Developing Platform Supply Chain Contract Coordination and a Numerical Analysis Considering Fresh-Keeping Services

Abstract

With changes in demand and the emergence of new distribution channels, consumer-centric buyer’s markets for many products have been formed. The platform supply chain has been continuously optimized and upgraded. Supply chain leaders have moved downstream to the end of the supply chain. The operational value has been further enhanced. The corresponding systematic construction of the platform supply chain has become an important driving force for future development. The model in this paper is different from the traditional supply chain contract model, which mainly focuses on suppliers or demand. In order to meet the requirements of fresh-keeping services and the goal of revenue sharing, we integrate the production and circulation characteristics of fresh produce into the design of a contract model. In this paper, a revenue-sharing contract model of the fresh produce supply chain is constructed based on the core position of retailers, the uncertainty of the market size, and the consideration of a fresh-keeping service. The model is mainly composed of the core retailer and the supplier. Through further numerical analysis, we verify the effectiveness of the revenue-sharing contract model in supply chain coordination. We also analyze the change trends in the optimal retail price, optimal freshness level, and optimal order quantity caused by changes in both the fresh-keeping service capacity and the revenue-sharing coefficient. The results show that after changing these two parameters, the supply chain can achieve coordination under the specified parameter values. The changed parameters will also lead to certain change trends in the optimal retail price, optimal freshness level, and optimal order quantity, and will have a corresponding impact on the stability of supply chain operation. This research provides a relevant theoretical and empirical basis for a fresh produce supply chain contract model with retailers at the core position. We also provide guidance and reference for optimizing the supply chain management mode and improving the overall operational efficiency of the fresh produce supply chain.

1. Introduction

With changes in consumer demand and the emergence of new distribution channels, the platform supply chain has been continuously optimized and upgraded, including integrating big data on the internet to achieve business updates and upstream and downstream collaboration. Further coordination of the relationships among members of the entire supply chain has created new consumption patterns, production methods, and business forms. The platform supply chain, with integrated trade and extensive management capabilities, is being rapidly shaped and developed.

The fresh produce platform supply chain is affected by production modes as well as consumption modes. Therefore, the supply chain must adapt to the production mode, the needs of consumers, and the characteristics of fresh produce [1]. In China, with the continuous development of the fresh produce supply chain and further improvements in distribution conditions, fresh produce is less affected by seasonal and regional constraints. As many constraints are overcome, different regions can circulate products with one another. The development of various retail formats allows more space for supply chain integration. Many items have become important factors and improved means of coordinating and optimizing the fresh produce platform supply chain, such as fresh-keeping service, revenue sharing, collaborative decision making, and cooperative incentives.

In contrast to traditional studies of the supply chain contract model of the fresh produce platform, which mainly focus on suppliers or demand, this paper focuses on supply chain preservation service requirements and revenue-sharing goals. As shown in Figure 1, we integrate the production and circulation characteristics of fresh produce preservation into the supply chain contract model design to construct a revenue-sharing contract model of the fresh produce supply chain based on the position of retailers at the core of the supply chain, the uncertainty of market size, and the consideration of fresh-keeping service. The model is mainly composed of core retailers and suppliers. Through further numerical analysis, the effectiveness of the revenue-sharing contract model in coordinating the supply chain is verified. Variation trends in price, optimal freshness level, and optimal order quantity are analyzed in detail.

The research innovations of this paper mainly include the following:

(1) According to the characteristics of fresh produce, such as perishability, large circulation loss, and high preservation requirements, this paper is different to traditional studies of fresh produce supply chain contract models focusing on suppliers or demand, and instead focuses on supply chain preservation service requirements and revenue-sharing goals. By integrating the production and circulation characteristics of fresh produce, such as freshness preservation, into the supply chain contract model design, a revenue-sharing contract model of the fresh produce supply chain is constructed based on the position of retailers at the core of the supply chain, the uncertainty of market size, and the consideration of freshness preservation services.

(2) Through further numerical analysis, the effectiveness of the revenue-sharing contract model in coordinating the supply chain, optimal retail price, and optimal freshness level and the change trend of order quantities is verified.

This paper is organized as follows: Section 1 describes the research background and significance, explains the connections to and differences between this study and previous studies, and introduces the main content and innovation points. Section 2 provides a review of the relevant literature and development background and describes the research contribution of this paper. Section 3 introduces the problem description and model assumptions in detail. Section 4 describes the revenue-sharing contract model construction, model derivation, research design, and so on. Section 5 presents the case discussion, data processing, program implementation, and results analysis. Notably, the example of the Fresh Networking project is used to verify the effectiveness of the model and algorithm. The conclusions and future research prospects are discussed in Section 6.

