A method for analyzing consumer brand loyalty based on multi-dimensional data

Abstract

The present invention relates to the field of market analysis, and discloses a method for analyzing consumer brand loyalty based on multi-dimensional data, comprising the following steps: firstly, collecting multi-dimensional data; then performing data preprocessing on the collected multi-dimensional data to obtain standardized multi-dimensional data; obtaining a purchasing behavior index based on purchasing behavior data processing, obtaining a consumer attribute index based on consumer attribute data processing, and obtaining a market environment index based on market environment data processing; inputting the purchasing behavior index, the consumer attribute index and the market environment index into a consumer brand loyalty analysis model to obtain consumer brand loyalty; comparing the obtained consumer brand loyalty with a preset consumer brand loyalty interval, and dividing the consumer brand loyalty evaluation into three levels of high, medium and low according to the comparison result; and formulating targeted marketing strategies according to the consumer brand loyalty evaluation results of different levels.

Description

Technical Field

[0001] The present invention relates to the field of market analysis, and in particular to a method for analyzing consumer brand loyalty based on multi-dimensional data. Background Art

[0002] With the development of social diversification, consumer behavior patterns have changed dramatically. Consumers no longer focus only on the functionality of products, but are increasingly pursuing the emotional value and social identity that brands carry. For example, some sports brands have cooperated with environmental protection organizations to launch products made of environmentally friendly materials, attracting consumers who value environmental protection concepts.

[0003] At the same time, consumers’ decision-making process has become more complex and is influenced by information from multiple online and offline channels, such as product reviews on social media and recommendations from friends.

[0004] However, traditional brand loyalty analysis methods are mostly based on simple single-dimensional data such as purchase frequency and number of repeated purchases. It is difficult to gain comprehensive and in-depth insights into consumers' complex behaviors and psychology, and cannot meet the needs of companies' precision marketing. Therefore, there is an urgent need for a method that can analyze consumer brand loyalty based on multi-dimensional data. Summary of the invention

[0005] The purpose of the present invention is to provide a consumer brand loyalty analysis method based on multi-dimensional data, so as to solve the technical problem that traditional brand loyalty analysis methods are mostly based on simple single-dimensional data such as purchase frequency and number of repeated purchases, which makes it difficult to fully and deeply understand the complex behavior and psychology of consumers and cannot meet the technical problems of enterprises' precision marketing needs.

[0006] The purpose of the present invention can be achieved through the following technical solutions: A method for analyzing consumer brand loyalty based on multi-dimensional data includes the following steps: S100, collecting multi-dimensional data, wherein the multi-dimensional data includes: purchase behavior data, consumer attribute data, and market environment data; S200, performing data preprocessing on the collected multi-dimensional data to obtain standardized multi-dimensional data; the data preprocessing includes data cleaning, removal of duplications, errors, missing values, and normalization; S300, obtaining a purchasing behavior index based on the purchasing behavior data, obtaining a consumer attribute index based on the consumer attribute data, and obtaining a market environment index based on the market environment data; S400, inputting the purchase behavior index, the consumer attribute index and the market environment index into a consumer brand loyalty analysis model to obtain consumer brand loyalty; S500, comparing the obtained consumer brand loyalty with a preset consumer brand loyalty interval, if the consumer brand loyalty> the preset consumer brand loyalty interval, evaluating the consumer brand loyalty as high-end, if the consumer brand loyalty∈the preset consumer brand loyalty interval, evaluating the consumer brand loyalty as medium-end, if the consumer brand loyalty<the preset consumer brand loyalty interval, evaluating the consumer brand loyalty as low-end; S600. Develop targeted marketing strategies based on the evaluation results of consumer brand loyalty at different levels.

[0007] The purchase behavior data includes purchase frequency, purchase amount, purchase time interval, purchase channel preference and brand conversion probability data; The consumer attribute data includes demographic characteristics and family life cycle information; The market environment data includes the number of competing products and the growth rate of competing products.

