Analysis of Influencing Factors on the Willingness and Behavioral Consistency of Chinese Consumers to Purchase Tea via E-Commerce Platforms

Abstract

Tea is a global economic crop. In the traditional sales model, the quality of tea is difficult to judge via external clues, and it basically relies on consumers to taste and experience it firsthand. However, currently, most e-commerce platforms can only provide consumers with product information and cannot provide experiential services, which strengthens consumers’ purchasing concerns and makes it difficult for them to take practical actions even if they have the desire to purchase tea online. Therefore, this article is based on a questionnaire survey of consumers in China, the world’s largest tea producing and selling country, using unordered multi classification logistic analysis data and calculating the marginal effect proportion of consistency occurrence. Through data analysis, it is shown that there is a serious inconsistency between consumers’ willingness and behavior in choosing online tea purchases. However, under the influence of some internal and external factors, there will also be positive changes; for example, the level of education, online shopping age, familiarity with tea, convenience, product diversification, online evaluation, and other variables will increase the motivation for consumers to convert their purchase intention into actual behavior. In addition, increasing the level of variables such as age, cultural association, cultural experience, convenience, information reliability, award-winning status, familiarity with tea, product diversity, online evaluation, and service attitude in online shopping can enhance consumers’ willingness to shop online and reduce extreme situations where there is neither intention nor behavior. This study provides evidence from a consumer perspective to enhance the efficiency of tea e-commerce marketing and proposes countermeasures and suggestions based on the research results, which could also provide ideas for marketing strategies for tea or other similar agricultural products.

1. Introduction

Tea is considered the second most popular beverage in the world after water [1,2]. According to a global market report, first, from the perspective of tea production and consumption, the global tea production in 2017 was 5.98 million tons, and the retail value of tea products was nearly 50 billion US dollars, maintaining a stable growth rate of about 4.4% per year in production and consumption for 10 consecutive years. Second, in terms of the number and economic importance of countries involved in tea production, there are a total of 48 countries in the world that produce tea, with China, Sri Lanka, and Kenya accounting for approximately half of the global tea production. Another 12 impoverished countries also produce tea, and most of the tea producers in these countries are small-scale farmers. It can be seen that tea production and consumption are of great significance for global agricultural economic development, especially for developing and impoverished countries.

China is the world’s largest producer and consumer of tea [3]. In the past few decades, the retail sales of tea in China have been growing at a high speed. Especially since the outbreak of the COVID-19 in 2020, the rise of Internet e-commerce has injected new impetus into the tea sales industry. Most residents have actively or passively chosen to accept isolation to avoid infection. The high-pressure situation has forced online shopping to become the main channel by which to meet daily consumption needs. Tea enterprises have gradually attached importance to the construction of their own e-commerce platforms. For example, China’s total tea e-commerce sales in 2020 reached 27.98 billion yuan, nearly twice that of 2019 before the COVID-19 outbreak. At present, tea has become an important agricultural pillar industry in the country [4,5,6].

Although e-commerce for agricultural products, including tea, has ushered in a good era of development, some scholars still point out that compared to traditional sales and other commodity categories of e-commerce, there is still significant room for improvement in the total agricultural product e-commerce market [7]. Accelerating the construction demand of the agricultural product e-commerce market has prompted more scholars to invest in the research of consumer online shopping psychological activity mechanisms. Existing research suggests that one of the purposes for consumers to purchase agricultural products through e-commerce channels is to reduce energy consumption and search costs. Therefore, perceived usefulness and perceived ease of use are also important factors affecting whether consumers use agricultural e-commerce [8]. In subjective variables, higher perceived value has an enhancing effect on consumers’ online shopping willingness and behavior, while perceived risk is the opposite. At the same time, perceived risk also negatively affects perceived value [9]. In addition, some external variables can also affect consumers’ willingness and behavior to purchase agricultural products through e-commerce channels, such as the speed and quality of logistics delivery [10], the subjective normative role in social networks [11,12], and the reputation of merchants’ e-commerce online platforms [13].

In addition, some scholars, from the perspective of the decision-making process from purchase intention to behavior transition, believe that one of the main reasons for causing market gap problems is that the severity of the inconsistency between intention and behavior varies across different channels or products [14,15]. Although online purchasing provides many conveniences, it also exacerbates the information asymmetry between buyers and sellers of agricultural products, placing consumers at a disadvantage in obtaining product information, being unable to judge the quality of agricultural products through actual contact. Therefore, constrained by other factors such as consumers’ individual characteristics, product perception, supply channels, product prices, quality certification, and social environment [16], consumers frequently exhibit a desire to purchase, but the decision-making process without purchasing behavior is inconsistent, thereby limiting the development of the consumer market for agricultural e-commerce [17,18].

Overall, existing research has extensively explored the factors that affect consumers’ willingness and behavior to purchase agricultural products through e-commerce. The research on the relationship, quantity, theory, and model construction of internal and external variables has also been very mature, making outstanding contributions to the field of e-commerce marketing. However, existing literature still has the following shortcomings: (1) In terms of research perspective, only a few scholars have realized the inconsistency between consumers’ willingness to purchase agricultural products in e-commerce and their actual behavior, and they have conducted limited research. However, they cannot fully explain the influencing factors of the inconsistency between willingness and behavior, and further exploration of this topic is needed in the future. (2) In terms of research subjects, most consumer studies on the consistency of willingness and behavior in purchasing agricultural products are mainly focused on vegetables, fruits, and livestock products [19,20], while there is a lack of research on the consistency of consumers’ willingness and behavior with respect to economic crops, especially online tea shopping. This cannot provide a solid preliminary research foundation for the development of marketing strategies for high-quality agricultural products.