2. Literature Review and Research Content

2.1. Research on Supply Chain Contract and Coordination

Previous research has defined supply chain coordination from a systems perspective, in terms of adopting appropriate methods to organize or adjust the system in an orderly fashion. Finally, the whole system is coordinated [2].

The main goal of supply chain coordination is to stabilize the relationships among several member subsystems. If the relationships among members are not properly handled through effective coordination and an agreement is not reached, the overall function of the supply chain system will be affected [3].

A supply chain contract is a document with legal effect that has been agreed upon by the internal members. It is mainly intended to solve two core problems in the supply chain. On the one hand, a bullwhip effect will occur if there is information asymmetry among supply chain members. On the other hand, supply chain members try to maximize their interests, which will cause a double marginal effect.

Cooperation among supply chain members has a direct and important impact on supplier performance. This shows that effective implementation of the contractual relationship contributes to cooperation and the gradual application of the contract in the supply chain.

Revenue sharing, buybacks, wholesale prices, and quantity elasticity contracts are the four main types of contractual relationships in the supply chain. Other contract types are derived from these four contractual mechanisms.

Table 1. Four main contractual relationships in fresh produce supply chain.

Contract Type	Main Content
Revenue-sharing contract	Upstream supply chain members sell fresh produce to downstream members at low wholesale prices and obtain sales revenue from downstream members at a fixed proportion after the sales period ends.
Buyback contract	After the product sales period ends, upstream supply chain members collect unsold fresh produce from downstream enterprises at repurchase price to encourage downstream members to expand their order range of fresh produce.
Wholesale price contract	Upstream member enterprises decide the wholesale prices of fresh produce, and downstream enterprises make order decisions based on market demand and wholesale prices of upstream companies.
Quantity flexibility contract	Downstream enterprises retain a certain amount of fresh produce for sale. When they receive more information about market demand, downstream members determine the final order quantity within the quantity provided by upstream enterprises.

presents the four main contractual relationships [4].

As an important means of supply chain coordination and operation, contracts have been favored by many researchers. Some researchers have used relationship theory to study the cooperative relationship in contracts. They have proposed and studied principal agent theory, the price coordination mechanism, the contract mechanism, the inventory control mechanism, and other types of cooperative contractual relationships. Bouncken [5] divided contracts in the supply chain into complete and incomplete forms. Wu et al. [6] carried out research on the issue of explicit contracts. Wan et al. [7] provided the explicit option coordination conditions for a disrupted supply chain under two supply chain structures and then explored the effects of the disruption and supply chain structure on the option coordination conditions. In a stochastic demand supply chain composed of a single manufacturer and a distributor, the contractual relationship cannot help the manufacturer achieve the goal of compatibility with the cost of all distributors’ efforts [8]. Revenue-sharing contracts [9,10], two-part tariff contracts [11], buyback contracts [12], and other contract forms have been used by many researchers to design effective incentives for the supply chain. Revenue-sharing and cost-sharing contracts are the basis of most contract studies [13]. Gualandris et al. [14] considered how the sustainability performance and buyer–supplier trust of key suppliers mediate and moderate such developments. Zhou et al. [15] studied an option contract model based on a basic model of the fresh agri-food supply chain and compared the production, profit, risk, and information sharing conditions of the supply chain in different cases. Scholars have designed different contracts to enable all supply chain members to form a vertical partnership, such as through revenue sharing, price discounts, or rebates [16]. These contracts encourage supply chain members to maximize their benefits in order to further improve the overall operational efficiency of the supply chain [17]. Kurpjuweit et al. [18] developed a typology of three supplier selection archetypes. Thomas et al. [19] decomposed social sustainability into dimensions of employee welfare and philanthropy to determine their effects on supplier selection.

2.2. Research on Platform Supply Chain Coordination Considering Fresh-Keeping Service

The coordination mode of the platform supply chain can effectively promote the operational level. However, many scholars have not considered or included perishable fresh produce and their characteristics in their studies. Currently, some research considers the impact of product loss and reduced freshness quality on the coordination of the fresh produce supply chain. Some scholars have proposed a supply chain optimization model that affects circulation loss. Based on this, some scholars put forward the loss problem of fresh produce, which is influenced by the two-stage effort in the transportation process. Cai et al. [20] studied checking the inventory and quality of fresh produce over time. Siddh et al. [21] provided a structured review of the existing literature on agri-fresh food supply chain quality (AFSCQ) and a platform for practitioners and researchers to identify the existing state of work, gaps in current research, and future directions in the field. Taleizadeh et al. [22] solved a chance–constraint platform supply chain problem with stochastic demand following a uniform distribution. Chen et al. [23] proposed a fresh platform model using a product production and sale system combined with IoT technology. Ju-ning et al. [24] proposed combined contracts and derived the value range of contract parameters that could realize coordination of the supply chain. Ghaemi et al. [25] constructed a new time-varying utility function around the loss of fresh produce and changes in consumer purchasing utility and quantity to maximize consumer utility.

