically updated; the conditions for the capability assessment model dynamic update are: the number of new orders placed by the timing or marketing personnel reaches a preset threshold.

[0229] Optionally, the collection unit 10 is specifically configured to:

[0230] Obtain preselected data at a preset frequency;

[0231] Perform data cleaning on the pre-selected data to obtain the target data.

[0232] Below, reference Figure 6 To describe the electronic device provided by the embodiment of the present application, the electronic device provided by the embodiment may include: at least one processor 100, at least one communication interface 200, at least one memory 300 and at least one communication bus 400;

[0233] In the embodiment of the present invention, the number of the processor 100, the communication interface 200, the memory 300, and the communication bus 400 is at least one, and the processor 100, the communication interface 200, and the memory 300 communicate with each other through the communication bus 400; obviously, Figure 6 The communication connections shown for the processor 100, communication interface 200, memory 300, and communication bus 400 are merely optional;

[0234] Optionally, the communication interface 200 may be an interface of a communication module, such as an interface of a GSM module; the processor 100 may be a central processing unit (CPU), or an application specific integrated circuit (ASIC), or one or more integrated circuits configured to implement an embodiment of the present invention.

[0235] The memory 300 may include a high-speed RAM memory, and may also include a non-volatile memory (non-volatile memory), such as at least one disk memory.

[0236] The processor 100 is specifically configured to execute the application program in the memory to implement the steps of the above-mentioned marketing personnel capability assessment method.

[0237] The basic principles of the present application have been described above in conjunction with specific embodiments. However, it should be noted that the advantages, strengths, and effects mentioned in this application are merely illustrative and not restrictive, and it should not be assumed that these advantages, strengths, and effects are required of each embodiment of this application. In addition, the specific details disclosed above are merely illustrative and facilitating understanding, and are not restrictive. The above details do not limit this application to necessarily being implemented using the above specific details.

[0238] The block diagrams of the devices, devices, equipment, and systems involved in this application are merely illustrative examples and are not intended to require or imply that they must be connected, arranged, or configured in the manner shown in the block diagrams. As will be appreciated by those skilled in the art, these devices, devices, equipment, and systems can be connected, arranged, or configured in any manner. Words such as "include," "comprise," "have," and the like are open-ended words, meaning "including but not limited to," and can be used interchangeably therewith. The words "or" and "and" used herein refer to the words "and / or" and can be used interchangeably therewith, unless the context clearly indicates otherwise. The word "such as" used herein refers to the phrase "such as but not limited to," and can be used interchangeably therewith.

[0239] It should also be noted that in the apparatus, device, and method of the present application, each component or each step can be decomposed and / or recombined, and such decomposition and / or recombination should be regarded as equivalent solutions of the present application.

Claims

1. A method for evaluating marketing personnel capabilities, characterized in that: include: Acquiring target data, wherein the target data is predetermined data related to the marketing personnel capability assessment; Determining evaluation characteristics for each marketing personnel based on the target data; Inputting the evaluation characteristics into a capability evaluation model to obtain capability evaluation results for each of the marketing personnel; The capability assessment model is a two-stage model consisting of an XGBoost model and a deep neural network model obtained through training based on historical target data; wherein, the XGBoost model is used to rank the importance of the assessment features, and the deep neural network model is used to assess the capability of each of the marketing personnel based on the ranked assessment features.

2. The method according to claim 1, characterized in that Determining evaluation characteristics for each marketing personnel based on the target data includes: Sorting the target data in time sequence to obtain time sequence target data; Analyzing the latest time series target data within a first preset time period based on an observation window that slides at a preset frequency to obtain a first eigenvalue; Calculating an arithmetic mean of the first eigenvalue and a second eigenvalue, where the second eigenvalue is a eigenvalue obtained by analyzing the time series target data within a latest second preset time period based on the observation window; the second preset time period is greater than the first preset time period; The arithmetic mean is used as the evaluation feature.

3. The method according to claim 1, characterized in that Inputting the evaluation characteristics into the capability evaluation model to obtain capability evaluation results for each marketing personnel includes: Inputting the evaluation features into the XGBoost model, and having the XGBoost model sort the importance of the evaluation features to obtain the sorted evaluation features; The sorted evaluation features are input into the deep neural network model, and the deep neural network model performs capability evaluation on the marketing personnel based on the mapping relationship to obtain the capability evaluation result; the mapping relationship is the correspondence between the evaluation features and the capability dimension that characterizes the capability of the marketing personnel.

4. The method according to claim 3, characterized in that The XGBoost model ranks the evaluation features by importance, including: Determining whether the evaluation feature satisfies any preset condition, the preset condition including: being greater than the mean of the evaluation feature and having a fluctuation less than a preset fluctuation range; If satisfied, the weight of the assessment feature in the capability dimension is increased compared to other assessment features; If not, the weight of the evaluation feature in the capability dimension is reduced compared to other evaluation features.

5. The method according to claim 3, characterized in that The XGBoost model ranks the evaluation features by importance, including: determining whether the assessment feature has a preset required weight; If yes, the weight value of the assessment feature within the capability dimension is determined according to the preset required weight.

6. The method according to claim 3, characterized in that The deep neural network model evaluates the marketing personnel's capabilities based on the mapping relationship, including: determining a combination of features present in the evaluation features; Based on the evaluation characteristics and / or the combination of characteristics, the marketing personnel are evaluated for their capabilities.

7. The method according to claim 3, characterized in that The capability assessment results include real-time capability rankings, a multi-dimensional capability radar chart generated based on scores corresponding to multiple capability dimensions, and the score gap between the current marketer and the benchmark marketer on the capability dimensions, where the benchmark marketer is a marketer whose score falls within the top preset percentage in the industry.

8. The method according to claim 1, characterized in that The capability assessment model is dynamically updated; the condition for the capability assessment model to be dynamically updated is: the number of new orders placed by the timing or marketing personnel reaches a preset threshold.

9. The method according to claim 1, characterized in that The acquiring target data includes: Obtain preselected data at a preset frequency; The preselected data is cleaned to obtain the target data.

10. A marketing personnel capability assessment system, characterized in that: include: A collection unit, configured to obtain target data, wherein the target data is predetermined data related to the marketing personnel capability assessment; A first processing unit, configured to determine evaluation characteristics for each marketing personnel based on the target data; A second processing unit is configured to input the evaluation characteristics into a capability evaluation model to obtain capability evaluation results for each of the marketing personnel; The capability assessment model is a two-stage model consisting of an XGBoost model and a deep neural network model obtained through training based on historical target data; wherein, the XGBoost model is used to rank the importance of the assessment features, and the deep neural network model is used to assess the capability of each of the marketing personnel based on the ranked assessment features.

Images