The Characteristics and Mode Selection of Agricultural Product
E-commerce Distribution Marketing
YUANYUAN CHU
Lyceum of the Philippines University, Manila 1002, PHILIPPINES
Abstract:—In recent years, China’s e-commerce market has become the largest in the world. E-commerce of
agricultural products can solve the problems of many traditional marketing modes and bring great convenience.
To study the e-commerce mode of agricultural products, the secondary indicator factors of the e-commerce
mode of agricultural products are constructed, and the structural equation of 7 hidden factors and 30
observation factors is designed based on the structural equation. On the basis of the above procedures, a
questionnaire on the selection of agricultural products e-commerce mode is designed. Through the calculation
and analysis of the mode, 6 recessive variables and 27 observation indicators are determined. After the analysis
of their relationship, the e-commerce mode selection strategy for agricultural products is finally proposed.
Key-Words: Agriculture products; E-commerce; Marketing characteristics; Mode
Received: March 8, 2022. Revised: November 4, 2022. Accepted: December 4, 2022. Published: December 15, 2022.
1. Introduction
Since China joined the WTO in 2001, the
Chinese economy has faced many pressures and
challenges in the process of keeping pace with the
world economy. E-commerce is characterized by
high efficiency and low consumption, thus it has
developed rapidly in the process of globalization. At
present, China is in a critical period of
comprehensive upgrading and reform, therefore
optimization and reform of industries is imperative.
In this context, the development of agricultural
products e-commerce (Electronic Commerce) needs
to meet the requirements of market development.
China has accelerated the construction of
infrastructure, however, even after the roads are built,
many remote rural areas are difficult to improve the
visibility of local agricultural products and sell them.
E-commerce can promote the agricultural product
market by forming an industrial chain of agricultural
products, thus the simultaneous development of
agricultural product planting, storage, warehousing,
transportation and other fields can be realized.
In 2018, China's online retail trading volume
reached 9 trillion CNY (China Yuan), accounting for
18.4% of the total retail sales of consumer goods, of
which the rural online retail sales reached 1.37
trillion CNY. E-commerce has become an important
way to fight poverty. Under the guidance of the
government and the Ministry of Commerce,
e-commerce marketing mode of agricultural products
has been promoted nationwide, [1-3]. In the "Double
11" event in 2018, the sales volume of rural
e-commerce network was 27.59 billion CNY, and the
main products included nuts, edible oil, milk, etc.
With the continuous development of e-commerce,
various e-commerce platforms and e-commerce
modes emerge one after another. Under this
background, how to choose a suitable e-commerce
mode for agricultural products is an urgent problem
to be solved, [4]. This research aims to solve this
problem. In addition, the research takes e-commerce
as a breakthrough to develop modern and
information agriculture, so as to solve the issues
relating to agriculture, rural areas and farmers in the
process of urbanization.
WSEAS TRANSACTIONS on SYSTEMS and CONTROL
DOI: 10.37394/23203.2022.17.59
Yuanyuan Chu
E-ISSN: 2224-2856
535
Volume 17, 2022
2. Related Works
Global agricultural development is directly
related to global food supply. Ensuring the normal
operation and development of agriculture is the basis
for human survival and social stability. Therefore, a
large number of researchers are conducting research
with the purpose of promoting the development of
agriculture. Zhao J and his team members studied the
preservation strategies of agricultural products and
analyzed it from the physical, chemical and
biological perspectives. They also analyzed the
ordering and storage strategies of fresh agricultural
products under the premise of limiting carbon
emissions. They believed that in the context of the
digital age, appropriate e-commerce models were
needed to realize the common interests of consumers,
farmers and dealers, [5]. At present, the development
of e-commerce is very fast, and many industries are
trying to conduct transactions in the form of
e-commerce. As an industry relying on cross regional
sales, agriculture has also received a lot of research
in recent years focusing on the combination of
agricultural product sales and e-commerce. You J
with his team studied the cross-border e-commerce
transactions of agricultural products in Sichuan, and
believed that the long delivery time and high cost of
cross-border trade hindered the development of
e-commerce of agricultural products in this region.
According to the research results, the current
cross-border e-commerce trade of agricultural
products in the region uses a single logistics method,
and expanding new logistics methods may improve
the problem, [6]. Liao et al. studied the theme of
using key opinion leaders to promote online sales of
agricultural products. They established an evolution
model of sales promotion strategy and verified the
effectiveness of the model through simulation, [7].
Liu and Kao's team studied the customer satisfaction
of characteristic agricultural products in specific
regions in online platform sales. They established the
influencing factor model and verified the hypothesis
of influencing factors. Finally, on the basis of this
model, the author proposed a strategy to improve the
online sales satisfaction of local agricultural products,
which promoted the development of e-commerce of
local agricultural products, [8]. Sun and Lei analyzed
the development status of a pepper agricultural
product, and summarized the problems in the
development of this agricultural product in terms of
multiple influencing factors, including brand effect,
natural environment factors, e-commerce platform
factors, e-commerce infrastructure factors, etc.