[0008] Furthermore, the expression of the consumer brand loyalty analysis model is: ; In the formula, Indicates consumer brand loyalty, Indicates The weight coefficient corresponding to the index is The values ​​range from 1 to 3, corresponding to the purchasing behavior index PB, consumer attribute index CA and market environment index ME respectively; Indicates The data influence coefficient corresponding to the index is: Indicates The current value of the index, Indicates The exponential decay coefficient, Indicates The trend change coefficient of an index; in, ; In the formula, Indicates The weight factor of each detection period, Indicates The index is The coefficient of dispersion within the detection period is Indicates the number of detection periods; in, ; In the formula, Indicates The stability coefficient of an index is based on the current moment Go back to the preset time period to get the marked time , Indicates The difference between the current time and the marked time of an index, , Indicates the The cumulative change of an index.

[0009] Furthermore, within a time period The cumulative change of the index The method to obtain is: Divide a time period into equal parts according to the set time period length preset time periods, and obtain the first The exponential curve over time ; By formula: , calculate the first The cumulative change of the index ; In the formula, Indicates the preset time period. Indicates The trend influence coefficient of the preset time period, and the The longer the end time of a preset period is from the current time, the The smaller.

[0010] Furthermore, the The index is The coefficient of dispersion within the detection period The expression is: ; In the formula, , Respectively represent The index is The maximum and minimum values ​​within the detection period, Indicates The index is The number of samples in a detection period, Indicates The observed value at sampling time, For the The average value of the index.

[0011] Furthermore, the method for obtaining the purchasing behavior index PB is: The pre-processed purchase frequency , Purchase amount , Purchase interval , purchase channel preference and brand switching probability Substitute the following formula: ; ; In the formula, is a judgment function about the purchase channel preference. If the online platform accounts for more than 1 / 2 of the current brand's sales, and the consumer's purchase channel preference is the online platform, then ,otherwise, ; If the offline sales of the current brand account for more than 1 / 2, and consumers prefer offline purchase channels, then ,otherwise, ; If the current brand's sales volume is the same for online platforms and offline platforms, then ; Indicates the number of times consumers purchase other brands during a set period of time; Indicates the number of times consumers purchase the current brand within a set period of time.

[0012] Furthermore, the method for obtaining the consumer attribute index CA is: Obtaining demographic characteristics and family life cycle information, wherein the demographic characteristics include age, income, and gender; the family life cycle information includes: single, married without children, and married with children; Input demographic characteristics and family life cycle information into the pre-trained neural network model and output the consumer attribute index CA; The neural network model is one of a multi-layer perceptron neural network, a radial basis function neural network and a long short-term memory network.

[0013] Furthermore, the method for obtaining the market environment index ME is: The number of competing products and the growth rate of competing products within a cycle Substitute the following formula: The market environment index ME is calculated; In the formula, Indicates the growth rate of reference competing products within a period, Indicates the number of reference competing products.

[0014] Beneficial effects of the present invention: (1) The present invention collects multi-dimensional data covering purchasing behavior, consumer attributes, and market environment, and comprehensively and deeply understands consumer behavior patterns, personal characteristics, and market dynamics, providing a rich data foundation for accurate analysis of brand loyalty and avoiding the one-sidedness of analysis caused by single-dimensional data; (2) The present invention converts various multi-dimensional data into corresponding purchasing behavior indexes, consumer attribute indexes and market environment indexes, realizes concise quantitative expression of complex data, facilitates subsequent model calculation and analysis, and highlights the comprehensive impact of various dimensional data on brand loyalty; then, using the consumer brand loyalty analysis model, the dimensional index is input as an input variable into the consumer brand loyalty analysis model, which can accurately predict consumer brand loyalty, improve the accuracy and generalization ability of the prediction, and provide strong data support for corporate decision-making; (3) The present invention clearly divides consumer brand loyalty into three levels: high, medium and low, by comparing with the preset consumer brand loyalty range, so that enterprises can quickly identify consumer groups with different loyalty levels and carry out targeted marketing activities. At the same time, based on the evaluation results of consumer brand loyalty at different levels, differentiated marketing strategies can be formulated; for high-loyalty consumers, emotional connections can be enhanced; for medium-loyalty consumers, consumption can be stimulated and brand awareness can be improved; for low-loyalty consumers, the reasons can be deeply analyzed and products or services can be improved, thereby effectively improving the utilization efficiency of marketing resources and enhancing brand competitiveness and market share. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] The present invention will be further described below in conjunction with the accompanying drawings.