Therefore, based on a review of the relevant research literature mentioned above, this article designs a research model framework based on previous research and collects consumers’ views on online tea shopping (in China, mainly referring to tea processed using traditional techniques, i.e., without adding other substances, and sold in the form of leaves) through questionnaire surveys in order to explore the influencing factors of consumers’ willingness and behavioral consistency in online tea shopping. First, compared to other agricultural products, tea is a high value-added agricultural product in China. Due to its unique cultural attributes and scarcity, it is often accompanied by far higher and expensive prices than other agricultural products, which means that there will be more influencing factors or a stronger impact that hinders the conversion of purchasing intention to purchasing behavior. Second, with the changing times, the objective reality of increasing confidence in Chinese culture and the emergence of new types of tea drinks, such as blended tea (milk tea, etc.), tea beverages, and reprocessed tea (jasmine tea, peach oolong tea, etc.), have given the market for tea opportunities to meet the taste and cultural pursuits of young people in the new era. Research on online tea shopping behavior is an important topic that satisfies current market and social needs. The main contributions of this article and its differences from previous research are as follows: (1) this article focuses on the impact of cultural perception on consumers’ inconsistent willingness and behavior in the online shopping of agricultural products and proposes countermeasures and suggestions, providing a theoretical basis for promoting the characteristic marketing of agricultural product e-commerce and expanding the consumer market; (2) it verifies the impact of existing theories and variables on the consistency of consumer willingness and behavior in agricultural e-commerce, rather than just discussing the direct impact on willingness or behavior, which constitutes an innovation of existing theories and variables; (3) taking tea consumers as the research object can provide market planning and development guidance for high-end agricultural products that will move in high value-added directions in the future.

2. Materials and Methods

2.1. Theoretical Basis and Research Hypothesis

Technology acceptance theory, perceived value theory, and clue theory are classic theories commonly used in the field of marketing. These theories indicate that consumers’ motivation to use a certain technology is usually composed of a combination of internal and external factors, including subjective judgments about the perceived ease of use, perceived usefulness, and value return of the technology, as well as behavioral responses to objective external cues and stimuli [21,22,23,24,25]. Previous studies have shown that the variables involved in the above theory have a significant impact on consumers’ willingness and behavior to purchase agricultural products [26,27], but there is still a lack of sufficient research on whether it can improve the efficiency of converting consumer willingness into actual behavior, which is particularly important for high-value agricultural products. Therefore, based on the above theoretical review, this study selected the following variables to screen for key factors that affect consumers’ willingness and behavior to purchase tea online, and then constructed a theoretical framework model:

(1)

Consumer personal characteristics

Previous studies have shown that personal characteristics, such as gender, age, Monthly average income, and level of product awareness, have a significant impact on consumers’ willingness and behavior to choose e-commerce online purchasing methods and can also affect the consistency of consumers’ online purchasing willingness and behavior [28,29,30,31].

H1. 

Personal characteristics have a significant impact on the consistency of consumers’ willingness and behavior to purchase tea online.

(2)

Cultural perception

In real consumption scenarios, a good atmosphere makes consumers more inclined to implement purchasing behavior, and consumers always associate products with corresponding cultural experiences during the shopping process [32]. Tea, unlike other agricultural products, has high cultural attributes and product added value. Therefore, whether consumers obtain a sense of cultural experience from shopping experiences or associate it with the cultural atmosphere of tea is different from other agricultural products, which affects their psychological needs and the value perception of online tea shopping willingness and behavior.

H2. 

Cultural perception has a significant impact on the consistency of consumers’ willingness and behavior to purchase tea online.

(3)

Perceived usefulness and perceived ease of use

Perceived usefulness refers to the degree of consumer recognition that online purchasing methods can improve purchasing efficiency, such as the types and quantities of optional products, transaction convenience, etc. [33]. Perceived ease of use refers to the degree to which consumers recognize the authenticity of product information sold on online platforms and believe that purchasing a product is easy [34]. Perceived usefulness and perceived ease of use are two classic research variables in technology acceptance models. Research has shown that the degree to which consumers perceive usefulness and perceived ease of use determines the decision-making process of their purchasing behavior. In other words, the more consumers perceive online shopping as an efficient, reliable, and easy-to-use purchasing method, the easier it is for them to put their purchasing intentions into action [35,36,37].

H3. 

Perceived usefulness has a significant impact on the consistency of consumers’ willingness and behavior to purchase tea online.

H4. 

Perceived ease of use has a significant impact on the consistency of consumers’ willingness and behavior to purchase tea online.

(4)

Online shopping preferences

Consumers’ online shopping preferences refer to the differences in their focus on information on online shopping platforms. The product information released by e-commerce sales platforms is an important way for consumers to obtain product information. Consumers may decide whether to engage in purchasing behavior by comparing the unique information elements of e-commerce agricultural products such as price, quality, origin, online comments, transaction volume, and customer service attitude across different product stores or purchasing channels [38,39].