Researchers have also comprehensively studied the key factors affecting the overall efficiency of the fresh produce supply chain and methods to improve the efficiency of supply chain integration [26]. Xu et al. [27] studied a new two-echelon supply chain, mainly using a model of short-term products with value loss, such as fresh produce. The retailer determines the order quantity according to the supplier’s quoted amounts. The supplier’s quoted amounts will help the retailer reduce the order cost. Ghiami et al. established a two-echelon supply chain for perishable fresh produce with limited storage space, where the market demand depends on the inventory level of retailers. They used genetic algorithms to find the best coordination strategy for allowing and not allowing shortage [28]. Trieneken et al. constructed a supply chain for perishable products comprising a manufacturer and multiple retailers. They obtained each retailer’s replenishment cycle length and the manufacturer’s replenishment time under the overall decision [29].

In summary, unlike an ordinary supply chain, a fresh produce platform supply chain has stricter requirements regarding freshness and loss, so its coordination and optimization issues are more complex. There are few studies on the coordination and optimization of a fresh produce supply chain, mainly because it is difficult to reflect its characteristics from one aspect alone. Some scholars have improved and innovated supply chain coordination and optimization methods according to the characteristics and actual situations, as shown in

Table 2. Classification of related papers.

Research Type	Main Related Papers
1. Supply Chain Contract and Coordination	Feng L, Govindan K, Li C, 2017 [2]; Swinnen J, Vos R, 2021 [3]; Fan C, Zhang Q S, Chen Y M, 2022 [4]; Bouncken R B, Ratzmann M, Tiberius V et al., 2020 [5]; Wu J, Zou L, Gong Y et al., 2021 [6]; Wan N, Li L, Wu X et al., 2021 [7]; Liang L, Wang X H, Gao J G, 2012 [8]; Cachon G P, Lariviere M A, 2005 [9]; Palsule-Desai O D, 2013 [10]; Wu X Y, Fan Z P, Cao B B, 2023 [11]; Hong X P, Gong Y M, Chen W Y, 2020 [12]; Yan B, Wu X, Ye B et al., 2017 [13]; Gualandris J, Kalchschmidt M, 2016 [14]; Zhou L N, Zhou G G, Qi F Z et al., 2019 [15]; Gokarn S, Kuthambalayan T S, 2019 [16]; Ma X L, Wang S Y, Islam S M N et al., 2019 [17], Kurpjuweit S, Wagner S M, Choi T Y, 2021 [18], Thomas R, Darby J L, Dobrzykowski D, 2021 [19]
2. Platform Supply Chain Coordination Considering Fresh-keeping Service	Cai X Q, Chen J, Xiao Y B et al., 2010 [20]; Siddh M M, Soni G, Jain R et al., 2017 [21]; Taleizadeh A A, Niaki S T A, Wee H M, 2013 [22]; Chen X, Chen R, Yang C, 2021 [23]; Ju-ning S U, Chen-guang L I U, Yong Y et al., 2015 [24]; Ghaemi Asl M, Adekoya O B, Rashidi M M, 2022 [25]; Fritz M M C, Schoggl J P, Baumgartner R J, 2017 [26]; Xu X H, Nie S Y, 2009 [27]; Ghiami Y, Williams T, Wu Y, 2013 [28]; Trienekens J, Zuurbier P, 2008 [29]

.

2.3. Research Contribution

As a perishable product, fresh produce has attracted the attention and interest of many scholars in China and other countries in the field of supply chain coordination, with a focus on certain coordination methods (contracts, incentives, evaluations, decision making, etc.). Current research is mainly content oriented. Uncertain delivery quantity, high fresh keeping service requirements, and large product loss are typical elements of fresh produce supply chain operations, and they are also the main topics that much of the current research is attempting to solve. At the same time, China is still in the stage of exploration and research on the supply chain of fresh produce, and research using numerical analysis and coordination optimization is also not systematic enough.

The research contribution of this paper is mainly that we focus on a revenue-sharing contract model of the fresh produce supply chain based on the position of the retailer at the core and the uncertainty of market size. We adjust the corresponding parameters and demonstrate through numerical analysis that the revenue-sharing contract can effectively coordinate the supply chain.

3. Problem Description and Model Assumptions

3.1. Problem Description

Diversity and uncertainty exist among the channels and members of the supply chain system. In the fresh produce supply chain, the demand can be a certain quantity or a distribution function, including one or more random variables (such as lead time). These uncertain variables are the main reason for income reduction in the fresh produce supply chain. Due to stakeholder pressure and potential market opportunities, the supply chain needs to establish effective cooperation through contract management and other ways [30]. At present, most of the research on contracts in the fresh produce supply chain is focused on suppliers, and research focusing on retailers is relatively lacking [31,32,33].