[0016] Figure 1 It is a diagram of the method steps of the present invention. DETAILED DESCRIPTION

[0017] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0018] See also Figure 1 As shown, the present invention is a method for analyzing consumer brand loyalty based on multi-dimensional data, comprising the following steps: S100, collecting multi-dimensional data, wherein the multi-dimensional data includes: purchase behavior data, consumer attribute data, and market environment data; S200, performing data preprocessing on the collected multi-dimensional data to obtain standardized multi-dimensional data; the data preprocessing includes data cleaning, removal of duplications, errors, missing values, and normalization; S300, obtaining a purchasing behavior index based on the purchasing behavior data, obtaining a consumer attribute index based on the consumer attribute data, and obtaining a market environment index based on the market environment data; S400, inputting the purchase behavior index, the consumer attribute index and the market environment index into a consumer brand loyalty analysis model to obtain consumer brand loyalty; S500, comparing the obtained consumer brand loyalty with a preset consumer brand loyalty interval, if the consumer brand loyalty> the preset consumer brand loyalty interval, evaluating the consumer brand loyalty as high-end, if the consumer brand loyalty∈the preset consumer brand loyalty interval, evaluating the consumer brand loyalty as medium-end, if the consumer brand loyalty<the preset consumer brand loyalty interval, evaluating the consumer brand loyalty as low-end; S600. Develop targeted marketing strategies based on the evaluation results of brand loyalty of consumers at different levels. For example: Highly loyal consumers: Provide exclusive membership services, such as priority purchase rights, customized product recommendations, and special birthday discounts, to further enhance their emotional connection with the brand; For medium-loyal consumers: Carry out promotional activities, such as discounts, free gifts, etc., to stimulate them to increase the frequency and amount of purchases; at the same time, strengthen brand publicity and interaction to improve brand awareness and reputation; Low-loyalty consumers: Analyze the reasons in depth, collect feedback through questionnaires, customer interviews, etc., and improve products or services in a targeted manner; if low loyalty is caused by product quality issues, strengthen quality control; if it is because competitors’ promotional activities have attracted consumers, formulate corresponding competitive strategies.

[0019] The purchase behavior data includes purchase frequency, purchase amount, purchase time interval, purchase channel preference and brand conversion probability data; The consumer attribute data includes demographic characteristics and family life cycle information; The market environment data includes the number of competing products and the growth rate of competing products.

[0020] In the present invention, by collecting multi-dimensional data covering purchase behavior, consumer attributes and market environment, a comprehensive and in-depth understanding of consumer behavior patterns, personal characteristics and market dynamic changes is obtained, providing a rich data basis for accurate analysis of brand loyalty and avoiding the one-sidedness of analysis caused by single-dimensional data; various multi-dimensional data are then converted into corresponding purchase behavior indexes, consumer attribute indexes and market environment indexes, achieving concise quantitative expression of complex data, facilitating subsequent model calculation and analysis, and highlighting the comprehensive impact of various dimensional data on brand loyalty; Then, the consumer brand loyalty analysis model is used to input the index of each dimension as an input variable into the consumer brand loyalty analysis model, which can accurately predict consumer brand loyalty, improve the accuracy and generalization ability of prediction, and provide strong data support for corporate decision-making; then, by comparing with the preset consumer brand loyalty range, consumer brand loyalty is clearly divided into three levels: high, medium, and low, so that companies can quickly identify consumer groups with different loyalty levels and carry out targeted marketing activities; Finally, based on the evaluation results of brand loyalty of consumers in different levels, differentiated marketing strategies are formulated; emotional connections are enhanced for highly loyal consumers, consumption is stimulated and brand awareness is improved for medium-loyalty consumers, and the reasons for low-loyalty consumers are deeply analyzed and products or services are improved, effectively improving the utilization efficiency of marketing resources and enhancing brand competitiveness and market share.