H5. 

Online shopping preferences have a significant impact on the consistency of consumers’ willingness and behavior to purchase tea online.

(5)

Trust endorsement

The authority guarantee implied by product information is the core content of trust endorsement. Packaging information, such as certification labels, brand image, and award status, will awaken the positive impression left by institutions, governments, or enterprises on consumers in long-term promotion and marketing activities. By generating a sense of trust, the degree of deviation caused by the uncertainty of unknown risks in consumer willingness and behavior decision-making can be reduced, and the positive proportion of purchase intention to purchase behavior and increase intention to behavior can be enhanced [40,41].

H6. 

Trust endorsement has a significant impact on the consistency of consumers’ willingness and behavior to purchase tea online.

The settings of all variables and influencing relationships in this study are shown in Figure 1.

2.2. Model Construction

This article takes the deviation of consistency between “consumers’ willingness and behavior to purchase tea online” as the dependent variable, and after selecting questionnaire items, it is mainly divided into three types of situations: unwillingness and no purchase behavior (negative consistency = 0); intention but no purchasing behavior (deviation = 1); willingness and purchasing behavior (positive consistency = 2). In addition, there are very few samples of unintentional purchasing behavior, and there is no sufficient theoretical explanation. Therefore, this study will not conduct research and exploration in this area [42]. According to the requirements of the research purpose, we selected the reference group with positive consistency. This study will use an unordered multiclass logistic regression model to analyze the influencing factors and differences of consumer willingness and behavior inconsistency during online tea shopping. The model is constructed as follows:

$$Z_i=\alpha +{\beta }_i{x}_i+\mu .$$

(1)

In Equation (1), α is the intercept, β is the estimated coefficient, x_i represents the explanatory variable, and μ is an error term. The probability of inconsistency between consumers’ willingness and behavior during online tea shopping is as follows:

$$P_i=F\left(\alpha +{\beta }_i{x}_i\right)=\frac{1}{1+e^{-\left(\alpha +{\beta }_i\right)}}.$$

(2)

In Equation (2), e represents the base of the natural logarithm. The estimation formula is as follows:

$$Ln\left(\frac{p_i}{1-p_i}\right)=z_i=\alpha +{\displaystyle \sum _{i=1}^6{\beta }_i{x}_i+\mu }$$

(3)

The event occurrence ratios in Equation (3) are as follows: the first type of event (negative consistency); the second type of event (deviation); and the third type of event (positive consistency). Meanwhile, based on previous research and the availability of actual data, this article introduces 6 categories of 21 explanatory variables. Among them, X1 represents personal basic characteristics, X2 represents the perception of tea culture during the online tea shopping process, X3 represents the perceived usefulness of whether using online channels can effectively improve purchasing efficiency, X4 represents the perceived ease of use of whether consumers can easily purchase tea through online channels, X5 represents consumers’ preference for different information during the online tea shopping process, and X6 represents consumers’ trust in authoritative endorsements.

2.3. Data Sources

China is not only the country with the longest history of planting, producing, and using tea in the world but also the country with the largest tea production in the world [5]. China has a large group of tea consumers and based on relevant research conducted in the Chinese market, the results are relatively representative of the industry market. This research questionnaire is based on the mature scale of previous literature [33,39,42,43,44]. At the same time, based on the characteristics of the research object, all variables except for personal characteristics were measured using a questionnaire based on the five point Likert scale: 1 represents “strongly disagree”; 2 represents “disagree”; 3 represents “average”; 4 represents “agree”; and 5 represents “strongly agree”.

This study conducted questionnaire survey data collection from May to July 2023 through China Questionnaire Star APP and online social media invitation filling. A total of 639 questionnaires were collected. After removing invalid questionnaires with obvious errors, random filling, and blank filling, 585 valid questionnaires remained, with an effective recovery rate of 91.55%. The variable description and preliminary statistical results are shown in

Table 1. Description of variables and preliminary statistical results.

Dimension	Items	Variable Description	Average	Standard Deviation	Minimum	Maximum
Dependent variable	Consistency	Negative consistency, Deviation, Positive consistency	1.21	0.85	0	2
Independent variable
Personal characteristic	Gender	Female, Male	1.36	0.48	1	5
	Age	20 and under, 21–30, 31–40, 41–50, 51 and above	2.25	1.06	1	5
	Education	Elementary and below, Middle school, Polytechnic school or High school, Junior college or Bachelor’s Degree, Graduate student and above	3.92	0.84	1	5
	Monthly average income	2000 and below, 2001~4000, 4001~6000, 6001~8000, 8001 and above.	2.26	1.35	1	5
	Online shopping age	1 year and below, 1~3 years, 4~6 years, more than 6 years	2.93	0.96	1	4
	Tea knowledge familiarity (multiple choice questions)	(1) Tea tree varieties; (2) Processing technology; (3) Health effects; (4) Cultural history; (5) Brewing method	2.40	1.20	1	5
Cultural perceptions	Cultural association	I can associate the tea culture of the tea I buy with the process of buying tea online	3.37	1.12	1	5
	Cultural experience	I can feel the tea culture connotation of the tea I purchased during the process of purchasing tea online	2.86	0.98	1	5
Perceived usefulness	Convenience	Buying tea online is easier and faster, saving time and effort	3.77	1.00	1	5
	Product diversification	More products to choose from when buying tea online	3.79	1.06	1	5
Perceived ease of use	Price transparency	Buy tea online for more transparent prices	3.55	1.09	1	5
	Credible information	The online information about the display of tea is true and reliable!	3.09	0.99	1	5
Online shopping preferences	price	I will compare the price difference between buying the same tea online and offline	3.51	1.10	1	5
	Quality	I will compare the difference in quality between the same tea purchased online and offline	3.59	1.03	1	5
	Origin	I will be looking at whether a particular tea is from a historical source or not	3.32	1.03	1	5
	Comment	I will be watching for tasting comments from other buyers in the store	3.56	1.11	1	5
	Trand volume	I will be looking at the sales of this tea in the store	3.62	1.03	1	5
	Business Services	I value the customer service shown in the tea shopping process Service attitude and professionalism	3.65	1.01	1	5
Trust endorsement	Certification labels	When I buy tea online, I will pay attention to whether the tea has obtained authorization Certification (organic, geo-protected, etc.)	3.61	1.09	1	5
	Brand image	When I buy tea online, I pay attention to the brand’s reputation in the industry	3.73	1.02	1	5
	Awards	I would look at how many awards the tea has won	3.05	1.07	1	5