According to the analysis results, an effective growth
strategy for online sales was proposed, [9].
By analyzing the relevant literature on the
combination of e-commerce and agricultural
products, it is found that the research in this field is
gradually getting more attention. However, a large
number of studies are limited to specific regions or
agricultural products, leading to the lack of
universality. Therefore, this research attempts to
build an e-commerce mode that can be applied to
various regions and agricultural products by means
of influencing factor analysis.
3. Research Design
Performance indicators of agricultural
e-commerce modes are analyzed, and these
indicators are used to evaluate and decide the choice
of e-commerce mode. Factors influencing
agricultural e-commerce modes are classified as
internal and external variables and analyzed
according to different attributes. The variables are
analyzed with structural equation modeling. This
model is also known as latent variable model or
covariance structure model. The model enables the
construction of hypothetical models based on
assumptions and validates the correctness of the
assumptions by collecting and analyzing variable
data. The hypotheses that are not correct enough are
modified based on the validation results to improve
the model performance. The validation method is to
test the difference between the true covariance of the
sample and the theoretical covariance of the model,
and the smaller the difference, the closer the model is
to the actual situation. After the correct model is
established, the correct e-commerce model for
agricultural products is selected and suggested based
on the model analysis results.
WSEAS TRANSACTIONS on SYSTEMS and CONTROL
DOI: 10.37394/23203.2022.17.59
Yuanyuan Chu
E-ISSN: 2224-2856
536
Volume 17, 2022
4. Agricultural Products E-commerce
Mode
E-commerce mode is a modern marketing mode
combining Internet technology and modern
marketing concept, which combines the operation
and distribution of the industrial chain. There are
many classifications of e-commerce modes.
E-commerce modes can be classified into B2B
(Business to Business), B2C (Business to Consumer)
and C2C (Consumer to Consumer) according to the
nature of e-commerce economic entities. It can also
be classified from other aspects, [10]. The
e-commerce mode of agricultural products is a
marketing mode of agricultural products based on the
conventional e-commerce mode. It is a new mode
based on efficiency and cost benefit distribution. In
this mode, the efficiency of e-commerce is evaluated
by time indicators, the benefits are evaluated by the
income of the business entity, the responsibilities are
evaluated by market risk, and the costs are evaluated
by profit space and profit rate, [11-12]. The design of
e-commerce mode for agricultural products takes
into account the advantages of economic entities and
the pattern of interest relations, reflecting the interest
distribution relationship among various economies,
as shown in Figure 1.
E-commerce mode of
agricultural products
Organizational composition of
economic entities Relationship pattern among
economic entities
Advantages of
single entity
Business
distribution
among entities
Benefits
distribution
among entities
Functions of a
single entity
Figure 1 The relationship among economies in the
agricultural E-commerce mode
When choosing the e-commerce mode of
agricultural products, it is necessary to judge
according to the main performance indicators of the
mode, and select the appropriate mode according to
its results. The main performance indicators are
shown in Figure 2.
Main
performance
indicators
Cost
Profit
Responsib
ility
Distributi
on
Efficiency Time
Balance
between
risk and
profit
Profit
margin
and scale
Key
performance
indicators
Judgment
basis
Figure 2 Performance indicators of agricultural
products E-commerce mode
5. Analysis on the Constituent
Factors of Agricultural Product
E-commerce Mode
Due to the different environment and
characteristics of different regions, the e-commerce
operation modes required by different regions are
also quite different. The corresponding cost control,
operation mode and profit source are quite different
as well, so there are many factors affecting the
operation mode of e-commerce, [13]. Through the
analysis of the characteristics of agricultural products
e-commerce, several main influencing factors are
sorted out, namely regional environment, platform,
product, scale, capital and risk.
Regional environmental factors have the most
significant impact on the choice of agricultural
product e-commerce modes. There are great
differences in consumption habits, consumption
levels, logistics networks and consumption concepts
in different regions. Even for the same business
entity or the same product, the agricultural product
operator still needs to choose different e-commerce
modes adapted to local conditions according to
different regions. With the development of modern
Internet technology, various e-commerce platforms
have been accepted by people, and platforms like
Taobao, JD (Jingdong), etc. have achieved great
success. These platforms can realize the large-scale
operation of outlets at relatively low cost. They can
obtain more profits through brand building and
WSEAS TRANSACTIONS on SYSTEMS and CONTROL
DOI: 10.37394/23203.2022.17.59
Yuanyuan Chu
E-ISSN: 2224-2856
537
Volume 17, 2022
promotion marketing. Different e-commerce
platforms have different service levels and different
operation modes, which will affect the business
operation of agricultural product operators. Therefore,
platform selection should also be considered as a
factor when selecting e-commerce modes.
Agricultural products have their uniqueness,
especially fresh products, which have higher
requirements for logistics transportation and storage.