[0021] The expression of the consumer brand loyalty analysis model is: ; In the formula, Indicates consumer brand loyalty, Indicates The weight coefficient corresponding to each index is determined based on historical data analysis. The values ​​range from 1 to 3, corresponding to the purchasing behavior index PB, consumer attribute index CA and market environment index ME respectively; Indicates The data impact coefficient corresponding to each index is formulated based on historical data and empirical data. The value range is 0 to 1. Indicates The current value of the index, Indicates The exponential decay coefficient, Indicates The trend change coefficient of an index; in, ; In the formula, Indicates The weight factor of each detection period ranges from 0 to 1 and is formulated based on historical data and empirical data. Indicates The index is The coefficient of dispersion within the detection period is Indicates the number of detection periods; in, ; In the formula, Indicates The stability coefficient of an index is determined based on historical data analysis and is based on the current moment. Go back to the preset time period to get the marked time , Indicates The difference between the current time and the marked time of an index, , Indicates the The cumulative change of an index.

[0022] In the present invention, by accumulating and summing the consumer purchasing behavior index PB, the consumer attribute index CA and the market environment index ME, the expression is: , thus calculating the consumer brand loyalty, obviously, not only through the current value of each index, but also combined with the discrete degree of each index's historical data and cumulative change , so that the changing trend of each index can be predicted, and the accurate assessment of consumer brand loyalty can be comprehensively improved to improve the accuracy of the consumer brand loyalty analysis model.

[0023] In a time period The cumulative change of the index The method to obtain is: Divide a time period into equal parts according to the set time period length preset time periods, and obtain the first The exponential curve over time ; By formula: , calculate the first The cumulative change of the index ; In the formula, Indicates the preset time period. Indicates The trend influence coefficient of the preset time period, and the The longer the end time of a preset period is from the current time, the The smaller.

[0024] The said The index is The coefficient of dispersion within the detection period The expression is: ; In the formula, , Respectively represent The index is The maximum and minimum values ​​within the detection period, Indicates The index is The number of samples in a detection period, Indicates The observed value at sampling time, For the The average value of the index.

[0025] The present invention provides a method for obtaining the cumulative change amount within a time period, first by dividing a time period into equal parts according to a set time period length preset time periods, and obtain the first The exponential curve over time , and then through the formula: , calculate the first The cumulative change of the index ; Obviously, if the cumulative change in each preset period The larger the value is, the greater the change in the current index is within the preset period. Substituting it into the formula The first The smaller the trend change coefficient of an index, the greater the probability that consumer brand loyalty will decrease; similarly, , calculate the first The index is The coefficient of dispersion within the detection period Obviously, if the dispersion of the current index in each detection period is greater, the current index is more unstable. Substituting it into the formula The smaller the attenuation coefficient obtained, the smaller the impact on consumer brand loyalty, and therefore the greater the probability of a decrease in consumer brand loyalty.

[0026] The method for obtaining the purchasing behavior index PB is as follows: The pre-processed purchase frequency , Purchase amount , Purchase interval , purchase channel preference and brand switching probability Substitute the following formula: ; ; In the formula, is a judgment function about the purchase channel preference. If the online platform accounts for more than 1 / 2 of the current brand's sales, and the consumer's purchase channel preference is the online platform, then ,otherwise, ; If the offline sales of the current brand account for more than 1 / 2, and consumers prefer offline purchase channels, then ,otherwise, ; If the current brand's sales volume is the same for online platforms and offline platforms, then ; Indicates the number of times consumers purchase other brands during a set period of time; Indicates the number of times consumers purchase the current brand within a set period of time.