.

2.4. Data Analysis

After the data collection is completed, the basic data of the survey subjects’ questionnaires are compiled using Excel, and then a single factor T-test is conducted using SPSS 22 software. The unordered multi classification logistic and marginal effects analysis is completed using Stata 15 software.

3. Results

3.1. Descriptive Statistical Analysis

The basic distribution of all survey object samples is shown in

Table 2. Basic characteristics of the sample.

Variant	Delineation Criteria	Sample Size	Frequency
Willingness	No	160	27.35%
	Yes	425	72.65%
Behavior	No	303	51.79%
	Yes	282	48.21%
Gender	Female	374	63.93%
	Male	211	36.07%
Age	20 years and under	135	23.08%
	21–30 years old	285	48.72%
	31–40 years old	77	13.16%
	41–50 years	62	10.60%
	51–60 years	26	4.44%
Education	Primary and below	12	2.05%
	Junior high school	21	3.59%
	Polytechnic school or High school	95	16.24%
	Junior college or Bachelor’s Degree	330	56.41%
	Graduate students and above	127	21.71%
Monthly average income	2000 and below	237	40.51%
	2001~4000	136	23.25%
	4001~6000	93	15.90%
	6001~8000	59	10.09%
	More than 8000	60	10.26%
Online shopping age	1 year and less	53	9.06%
	1~3 years	134	22.91%
	4~6 years	201	34.36%
	More than 6 years	197	33.68%
Tea knowledge familiarity	Not familiar	153	26.15%
	A little familiar	196	33.50%
	General familiarity	136	23.25%
	Quite familiar	50	8.55%
	Very familiar	50	8.55%

. The sample number of consumers willing to purchase tea online was 425, accounting for 72.65% of the total survey respondents. However, only 282 people actually took action, a decrease of 24.44% compared to the number of people willing to purchase tea online. From the sample distribution of personal characteristics, it can be seen that tea online shopping consumers are mainly young and middle-aged people aged 20–40, with a monthly disposable income of less than 4000 yuan and have a college degree or above. At the same time, most consumers have more than one year of online shopping experience, but they have only reached a level of familiarity or general familiarity with tea related knowledge. Through observation and analysis of the basic distribution of the samples, it is not difficult to infer that the current online tea shopping consumers are mainly students with higher education, so their disposable income is generally not high. Although they have rich online shopping experience, their understanding of tea knowledge is not sufficient. Therefore, even if they show a high willingness to purchase tea online, they have not fully transformed their actual behavior.

Then, although descriptive analysis can preliminarily determine the reasons for the bias in the willingness and behavior of tea online shopping consumers, it is slightly insufficient as the only measurement standard. Therefore, this article analyzes whether consumers’ willingness and behavior towards online tea shopping are consistent (positive consistent, negative consistent) or inconsistent [45]. The results are shown in

Table 3. Consumers and their subgroups’ willingness and behavioral bias towards online tea shopping.

Individual Consumer Characteristics		Proportion (%)			Average Value	Chi-Square	T-Value
		Negative Consistency	Deviation	Positive Consistency
Gender	Women	24.06	25.94	50.00	1.26	45.417 ***	68.477 ***
	Male	33.18	20.80	46.02	1.13
Age	20 years and under	22.96	25.19	51.85	1.29	354.308 ***	51.160 ***
	21–30 years old	25.61	19.30	55.09	1.29
	31–40 years old	29.87	36.36	33.77	1.04
	41–50 years	40.32	27.42	32.26	0.92
	51–60 years	30.77	34.62	34.61	1.04
Education	Primary and below	33.33	25.00	41.67	1.08	565.761 ***	113.143 ***
	Junior high school	42.86	14.29	42.85	1.00
	Secondary and high school	27.37	31.58	41.05	1.14
	College and Bachelor’s Degree	25.15	22.73	52.12	1.27
	Graduate students and above	29.92	25.20	44.88	1.15
Average income	2000 and below	21.94	23.63	54.43	1.32	187.607 ***	40.527 ***
	2001~4000	31.62	23.53	44.85	1.13
	4001~6000	31.18	27.96	40.86	1.10
	6001~8000	33.90	23.73	42.37	1.08
	More than 8001	26.67	25.00	48.33	1.22
Online shopping age	1 year and less	43.40	28.30	28.30	0.85	98.590 ***	73.621 ***
	1~3 years	32.09	20.15	47.76	1.16
	4~6 years	22.89	29.35	47.76	1.25
	More than 6 years	24.37	21.32	54.31	1.30
Tea knowledge familiarity	Not familiar	31.37	29.41	39.22	1.08	144.239 ***	48.189 ***
	A little familiar	25.51	25.00	49.49	1.24
	General familiarity	25.74	25.00	49.26	1.24
	Quite familiar	22.00	18.00	60.00	1.38
	Very familiar	32.00	12.00	56.00	1.24
Full sample		27.35	24.10	48.55	1.21	-	-