The logistics costs of different products are different,
thus there are certain differences when choosing the
e-commerce operation mode. Most of the agricultural
products of e-commerce business entities are planted
for off-site sales, and through e-commerce platforms
they can quickly complete the recovery of funds and
obtain profits. Most farmers’ capital and risk
tolerance are limited, which makes it difficult for
them to operate with large-scale companies and
enterprises. The normal operation of an e-commerce
platform requires sufficient technical and
management capabilities as well as capital
investment, so capital and risk factors must be
considered. The planting and sales of agricultural
products are characterized by scale and
intensification. Only scale operation can
fundamentally reduce marginal cost, improve
operational efficiency and obtain more product
income.
6. Assumption and Construction of
Agricultural Products E-commerce
Mode
A. Assumptions affecting the choice of
agricultural products E-commerce mode
According to the constituent factors of
agricultural product e-commerce mode, we can
analyze the factors that affect the choice of
agricultural product e-commerce mode, for which the
following assumptions are proposed:
Assumption 1: Consumer experience has an
impact on the choice of agricultural product
e-commerce mode. A good agricultural product
e-commerce mode can bring a good experience to
consumers. An e-commerce mode is feasible only
when consumers have a good experience. Consumer
experience is related to multiple factors of
agricultural products. The quality, price, reputation,
logistics of agricultural products and other factors
together constitute the influencing factors of
consumer experience.
Assumption 2: Partner satisfaction has an
impact on the choice of agricultural product
e-commerce mode. A good e-commerce mode must
be able to give consideration to all aspects of the
business chain and meet the demands of different
partners. There are many factors that determine
partner satisfaction, including profit, operating cost,
risk, etc.
Assumption 3: Government satisfaction has an
impact on the choice of agricultural product
e-commerce mode. The development of e-commerce
can not be separated from the support of the
government. A good e-commerce mode should be
legal and can pay tax contributions to the
government. Therefore, government satisfaction will
determine the choice of agricultural product
e-commerce mode to a certain extent.
Assumption 4: The main characteristics of
agricultural products have an impact on the choice of
agricultural e-commerce mode. Agricultural products
may have their own unique product characteristics,
such as origin, scale, branding, production cycle, etc.
These characteristics will have an impact on the
marketing mode and process of agricultural products
e-commerce.
Assumption 5: The convenience of the
marketing subject has an impact on the choice of
agricultural product e-commerce mode. For the
marketing subject of agricultural products, different
e-commerce modes mean different costs and effects.
A convenient mode for the marketing subject can
effectively reduce costs and improve efficiency.
Assumption 6: The cooperation degree of
circulation subjects has an impact on the choice of
agricultural product e-commerce mode. The
marketing mode of e-commerce can give full play to
the characteristics of all people in the business chain,
which reflects a multi-party cooperation. Therefore,
WSEAS TRANSACTIONS on SYSTEMS and CONTROL
DOI: 10.37394/23203.2022.17.59
Yuanyuan Chu
E-ISSN: 2224-2856
538
Volume 17, 2022
factors such as circulation efficiency and spillover
benefits of circulation subjects have an impact.
Assumption 7: The performance of e-commerce
platforms has an impact on the choice of agricultural
products e-commerce mode. There are many
e-commerce platforms that can be chosen, and
different e-commerce modes have different
performance requirements for e-commerce platforms.
The performance of the e-commerce platform
includes the click through rate, conversion rate and
technical quality standards of the platform website.
Combining these assumptions, the main latent
variables of agricultural product e-commerce mode
can be obtained. Three exogenous latent variables
and 12 corresponding measurable variables are
obtained, in addition, there are four endogenous
latent variables and 18 corresponding measurable
variables.
B. Model construction
Structural equation model can deal with a lot of
dependent variables and has the advantage of good
fitting. The number of Latent variables to be
analyzed in this study is 7, so structural equation
model is selected. In order to clearly describe the
interrelationship between various factors, the
knowledge of marketing management discipline and
the characteristics of agricultural product sales are
analyzed together, and a structural equation diagram
is constructed, as shown in Figure 3.
Operation difficulty
E-commerce platform standards
and specifications
Website click-through rate and
conversion rate
Technical capability of e-
commerce platform
Cooperative spillover effect
Convenience of Collaboration
Circulation efficiency
Efficiency of information
transmission
Marketing cost
Marketing implementation
difficulty
Marketing effect
Lead time
Distribution distance of
production and marketing places
Product features are distinctive or
not
Production scale
Branding level
Packaging and storage
requirements
Performance of e-
commerce platform
Convenience of
marketing subject
Main attributes of
agricultural
products
Circulation
subject
cooperation
Consumer
Experience
Price of agricultural products
Waiting time of logistics
Service
Reputation
Product quality
Profit
Balance between risk and profit
Whether to give full play to their
respective business abilities
Operating capital
ε1Content design of e-commerce
platform website
ε2
ε3
ε4
ε5
ε6
ε7
ε8
ε9
ε10
ε11
ε12
ε13
ε14
ε15
ε16
ε17
ε18
ζ1
ζ2
ζ3
ζ4
δ1
Government
support
Satisfaction of
partners
Conforms to consumption habits
or not
Is it morally and legally accepted
Tax promotion
δ2
δ3
δ4
δ5
δ6
δ7
δ9
δ8
δ10
δ11
δ12
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
Figure 3 Structural equation of influencing Factors of Agricultural Product E-commerce model
According to the specific characteristics of
agricultural products e-commerce, relevant variables
are determined, as shown in Table 1.