[0027] In the present invention, the five purchasing behavior parameters of consumers are integrated and then substituted into the formula Calculation can get the purchase behavior index PB. Obviously, if the purchase frequency The higher the value, the more times consumers buy the current brand in the same period of time, so their loyalty to the current brand is obviously higher, and the purchase amount The larger the value, the more money consumers spend on the current brand in the same period of time. Therefore, their loyalty to the current brand is obviously higher. The time interval between consumers purchasing the current brand in the same period of time is The shorter the purchase channel, the higher the loyalty to the current brand. If the consumer's purchase channel preference is more consistent with the channel where the current brand's sales account for a larger proportion, it is obvious that the loyalty to the current brand is higher, and the consumer's brand switching probability is higher. The smaller it is, the higher the loyalty to the current brand, which has achieved the purpose of comprehensively analyzing consumer purchasing behavior.

[0028] The method for obtaining the consumer attribute index CA is: Obtaining demographic characteristics and family life cycle information, wherein the demographic characteristics include age, income, and gender; the family life cycle information includes: single, married without children, and married with children; Input demographic characteristics and family life cycle information into the pre-trained neural network model and output the consumer attribute index CA; The neural network model is one of a multi-layer perceptron neural network, a radial basis function neural network and a long short-term memory network. By assigning values ​​to demographic characteristics and family life cycle information according to experience and then inputting them into a pre-trained neural network model, the consumer attribute index CA can be obtained; For example: Demographic characteristics: Age is divided into four age groups: under 20, 20-40, 40-60, and over 60, and the values ​​are assigned as 0, 3, 2, and 1 respectively; income does not need to be assigned, it can be collected directly; gender can be assigned as 0 for male and 1 for female; it indicates a consumer hospital; Family life cycle information: singles are assigned a value of 1, newly married people are assigned a value of 2, and people with children are assigned a value of 3; it indicates the willingness to change brands, and the lower the willingness to change brands, the higher the value; The above assignments are only for example, and are actually determined based on statistical data analysis. Obviously, the larger the consumer attribute index CA is, the higher the consumer's willingness to consume and the lower the probability of changing brands, so the greater the brand loyalty.

[0029] The method for obtaining the market environment index ME is: The number of competing products and the growth rate of competing products within a cycle Substitute the following formula: The market environment index ME is calculated; In the formula, Indicates the growth rate of reference competing products within a period, Indicates the number of reference competing products. The greater the growth rate of competing products in a cycle and the more competing products there are, the greater the competition for the current brand, so the wider the range of choices for consumers in the purchase process. Obviously, the greater the growth rate of competing products in a cycle and the more competing products there are, the smaller the market environment index ME is, and the smaller the impact on consumer brand loyalty is.

[0030] It should be noted that the calculation formulas and various parameters involved in the calculation in the present invention have been dimensionally processed in advance, and the process of dimensionless processing is well known in the industry and will not be described here.

[0031] The above is a detailed description of an embodiment of the present invention, but the content is only a preferred embodiment of the present invention and cannot be considered to limit the scope of implementation of the present invention. All equivalent changes and improvements made within the scope of the present invention should still fall within the scope of the patent coverage of the present invention.

Claims

1. A method for analyzing consumer brand loyalty based on multi-dimensional data, characterized in that: The steps include: S100, collecting multi-dimensional data, wherein the multi-dimensional data includes: purchase behavior data, consumer attribute data, and market environment data; S200, performing data preprocessing on the collected multi-dimensional data to obtain standardized multi-dimensional data; the data preprocessing includes data cleaning, removal of duplications, errors, missing values, and normalization; S300, obtaining a purchasing behavior index based on the purchasing behavior data, obtaining a consumer attribute index based on the consumer attribute data, and obtaining a market environment index based on the market environment data; S400, inputting the purchase behavior index, the consumer attribute index and the market environment index into a consumer brand loyalty analysis model to obtain consumer brand loyalty; S500, comparing the obtained consumer brand loyalty with a preset consumer brand loyalty range, and dividing the consumer brand loyalty evaluation into three levels: high, medium, and low according to the comparison result; S600. Develop targeted marketing strategies based on the evaluation results of consumer brand loyalty at different levels.