Note: *** represent p < 0.01, respectively, i.e., the analysis results are significant at 10%, 5% and 1% levels (the same below).

. Except for consumers who have a general understanding of tea, the proportion of positive consistency in willingness and behavior is over 60% for all survey subjects in each feature segment, and the vast majority of subgroups have more than 20% of consumers with a deviation in willingness and behavior. Taking the deviation state of consumer subgroups’ willingness and behavior consistency as an example, the results based on chi square test, T-test, and mean willingness and behavior consistency show that overall, the deviation scores of each subgroup are between 0.5 and 1.5, indicating that there is generally a high degree of deviation in willingness and behavior among survey subjects from different groups in this study. Among them, the deviation situation of the female subgroup is more severe than that of the male subgroup; this indicates that women are more likely than men to have intention but no behavior during the online tea shopping process. The deviation of the 31–40 year old subgroup is the most severe, with significant differences between different subgroups. The subgroup of consumers who are very familiar with tea knowledge has the lowest proportion of deviant individuals and shows a trend of decreasing the proportion of deviant individuals and increasing the proportion of positive consistency as familiarity gradually deepens.

By analyzing the differences in the proportion of consumers who deviate from their willingness and behavior to purchase tea online, the following information can be inferred. First, there is a high proportion of deviation among consumers, both overall and among different characteristic subgroups, indicating that there are other interfering factors that have a significant impact on the process of transforming consumers’ willingness to purchase tea online into actual behavior, resulting in weaker predictive power of willingness on behavior. Second, although women are more likely to develop purchasing intentions compared to men, they are also more likely to fall into a choice dilemma due to considering too many factors, resulting in women’s strong purchasing intentions not being converted into purchasing behavior, instead being in a wait-and-see state. Third, the more familiar consumers are with tea, the stronger their views on online tea shopping behavior, and the stronger their willingness to predict behavior. More often, there is only a negative or positive consensus state of buying or not buying, rather than a contradictory state of willingness to buy without taking action.

3.2. Regression Results

We performed unordered multi classification logistic regression analysis on data based on Stata software. Model 1 analyzed the influencing factors of the ratio of negative consistency to positive consistency. Model 2 analyzed the influencing factors of the ratio of deviation to positive consistency. As shown in

Table 4. Influencing factors of consumers’ willingness and behavioral bias in online tea purchasing.

Variant	Items	Model 1		Model 2
		Ratio	Standard Deviation	Ratio	Standard Deviation
Personal characteristic	Gender	0.114	0.237	−0.306	0.251
	Age	0.225 *	0.128	0.107	0.131
	Education	0.210	0.146	0.262 *	0.149
	Monthly average income	−0.018	0.095	−0.073	0.097
	Online shopping age	−0.269 **	0.126	−0.138	0.129
	Tea knowledge familiarity	−0.050	0.095	−0.247 **	0.103
Cultural perceived	Cultural association	−0.283 *	0.158	0.119	0.160
	Cultural Experience	−0.118	0.163	0.574 ***	0.168
Perceived usefulness	Convenience	−0.316 **	0.156	−0.143	0.163
	Product diversification	−0.280 *	0.154	−0.420 ***	0.159
Perceived ease of Use	Price transparency	−0.138	0.130	−0.067	0.139
	Credible information	−0.276 *	0.147	0.078	0.155
Online Shopping Preferences	Price	0.015	0.150	−0.061	0.151
	Quality	0.011	0.154	0.002	0.159
	Origin	−0.129	0.144	0.057	0.154
	Comment	−0.213	0.137	−0.310 **	0.141
	Trand volume	0.072	0.155	0.013	0.159
	Business Services	−0.009	0.152	−0.326 **	0.156
Trust endorsement	Certification labels	0.238	0.151	0.075	0.157
	Brand image	−0.143	0.167	−0.064	0.168
	Awards	−0.275 **	0.130	0.045	0.138
Observed value		585
Log-likelihood		−522.365
LR chi2		184.59
Prob > chi2		0.000 ***
Pseudo R2		0.1502

Note: *, **, *** represent p < 0.10, p < 0.05 and p < 0.01, respectively, i.e., the analysis results are significant at 10%, 5% and 1% levels.

, the overall Log -likelihood value of the model is −522.365, and the LR chi2 value is 145.84, corresponding to a p-value of 0.000 < 0.01. This indicates that the fitting degree is good, and the model is significant and effective. In addition, from Table 4, it can be seen that among the six variables of personal characteristics—cultural perception, perceived usefulness, perceived ease of use, online shopping preferences, and trust endorsement—all sub variables significantly affect the consistency of consumers’ willingness and behavior to purchase tea online. Therefore, the initial assumptions H1~H6 are partially valid, and the analysis results further support and confirm the correctness of the original theory.