WSEAS TRANSACTIONS on SYSTEMS and CONTROL
DOI: 10.37394/23203.2022.17.59
Yuanyuan Chu
E-ISSN: 2224-2856
539
Volume 17, 2022
Table 1 Variable definition table
Endogenous variables
External variables
Variables
Explanation
Variables
Explanation
η1
Performance of e-commerce
platform
ξ1
Consumer Experience
η2
Circulation subject cooperation
ξ2
Satisfaction of partners
η3
Convenience of marketing subject
ξ3
Government support
η4
Main attributes of agricultural
products
Y1
Content of e-commerce platform
website
X1
Price of agricultural products
Y2
E-commerce platform standards
and specifications
X2
Waiting time of logistics
Y3
Operation difficulty
X3
Product quality
Y4
Website click-through rate and
conversion rate
X4
Service
Y5
Technical capability of
e-commerce platform
X5
Reputation
Y6
Circulation efficiency
X6
Conforms to consumption habits or
not
Y7
Convenience of Collaboration
X7
Profit
Y8
Cooperative spillover effect
X8
Balance between risk and profit
Y9
Efficiency of information
transmission
X9
Whether to give full play to their
respective business abilities
Y10
Marketing cost
X10
Operating capital
Y11
Marketing implementation
difficulty
X11
Is it morally and legally accepted
Y12
Marketing effect
X12
Tax promotion
Y13
Lead time
Y14
Distribution distance of production
and marketing places
Y15
Product features
Y16
Production scale
Y17
Branding level
Y18
Packaging and storage
requirements
Circulation efficiency regards the e-commerce
model of agricultural products as an intermediary
variable, and sets the vectors composed of external
indicators and endogenous indicators as
x
and
y
respectively. Assume that the endogenous latent
variable vector is
, the factor load matrix of the
endogenous indicator is
y
, the external latent
variable vector is
, and the factor load matrix of
the external indicator is
x
. Then let
,

be
residual terms reflect the unexplained part of the
equation, [14].
The measurement equation between indicators
WSEAS TRANSACTIONS on SYSTEMS and CONTROL
DOI: 10.37394/23203.2022.17.59
Yuanyuan Chu
E-ISSN: 2224-2856
540
Volume 17, 2022
and latent variables can be obtained as follows:
x
x
(1)
y
y
(2)
Assuming that the influence of external latent
variables on endogenous latent variables is
, the
relationship between endogenous latent variables is
B
, and the residual term of structural equation is
,
then the structural equation can be obtained as
follows:
B
(3)
7. Data Analysis
7.1 Analysis of data reliability and validity
The questionnaire was designed according to
the assumptions and corresponding indicator
variables mentioned above, and the corresponding
questionnaire survey was carried out from April 2019
to July 2019. The survey was conducted in the form
of online e-mails. Totally 100 questionnaires were
distributed, 70 valid questionnaires were recovered,
and the availability rate was 70%. SPSS (Statistical
Product Service Solutions) software was used to
analyze the reliability and validity of the data. The
specific results are shown in Table 2 and Table 3.
Table 2 Reliability analysis
Cronbach's Alpha
N of Items
0.898
30
In general, Cronbach's Alpha reliability
coefficient of 0.8 means very good reliability, and
generally it is not less than 0.6. As can be seen from
the data in Table 2, α= 0.898>0.8, indicating that the
reliability analysis is effective.
Table 3 Validity analysis
Test result of Kaiser-Meyer-Olkin (KMO)
0.698
Sphericity test
Chi square value
1098.042
df
435
Sig.
0
It can be seen from Table 3 that the KMO and
Bartlett sphericity test values are 0.698>0.5,
indicating that the factor analysis has good validity.
It can be further seen that P=0.000<0.001, indicating
that the correlation coefficient matrix can explain
most of the variance and has good validity.
7.2 Fit analysis of the model
When using the statistical data of the
questionnaire survey for analysis, the original
assumption mode needs to be partly adjusted to meet
the actual situation. The revised structure diagram is
as follows in Figure 4.