2. The method for analyzing consumer brand loyalty based on multi-dimensional data according to claim 1, characterized in that: The purchase behavior data includes purchase frequency, purchase amount, purchase time interval, purchase channel preference and brand conversion probability data; The consumer attribute data includes demographic characteristics and family life cycle information; The market environment data includes the number of competing products and the growth rate of competing products.

3. The method for analyzing consumer brand loyalty based on multi-dimensional data according to claim 2, characterized in that: The expression of the consumer brand loyalty analysis model is: ; In the formula, Indicates consumer brand loyalty, Indicates The weight coefficient corresponding to the index is The values ​​range from 1 to 3, corresponding to the purchasing behavior index PB, consumer attribute index CA and market environment index ME respectively; Indicates The data influence coefficient corresponding to the index is: Indicates The current value of the index, Indicates The exponential decay coefficient, Indicates The trend change coefficient of an index; in, ; In the formula, Indicates The weight factor of each detection period, Indicates The index is The coefficient of dispersion within the detection period is Indicates the number of detection periods; in, ; In the formula, Indicates The stability coefficient of the index is based on the current moment Go back to the preset time period to get the marked time , Indicates The difference between the current time and the marked time of an index, , Indicates the The cumulative change of an index.

4. The method for analyzing consumer brand loyalty based on multi-dimensional data according to claim 3 is characterized in that: In a time period The cumulative change of the index The method to obtain is: Divide a time period into equal parts according to the set time period length preset time periods, and obtain the first The exponential curve over time ; By formula: , calculate the first The cumulative change of the index ; In the formula, Indicates the preset time period. Indicates The trend influence coefficient of the preset time period, and the The longer the end time of a preset period is from the current time, the The smaller.

5. The method for analyzing consumer brand loyalty based on multi-dimensional data according to claim 4, characterized in that: The said The index is The coefficient of dispersion within the detection period The expression is: ； In the formula, , Respectively represent The index is The maximum and minimum values ​​within the detection period, Indicates The index is The number of samples in a detection period, Indicates The observed value at sampling time, For the The average value of the index.

6. The method for analyzing consumer brand loyalty based on multi-dimensional data according to claim 2, characterized in that: The method for obtaining the purchasing behavior index PB is as follows: The pre-processed purchase frequency , Purchase amount , Purchase interval , purchase channel preference and brand switching probability Substitute the following formula: ； ； In the formula, is a judgment function about the purchase channel preference. If the online platform accounts for more than 1 / 2 of the current brand's sales, the consumer's purchase channel preference is the online platform. ,otherwise, ; If the offline sales of the current brand account for more than 1 / 2, and consumers prefer offline purchase channels, then ,otherwise, ; If the current brand's sales volume is the same for online platforms and offline platforms, then ; Indicates the number of times consumers purchase other brands during a set period of time; Indicates the number of times consumers purchase the current brand within a set period of time.

7. The method for analyzing consumer brand loyalty based on multi-dimensional data according to claim 2, characterized in that: The method for obtaining the consumer attribute index CA is: Obtaining demographic characteristics and family life cycle information, wherein the demographic characteristics include age, income, and gender; the family life cycle information includes: single, married without children, and married with children; Input demographic characteristics and family life cycle information into the pre-trained neural network model and output the consumer attribute index CA; The neural network model is one of a multi-layer perceptron neural network, a radial basis function neural network and a long short-term memory network.

8. The method for analyzing consumer brand loyalty based on multi-dimensional data according to claim 2, characterized in that: The method for obtaining the market environment index ME is: The number of competing products and the growth rate of competing products within a cycle Substitute the following formula: The market environment index ME is calculated; In the formula, Indicates the growth rate of reference competing products within a period, Indicates the number of reference competing products.

Images