In order to further quantify the impact of explanatory variables on consumers’ willingness and behavioral consistency in online tea shopping, this study calculated the marginal effects of each significant variable in the regression model, as shown in

Table 5. Marginal effects of significant variables.

Items	Negative Consistency		Deviation		Positive Consistency
	Efficiency Value	Standard Deviation	Efficiency Value	Standard Deviation	Efficiency Value	Standard Deviation
Age	0.031	0.019	0.002	0.018	−0.033	0.022
Education	0.018	0.022	0.028	0.021	−0.046 *	0.025
Online shopping age	−0.037 *	0.019	−0.004	0.018	0.041 *	0.021
Tea knowledge familiarity	0.008	0.015	−0.037 **	0.015	0.028 *	0.016
Cultural association	−0.057 **	0.024	0.039 *	0.023	0.018	0.026
Cultural Experience	−0.060 **	0.024	0.101 ***	0.023	−0.041	0.027
Convenience	−0.044 *	0.024	−0.001	0.023	0.046 *	0.026
Product diversification	−0.019	0.023	−0.049 **	0.023	0.068 ***	0.026
Reliable information	−0.053 **	0.023	0.032	0.023	0.021	0.024
Comment	−0.015	0.021	−0.036 *	0.020	0.051 **	0.023
Business services	0.021	0.024	−0.052 **	0.023	0.031	0.025
Awards	−0.050 **	0.020	0.026	0.020	0.024	0.022

Note: *, **, *** represent p < 0.10, p < 0.05 and p < 0.01, respectively, i.e., the analysis results are significant at 10%, 5% and 1% levels.

.

According to the analysis results in Table 4 and Table 5, the factors that significantly affect consumers’ willingness to purchase tea online are shown below.

(1)

Personal characteristics: The education level variable in Model 2 is significant at the 10% level, which means that for each level of education improvement, the probability of positive consistency between consumer willingness and behavior decreases by 4.6%. The age variable of online shopping in Model 1 is significant at the 5% level, which means that for every level of increase in online shopping age, the probability of negative consistency in consumer willingness and behavior decreases by 3.7%, while the probability of positive consistency increases by 4.1%. At a level of 5% familiarity with tea, there is a significant negative impact on the bias of consumers’ willingness and behavioral consistency in online tea shopping. This means that for each level of familiarity with tea knowledge, the probability of deviation between willingness and behavior decreases by 3.7%, while the probability of positive consistency increases by 2.8%. Chinese tea can be divided into six main tea categories: green tea; black tea; black tea; white tea; yellow tea; and oolong tea. As a result, there are many standard categories for evaluating the quality of tea, which poses a high cultural knowledge barrier for consumers. This may cause fear for consumers who lack tea knowledge or experience in online tea shopping, but it also creates strong knowledge confidence for consumers who have mastered judgment skills. With the increase in education level, online shopping age, and familiarity with tea knowledge, consumers will gradually form a firm knowledge confidence, believing that they have sufficient information processing and product quality identification abilities in the process of online tea shopping, and will be able to fully enjoy the convenience brought by online tea shopping behavior while also buying satisfactory tea products. Thus, experienced consumers are less likely to be disturbed by other influencing factors in the decision-making process of converting their purchase intention into purchase behavior, resulting in a higher probability of a positive alignment between their willingness and behavior.

(2)

Cultural perception: The cultural association variable in Model 1 is significant at the 10% level, which means that for every level of increase in consumer perception of cultural association, the probability of negative agreement between consumer willingness and behavior decreases by 5.7%, while the probability of deviation increases by 3.9%. The cultural experience variable in Model 2 is significant at the 10% level, which means that for every level of improvement in the tea online shopping process that consumers can feel, the probability of negative consistency between their willingness and behavior decreases by 6%, while the probability of being in a deviant state increases by 10.1%. This indicates that consumers attach great importance to the value of tea cultural attributes. The good cultural and shopping atmosphere created by tea merchants can arouse consumers’ curiosity about tea culture and their desire to purchase tea, but at the same time, high cultural value tea is often accompanied by high prices, which also suppresses consumers’ motivation to put their purchasing intentions into practice. Therefore, although the full display of tea culture on online platforms can only increase the proportion of people who are willing but have no purchasing behavior, it exacerbates the contradiction between consumers’ willingness and behavior.