WSEAS TRANSACTIONS on SYSTEMS and CONTROL
DOI: 10.37394/23203.2022.17.59
Yuanyuan Chu
E-ISSN: 2224-2856
541
Volume 17, 2022
Operation difficulty
E-commerce platform standards
and specifications
Technical capability of e-
commerce platform
Cooperative spillover effect
Convenience of Collaboration
Circulation efficiency
Efficiency of information
transmission
Marketing cost
Marketing implementation
difficulty
Marketing effect
Lead time
Distribution distance of
production and marketing places
Product features are distinctive or
not
Production scale
Branding level
Packaging and storage
requirements
Performance of e-
commerce platform
Convenience of
marketing subject
Main attributes of
agricultural
products
Circulation
subject
cooperation
Consumer
Experience
Price of agricultural products
Waiting time of logistics
Service
Reputation
Product quality
Profit
Balance between risk and profit
Whether to give full play to their
respective business abilities
Operating capital
ε1Content design of e-commerce
platform website
ε2
ε3
ε5
ε6
ε7
ε8
ε9
ε10
ε11
ε12
ε13
ε14
ε15
ε16
ε17
ε18
ζ1
ζ2
ζ3
ζ4
δ1
Satisfaction of
partners
δ2
δ3
δ4
δ5
δ7
δ9
δ8
δ10
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
Figure 4
Modified structural equation diagram
The results of the fit analysis of the assumption model are shown in Table 4.
Table 4 Fit analysis of the model
Indicator
Indicator value
Indicator
Indicator value
DF
(Degree of freedom)
346.000
IFI (Incremental fit index)
0.881
X2 (
Chi-Square Statistic)
650.587
GFI (Goodness-of-fit index)
0.877
P (P-Value)
0.000
AGFI (Adjusted goodness-of-fit index)
0.887
NFI (Normative fit
index)
0.858
RFI (Relative fitting index)
0.882
NNFI (Non-normed fit
index)
0.884
RMR (Root mean square residual)
0.026
CFI (Comparative fit
index)
0.901
RMSEA (Root-mean-square error of
approximation)
0.076
IFI, RMSEA, GFI, AGFI and RMR in Table 4
are all absolute fitting indexes, which are ideal
structural equation model evaluation indexes. NFI,
NNFI, CFI and RFI are value-added fitting indexes,
WSEAS TRANSACTIONS on SYSTEMS and CONTROL
DOI: 10.37394/23203.2022.17.59
Yuanyuan Chu
E-ISSN: 2224-2856
542
Volume 17, 2022
which are widely used structural equation model
evaluation indicators. The larger the value, the better
the fitting. X2 and P are traditional statistical
significance evaluation indicators, which are applied
in all data statistics. DF is the difference between the
amount of information provided by the sample data
and the number of parameters to be estimated. In
general, the X2/DF ratio is 2:1-3:1, X2/DF of this
model is 2.06, P=0.000<0.1. According to the above
conditions, combined with the fitting conditions of
the structural equation model, this result can be
considered acceptable, [15]. Further analysis of other
fitness analysis indicators shows that RMR<0.035,
RMSEA<0.08. Besides, NFI, CFI and other
indicators are less than 0.08, indicating that the
model and data fit well. Among the 6 variables and
27 measurement items in the model, some indicators
cannot be less than 0.9, but considering the sample
size, they are considered acceptable. The above
results show that the hypothetical model is available
for fitting.
7.3 Model analysis results
Table 5 and Table 6 are the analysis results of
the model. In Table 5, considering the limited
number of samples, the confidence interval of
P<0.05 is set as the criterion of significance. ***
indicates passing the inspection. The standard
deviation, critical ratio, nonstandard estimate and
significance P value of all data are included in the
Regression Weights.
Table 5 Regression weights of variables
C.R.
S.E.
Estimate
P
η4
←--
ξ2
2.789
.197
.551
***
η4
←--
ξ1
2.346
.204
.483
***
η3
←--
ξ2
2.107
.204
.431
***
η2
←--
ξ2
3.913
.229
.897
***
η2
←--
ξ1
2.722
.198
.539
***
η3
←--
ξ1
1.984
.115
.182
***
η1
←--
ξ1
2.384
.271
.646
***
η1
←--
ξ2
1.961
.135
.258
***
y6
←--
η2
3.736
.255
.953
***
y8
←--
η2
1.000
y7
←--
η2
4.447
.238
1.059
***
y9
←--
η3
1.000
y12
←--
η3
2.019
.900
1.817
***
y11
←--
η3
1.962
.414
.482
***
y10
←--
η3
2.104
1.230
2.588
***
y15
←--
η4
2.861
.457
1.308
***
y16
←--
η4
2.169
.405
.879
***
y18
←--
η4
2.485
.390
.970
***
y5
←--
η1
2.240
.461
1.033
***
y3
←--
η1
2.666
.601
1.602
***
y2
←--
η1
2.338
.473
1.106
***
y1
←--
η1
1.000
y13
←--
η4
1.000
y14
←--
η4
1.961
.3294
.501
***
x9
←--
ξ2
1.983
.200
.378
***
WSEAS TRANSACTIONS on SYSTEMS and CONTROL
DOI: 10.37394/23203.2022.17.59
Yuanyuan Chu
E-ISSN: 2224-2856
543
Volume 17, 2022
x10
←--
ξ2
3.687
.228
.841
***
x3
←--
ξ1
3.363
.341
1.146
***
x4
←--
ξ1
3.187
.165
.526
***
x5
←--
ξ1
3.1417
.352
1.204
***
x8
←--
ξ2
3.557
.256
.909
***
x7
←--
ξ2
1.000
x17
←--
η4
1.964
.325
.502
***
x2
←--
ξ1
2.851
.304
.867
***
x1
←--
ξ1
1.000
η4
←--
ξ2
2.789
.197
.551
***
When the number of samples is too small, some
data in Table 6 will be close to 1. If the number of
samples is sufficient and the value is greater than 1,
it means that the latent variable of the structural
equation has collinearity. Although there is no case
greater than 1 in this table. The number of samples
may be too small to explain the structural equation.