(3)

Perceived usefulness and perceived ease of use: The convenience variable in Model 1 is significant at the 5% level, which means that for every level of improvement in consumers’ perception of the convenience of online tea shopping, the probability of negative agreement between consumer willingness and behavior decreases by 4.4%, while the probability of positive agreement increases by 4.6%. The variable of product diversification in both Model 1 and Model 2 shows a significant negative bias towards the consistency of consumers’ willingness and behavior in online tea shopping. This means that for each level of satisfaction with product diversification on online shopping platforms, the probability of the deviation between consumer willingness and behavior decreases by 4.9%, while the probability of positive consistency increases by 6.8%. The information reliability variable in Model 1 is significant at the 5% level, which means that for every level of improvement in consumers’ perception of the reliability of information published by tea online stores, the probability of negative consistency between consumer willingness and behavior decreases by 5.3%. In real consumption scenarios, consumers need to wait for tea to be brewed for a long time in tea shops and cannot try as many tea products as possible within the limited shopping time. In addition, online platforms provide consumers with a great experience of one-stop shopping at home without leaving their homes, allowing them to quickly watch prerecorded display images or videos from merchants online. This approach can help consumers save time while also enabling them to choose tea products that they are satisfied with. Therefore, on the premise that consumers believe in the credibility of merchants, the more they believe that purchasing tea online can enjoy better convenience and product diversification, the more willing they will be to purchase tea online. Not only does it increase the proportion of people who have intention and actual purchasing behavior; it also reduces the probability of negative agreement between intention and behavior, while the probability of deviation and positive agreement increases.

(4)

Online shopping preferences: The online evaluation and service attitude variables in Model 2 are significantly negatively correlated with the deviation in the consistency of consumers’ willingness and behavior to purchase tea online. This means that for each level of improvement in consumers’ online evaluation of products and their attention to store service attitude and professionalism, the probability of deviation between consumers’ willingness and behavior will decrease by 3.6% and 5.2%, respectively. At the same time, the level of attention to online evaluations will correspondingly increase the probability of consistent occurrence of willingness and behavior by 6.8%. A possible explanation for this phenomenon is that for consumers, online tea shopping is a pure online experience, and there is no way to directly observe the quality and price comparison of products. It is necessary to understand the product through interaction with others’ information. Therefore, paying attention to other consumers’ online comments on merchants and professional explanations from customer service can promote consistency in consumer willingness and behavior and reduce the probability of deviation.

(5)

Trust endorsement: The variable of award status in Model 1 is significant at the 5% level, which means that for each level of increase in consumer attention to whether tea products in online stores have won awards, the probability of negative consistency between consumer willingness and behavior decreases by 5.0%. Possible reasons include two aspects: First, the award status represents the product quality competitiveness of tea in the industry; the more consumers pay attention to the award status, the more they trust the quality assurance implied behind the award label. Second, as mentioned earlier, tea is a cultural product, and having famous cultural products or tea leaves can cause consumers to conduct impulsive consumption out of vanity, which is essentially the same. Therefore, increasing attention to the award-winning situation will reduce the probability of negative consensus among consumers who are unwilling to purchase tea online.

4. Conclusions and Discussion

This study adopts empirical methods to explore the influencing factors of inconsistent purchase intention and behavior among consumers during the online tea shopping process. Based on the summary and reference of previous studies, a framework for the study of willingness and behavior consistency was constructed, and data were collected through questionnaire surveys. Finally, the influence and marginal effects of each variable on willingness and behavior consistency were calculated through unordered multi classification logistic regression analysis.

The results indicate that the assumptions H1~H6 are partially valid. Among them, consumers’ gender, age, education level, online shopping age, familiarity with tea, feelings of cultural association, cultural experience, convenience, product diversity, and reliable information in online shopping experience, as well as attention to online evaluations, service attitudes, and awards, all significantly increase consumers’ willingness to purchase tea through online channels. The increase in the level of variables such as education level, online shopping age, familiarity with tea knowledge, convenience, product diversification, and online evaluation will increase the positive proportion of consumer willingness and behavior. An increase in the level of variables such as online shopping age, cultural association, cultural experience, convenience, reliable information, and award status will reduce the negative consistency ratio between willingness and behavior. The increasing level of familiarity with tea knowledge, product diversification, online evaluation, and service attitude variables will increase the proportion of deviation between willingness and behavior. Cultural associations and experiences can also increase the proportion of deviations.

First, the conclusion of this article confirms that some of the influencing factors that have been confirmed in previous studies are also applicable in the study of tea e-commerce; for example: (1) young women are more sensitive and impulsive consumers, and under the influence of special consumption scenarios, online shopping behavior is more direct, fast maturing, and frequent [46,47]; (2) the higher the consumers’ expectations for the functionality of e-commerce platforms themselves (such as information reliability, convenience, and product diversification), as well as their level of attention to various types of information on the platform (such as online evaluations and service attitudes), the more likely they are to be willing to purchase products through online channels [13,48,49]. When the above two consumer groups have online shopping intentions, these will transform into online shopping behavior in a shorter period of time, and the probability of their online tea shopping intentions and behavior being consistent is also higher compared to other sub groups within their respective groups.

Second, based on the characteristics of the tea market, this article also draws some differential conclusions that are different from before: (1) For young people or low educated or inexperienced consumer groups, due to cultural barriers in tea knowledge, it is impossible to distinguish the true quality of tea products whether purchased online or offline channels. Therefore, in order to save costs and the need to refer to the evaluations of others, consumers are more willing to purchase tea online, which also leads to the higher willingness and behavior of these consumers compared to the elderly or the consumer group with rich tea knowledge reserves. (2) Cultural attribute-related variables (cultural association, cultural perception, and award status) can reduce the probability of inconsistency between consumer willingness and behavior. This is consistent with the hypothetical logic deduction in this article. Most consumers believe that tea is not only an agricultural product but also a cultural product. The functional attributes of cultural products mainly involve satisfying consumers’ psychological needs such as enjoyment, flaunting, and vanity. Therefore, the more consumers hope to obtain a sense of cultural experience through online tea shopping behavior, the higher their willingness to purchase tea online. The group with consumption intention is the quantitative basis of the group with consumption behavior, so the group of consumers with these characteristics will more likely experience situations where their intention and behavior deviate (with intention but without behavior) or are positively consistent (with intention but with behavior).