Table 6 Standard regression weights of each variable
Estimate
η4
←--
ξ2
.942
η4
←--
ξ1
.695
η3
←--
ξ2
.984
η2
←--
ξ2
.956
η2
←--
ξ1
.530
η3
←--
ξ1
.503
η1
←--
ξ1
.806
η1
←--
ξ2
.382
y6
←--
η2
.477
y8
←--
η2
.543
y7
←--
η2
.613
y9
←--
η3
.178
y12
←--
η3
.328
y11
←--
η3
.103
y10
←--
η3
.424
y15
←--
η4
.501
y16
←--
η4
.291
y18
←--
η4
.367
y5
←--
η1
.408
y3
←--
η1
.590
y2
←--
η1
.441
y1
←--
η1
.405
y13
←--
η4
.346
y14
←--
η4
.207
x9
←--
ξ2
.257
x10
←--
ξ2
.564
x3
←--
ξ1
.630
WSEAS TRANSACTIONS on SYSTEMS and CONTROL
DOI: 10.37394/23203.2022.17.59
Yuanyuan Chu
E-ISSN: 2224-2856
544
Volume 17, 2022
x4
←--
ξ1
.570
x5
←--
ξ1
.652
x8
←--
ξ2
.537
x7
←--
ξ2
.579
x17
←--
η4
.184
x2
←--
ξ1
.476
x1
←--
ξ1
.510
η4
←--
ξ2
.942
After visualizing the weight data in Table 5, a
structural equation diagram marked with regression
parameters is obtained, as shown in Figure 5.
Operation difficulty
E-commerce platform standards
and specifications
Technical capability of e-
commerce platform
Cooperative spillover effect
Convenience of Collaboration
Circulation efficiency
Efficiency of information
transmission
Marketing cost
Marketing implementation
difficulty
Marketing effect
Lead time
Distribution distance of
production and marketing places
Product features are distinctive or
not
Production scale
Branding level
Packaging and storage
requirements
Performance of e-
commerce platform
Convenience of
marketing subject
Main attributes of
agricultural products
Circulation
subject
cooperation
Consumer
Experience
Price of agricultural products
Waiting time of logistics
Service
Reputation
Product quality
Profit
Balance between risk and profit
Whether to give full play to their
respective business abilities
Operating capital
ε1Content design of e-commerce
platform website
ε2
ε3
ε5
ε6
ε7
ε8
ε9
ε10
ε11
ε12
ε13
ε14
ε15
ε16
ε17
ε18
ζ1
ζ2
ζ3
ζ4
δ1
Satisfaction of
partners
δ2
δ3
δ4
δ5
δ7
δ9
δ8
δ10
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1.00
1.00
.94
1.05
.76
.59
.98
1.22
1.22
.57
.10
1.09
.82
.78
.23
1.06
.89
1.00
1.11
1.60
1.03
1.00
1.06
.95
1.00
2.59
.48
1.82
1.00
.50
1.31
.88
.50
.37
.04
.12
.05
.05
.65
.54
.26
.18
.90 .55
.48
.55
.48
.43
1.00 .87
1.15
.53
1.20
1.00
.91
.38
.84
.87
.79
.61
.13
.80
.56
.88
.87
.66
Figure 5
Structural equation diagram with regression parameters
It can be seen from the above analysis that
excluding the latent variable of government support
has little impact on the model. This is mainly
because China is still in the primary stage of
e-commerce development, and e-commerce
development is just starting, thus the government
lacks strictness and experience in supervision. The
reason why consumers' habits are excluded from the
consumer experience is that it is found that there are
two extreme situations of this indicator in the
questionnaire survey, either the impact is minimal or
it is very important, which will result in excessive
WSEAS TRANSACTIONS on SYSTEMS and CONTROL
DOI: 10.37394/23203.2022.17.59
Yuanyuan Chu
E-ISSN: 2224-2856
545
Volume 17, 2022
variance. From the perspective of consumers, their
consumption behavior may be out of consumption
habits, or it may be to try new things. This kind of
consumption behavior has a certain randomness, so it
does not have a great impact on the choice of models.
Further analysis of Latent variable influencing
factors in the main performance of e-commerce
platforms shows that there is a significant correlation
between click-through rate, conversion rate and
marketing effect, which makes the equation
collinearity enhanced. Therefore, this influencing
factor is eliminated. Considering that the correlation
between consumer experience and satisfaction of
partner business entities is small, the corresponding
correlation analysis is also ignored.
The load coefficient of Satisfaction of partners
on the main attributes of agricultural products is 0.55,
and the load coefficient on the degree of
collaboration is 0.9, indicating that its impact is
significant. The collaborative load coefficient of
consumer experience and Circulation subject
cooperation is 0.65, and the collaborative load
coefficient of e-commerce platform and Circulation
subject cooperation is 0.54, which shows that they
are closely related. The relationship load between
consumer experience and Operation difficulty is 1.6,
and the relationship load between consumer
experience and Convenience of marketing subject is
1.06, which fully demonstrates that consumer
experience has high requirements for e-commerce
platform operation and efficiency.In addition, the
load coefficient of partner satisfaction and profit is 1,
the load coefficient of reputation is 1.2, and the load
coefficient of product quality is 1.15. Marketing cost
has a great impact on the convenience of the
marketing subject. The corresponding load
coefficient is 2.59, followed by the load coefficient
of marketing effect, which is 1.52.
According to the analysis, there are six
measurable main influencing factors that affect the
choice of agricultural product e-commerce model,
including marketing cost, profit, reputation, main
attributes of agricultural products, the convenience of
circulation subject, and the operation difficulty. After
removing some factors that are less relevant, the five
hidden factors are confirmed, namely, circulation
subject cooperation, the attributes of agricultural
products, consumer experience, satisfaction of
partners and the performance of e-commerce
platform subjects.
8. Suggestions on the Selection of
E-commerce Mode for Agricultural
Products
It can be seen from the above analysis that there
are many influencing factors that need to be
considered when choosing the e-commerce mode for
agricultural products. In practical application, the
circulation subject cooperation and the satisfaction of
the partners should be considered from a macro
perspective. The e-commerce platform is a platform
that shares resources and can achieve mutual benefit.
Five negative attributes were determined, including
consumer experience, partner satisfaction, etc. The
quality of agricultural products and the reputation of
enterprises have a great impact on the experience of
consumers. Especially in the e-commerce mode,
more attention should be paid to the reputation of
enterprises and branding of products, and stricter
quality standards should be implemented, [16-18].
In order to realize the efficient operation of the
agricultural product e-commerce chain, it is
necessary to fully consider the profits of partners and
make all participants have significant benefits. A
mode that takes into account both risks and benefits
can enable them to actively participate in the
operation of agricultural products. When choosing
e-commerce platforms, the difficulty of operation is a
very important indicator. As online shoppers and
marketing agents may come from rural areas, their
computer operating skills may be poor. Therefore,
the e-commerce platform should be designed
according to the preferences and computer skills of
users, so that various users can have a better
experience in the operation process, [19]. Lead time
also has a direct impact on the sales mode of
agricultural products. Order farming mode and cycle
purchase mode can be adopted.
WSEAS TRANSACTIONS on SYSTEMS and CONTROL
DOI: 10.37394/23203.2022.17.59
Yuanyuan Chu
E-ISSN: 2224-2856
546
Volume 17, 2022
Here, China is analyzed as a representative
country. China has a large land area and relatively
developed agriculture, so the marketing of
agricultural products is very important in that country.
The current e-commerce platform and the
corresponding distribution system in China are quite
developed, which means that the performance of the
main e-commerce platform, the degree of
collaboration of the distribution body and the ease of
marketing are high. However, there is currently a
problem of low price competition among Chinese
agricultural e-merchants, which undermines the
profitability and development potential of this market,
and the negative effects of low price competition
need to be avoided by improving the e-commerce
model.
9. Conclusion
This study explored the influencing factors of
agricultural e-commerce mode selection and
established an influencing factor model through SEM.
The results of the study identified five latent
variables and their corresponding observed variables,
and suggestions for the selection of agricultural
products e-commerce models were made based on
the modeling results. The agricultural products
e-commerce mode should fully consider the benefits
and risks of all participants, and fully consider
customers' operational capabilities and shopping
experience. In addition, the agricultural products
e-commerce mode should fully consider the
characteristics of the goods. For this point, other
studies in this field have reached the same conclusion,
[20]. This study has achieved successful results, but
due to the small sample size, there is still room to
improve the accuracy of the final results when
conducting statistical analysis. In the follow-up work,
the sample size can be further increased to obtain a
more realistic agricultural products e-commerce
model.
Reference
[1] Li X, Zhu B. Influence of environmental
regulation on cross-border ecommerce export of
agricultural products. Journal of Environmental
Protection and Ecology, Vol.22, No.3,2021.
pp.1347-1357.