5. Suggestions

At present, the agricultural product e-commerce industry is constantly developing, and the tea industry is gradually joining it. Compared to other agricultural products, there is a more prominent issue of inconsistency in willingness and behavior. In the field of online tea purchasing, it is necessary to continuously explore the reasons and introduce reforms to overcome the difficulties. Therefore, based on the research conclusions of this article, the following suggestions are proposed.

First, there is a serious deviation between consumers’ willingness and behavior in online tea shopping. When formulating market strategies in the future, it is not recommended to rely solely on online sales platforms. Instead, a development approach that integrates physical stores and e-commerce should be considered, both online and offline. For example, when fully displaying products on e-commerce platforms, convenient offline experience appointments can be provided, which can save consumers time in selecting products and enhance their shopping experience, achieving complementary marketing effects.

Second, further segment the consumer population, analyze consumer profiles, and identify market positioning. Young people are the main force of tea online shopping consumption, but they basically lack sufficient knowledge reserves. Therefore, in the display of tea products on online platforms, it is not only necessary to label key quality assurance related information; it is also necessary to further display the official explanations and screening methods of the meanings represented by this information. On the one hand, it can help customers save information acquisition costs; and on the other hand, it can effectively transmit product information to consumers, improving marketing efficiency.

Third, focus on creating a sense of atmosphere for consumers to experience tea culture. Through online channels, information with distinct humanistic and natural tea culture concepts such as tea garden landscapes, tea processing, tea knowledge, and tea development history can be displayed on sales platforms in vivid ways such as via videos and VR, creating an immersive feeling for consumers. While enhancing cultural appeal, it promotes consumers’ desire for impulsive consumption in the cultural atmosphere, shortens decision-making time, and improves the efficiency of quickly converting tea online shopping intentions into actual behavior.

Fourth, develop diversified products and improve the transparency of product information on online platforms. Diversified product categories are a distinct feature that sets tea apart from other agricultural products, bringing market segmentation advantages to tea marketing. At the same time, it also brings consumers the drawbacks of high difficulty in quality identification and particularly energy consuming. Therefore, it is recommended to further improve the construction and exposure of product identity information, such as product traceability systems and quality control systems, while maintaining the advantages of e-commerce platforms that can simultaneously showcase diverse products. Every rigorous and comprehensive quality information will enhance consumers’ trust in the brand. Even with the emergence of new tea products, it will significantly reduce consumers’ doubts about the risks of purchasing behavior and promote the conversion of consumers’ willingness to purchase tea online into action.

Fifth, do a good job in controlling online word-of-mouth and maintaining brand advantages. To this end, first, it is necessary to provide good product services. Agricultural products have the characteristics of seasonal differences in quality, difficulty in storage and transportation, and may deteriorate during the express delivery process, which affects the positive evaluation of consumers. Therefore, on the one hand, it is necessary to do a good job in quality control, as ensuring the stability of quality is the basis for reducing negative evaluations; on the other hand, protective measures should be taken during transportation, and if necessary, cold-chain transportation can be adopted to ensure that the product is intact in appearance and of the same quality when delivered to consumers. Second, it is necessary to enhance customer service tea knowledge and professional literacy. The insufficient knowledge reserve of tea consumers in e-commerce channels makes this group have a strong risk perception. If customer service can provide professional and effective answers and solutions to consumers’ problems, it will greatly increase the probability of online positive feedback. Finally, it is important to highlight product advantages in marketing promotion. By showcasing awards or certificates issued by authoritative organizations in industry competitions, authoritative endorsement can be provided for product quality, which is an effective way to reduce consumers’ doubts about product quality. Therefore, it is recommended that businesses actively participate in various industry competitions and apply for well-known recognition labels. Through brand building via the above three methods, businesses can effectively reduce the risk and doubt level of consumers’ online shopping willingness in terms of its conversion into behavioral decision-making, increase the proportion of actual behavior occurrence, and prevent the loss of potential consumers.

6. Research Limitations and Prospects

This article focuses on examining the psychological mechanism of consumers’ online tea shopping at the micro level. Based on the analysis of the results and conclusions, there are still some areas that can be expanded in this study: (1) The perception of tea culture did not significantly increase the probability of a positive consensus situation where willingness is converted into behavior. This may be because willingness to purchase does not incur actual costs, but purchasing behavior does. However, this study did not explore the impact of cultural perception on the consistency of tea online shopping willingness and behavior under different price levels. Continuing to explore this issue in the future will further advance understanding of how cultural perception affects consumers’ tea purchasing decision behavior in different consumption scenarios. (2) As stated in the full text, tea is not only an agricultural product in China but also a cultural product. However, as is well known, no country in the world has as long a history of tea culture as China. In most cases, international tea trade and circulation types consider tea as a bulk commodity. So, if it is necessary to verify whether the theoretical derivation and empirical results of this study are applicable to all countries that consume tea, it is necessary to ensure that the proposed conclusions and sales strategies are not only effective but also applicable within the local market in the constantly changing geographical and cultural background of sample collection.