[2] Ning X. Financing decisions analysis of
e-commerce supply chain finance for fresh
agricultural products in China. Journal of
Finance Research, 2021, Vol.5, No.2, 2021.
pp.1-5.
[3] Wang N. Analysis on regional development
model of rural E-commerce. Asian Agricultural
Research, Vol.11, No.7, 2019, pp.5-7.
[4] Xia T, Minghua G U, Liang J, et al. Reflections
on the high-quality development of rural areas in
Guangshan County, Xinyang City. Asian
Agricultural Research, Vol.14, No.6, 2022,
pp.12-14.
[5] Zhao J, Chen L, Zhang P. Discussion on fresh
agricultural products ordering and preservation
strategy under new normal carbon emission
constraint. Plant Diseases and Pests, Vol. 12,
No.4, 2021, pp.33-36.
[6] You J, Chen H. A study on establishing a
cross-border E-commerce logistics system for
agricultural products in Sichuan Province in the
context of the belt and road initiative.
Contemporary Social Sciences, Vol.6, No.5,
2021, pp.13-28.
[7] Liao M J, Zhang J, Wang R M, et al. Simulation
research on online marketing strategies of
branded agricultural products based on the
difference in opinion leader attitudes.
Information Processing in Agriculture, Vol.8,
No.4, 2021, pp.528-536.
[8] Liu X, Kao Z. Research on influencing factors of
customer satisfaction of e-commerce of
characteristic agricultural products. Procedia
Computer Science, Vol.199, 2022 pp.1505-1512.
[9] Sun J, Lei H. Research on the development of
agricultural products under the background of
"Internet +". IOP Conference Series: Earth and
Environmental Science, Vol.571,No.1, 2020,
pp1-6.
WSEAS TRANSACTIONS on SYSTEMS and CONTROL
DOI: 10.37394/23203.2022.17.59
Yuanyuan Chu
E-ISSN: 2224-2856
547
Volume 17, 2022
[10] Zeng Y, Song S, Peng W. Optimal add-on items
recommendation service strength strategy for
e-commerce platform with
full-reduction-promotion. RAIRO - Operations
Research, 2022, Vol.56, No.2, 2022,
pp.1031-1049.
[11] Peng C, Biao M A, Zhang C. Poverty alleviation
through e-commerce: Village involvement and
demonstration policies in rural China. Journal of
Integrative Agriculture, No.4, 2021,
pp.998-1011.
[12] Cao S, Yujun H U, Peijun F U. Present situation
and problems of application of new media in
rural E-commerce: A case study of Anhui
Province. Asian Agricultural Research, Vol.13,
No.10, 2021, pp.12-17.
[13] Yang H, Ju H E. The upward issue of agricultural
products based on farmer cooperatives in the
context of internet. Asian Agricultural Research,
Vol.12, No.3, 2020, pp.6-9.
[14] Ma G X, Zhu L, Lin T R, et al. Multilevel
Pathways of Colorectal Cancer Screening
Among Low-Income Vietnamese Americans: A
Structural Equation Modeling Analysis. Cancer
Control, Vol.28, No.10, 2021, pp.48-84.
[15] Mcneish D. Specifying Location-Scale Models
for Heterogeneous Variances as Multilevel
SEMs. Organizational Research Methods,
Vol.24, No.3, 2020, pp.630-653.
[16] Shi G. Enterprise management innovation and
organizational model investigation based on the
E-commerce environment. Proceedings of
Business and Economics Studies, Vol.3, No.4,
2020, pp.32-34.
[17] Zhao J, Chen L, Zhang P. Analysis on
development characteristics of agricultural
product processing industry under the new
normal. Plant Diseases and Pests, 2021, Vol.12,
No.5, 2021, pp.35-38.
[18] Hai-Ying G U, Wang C W. Impacts of the
COVID-19 pandemic on vegetable production
and countermeasures from an agricultural
insurance perspective. Journal of Integrative
Agriculture, No.12, 2020, pp.2866-2876.
[19] Peng C, Biao M A, Zhang C. Poverty alleviation
through e-commerce: Village involvement and
demonstration policies in rural China.
Agricultural Sciences in China, Vol.20, No.4,
2021, pp.998-1011.
[20] Wyckhuys K, Lu Y, Zhou W, et al. Ecological
pest control fortifies agricultural growth in
Asia-Pacific economies. Nature Ecology &
Evolution, Vol.41, No.11, 2020, pp.1567-1567.
Creative Commons Attribution License 4.0
(Attribution 4.0 International, CC BY 4.0)
This article is published under the terms of the Creative
Commons Attribution License 4.0
https://creativecommons.org/licenses/by/4.0/deed.en_US
WSEAS TRANSACTIONS on SYSTEMS and CONTROL
DOI: 10.37394/23203.2022.17.59
Yuanyuan Chu
E-ISSN: 2224-2856
548
Volume 17, 2022