(1) With the rapid development of fresh e-commerce and community fresh marketing, retailers have increasingly occupied a core position in the fresh produce supply chain as the main investors. On this basis, it is necessary to establish a revenue-sharing contract model focusing on retailers. We need to reduce the impact of market size and other uncertainties on the supply chain. Meanwhile, it is also necessary to reasonably distribute the benefits of the supply system in order to reduce or eliminate conflicts of interest between the supply chain and its members. Further, a retailer-led fresh produce supply chain system should be specifically coordinated through the contractual relationships among members.

(2) In practice, the size of the market for fresh produce is usually uncertain. It is necessary to combine the characteristics of fresh produce (long production cycle, short sales cycle, and high requirements for fresh-keeping service) and strong cooperative objectives among supply chain members. Through revenue-sharing contract model research, an effective supply chain coordination method is established for the fresh-keeping service requirements.

This paper is based on the requirements of supply chain cooperation and the core position of the retailer. In order to meet the goal of revenue sharing, we also consider the product characteristics and circulation characteristics of fresh produce, such as perishability, large circulation loss, and high preservation requirements. Considering the fresh-keeping service, we discuss how to use a revenue-sharing contract to coordinate the fresh produce supply chain. Then, we analyze the impact trends and characteristics of the changing value of key factors, including the revenue-sharing coefficient and fresh-keeping service capacity.

3.2. Model Description and Assumptions

3.2.1. Model Description

The upstream and downstream members of the fresh produce supply chain are core retailers and suppliers. The following assumptions are made:

(1) Fresh produce in the supply chain consists of single-period products. Demand is a function of market price and freshness level, and the market demand is positive.

(2) The supply chain information is symmetrical. The order quantity of retailers is equal to the market demand, which is not affected by fluctuations in market size. Losses due to shortages and the residual value at the end of the period are not considered. Suppliers’ product availability and fresh-keeping capacity can fully meet retailers’ needs. Both parties recognize an optimal freshness level to ensure that expectations will be maximized and there will be no shortages.

(3) Supply chain members take the maximization of their expected utility as the decision criterion. Retailers with a risk-neutral attitude are the main investors in the supply chain. As the core members of the supply chain, retailers lead, formulate the revenue-sharing contract, and have a good anti-risk ability. As the followers, suppliers that are risk-averse accept or reject the contract.

(4) As the main investors in the fresh produce supply chain, retailers decide on order quantities and retail prices and receive a proportion of the revenue. Suppliers determine the freshness level and fresh-keeping service capacity.

3.2.2. Definitions of Notations

For the model construction and case analysis, we set some notations and defined some parameters, as shown in

Table 3. Parameters of the contract model.

Notation	Definition
$a$	Uncertain market size
$t$	Freshness level of fresh produce
$c$	Unit production (service) cost of supplier
$p$	Retail price of fresh produce
$d$	Market demand
$\eta$	Level of accepted risk of supplier
$q$	Quantity of fresh produce ordered by retailer from supplier
$w$	Wholesale price obtained by retailer from supplier
$\alpha$	Coefficient of influence of freshness on market demand
$\beta$	Coefficient of influence of retail price on market demand
$\phi$	Fresh-keeping service capability of supplier
$\theta$	Revenue-sharing coefficient of retailer
$\lambda$	Cost coefficient of supplier’s fresh-keeping service
$c_s(t)$	Fresh-keeping service cost required by supplier $t$
$\mu$	Average value of uncertain market size
${\sigma }^2$	Variance of uncertain market size
${\Pi }_r^1$	Total profit for retailer in revenue-sharing contract
${\Pi }_s^1$	Total profit for supplier in revenue-sharing contract
$U_r^1$	Utility of retailer in revenue-sharing contract
$U_s^1$	Utility of supplier in revenue-sharing contract
$U_{sc}^1$	Total utility of supply chain in revenue-sharing contract

.

The parameters of the model meet the following conditions: $c<p$; $d=a+\alpha t-\beta p$; $w=\phi p+(1-\phi )c$; $c_s(t)$ = $\lambda t^2$; $\alpha >0$; $\beta >0$; $\lambda >0$; and $0\le \theta \le 1$. The larger $\theta$ is, the higher the proportion of sales revenue the retailer receives; on the contrary, when $\theta$ is smaller, the supplier will receive a higher proportion of sales revenue.

For $0\le \phi \le 1$, when $\phi$ increases, the supplier’s fresh-keeping service capacity becomes stronger, and the service level becomes higher, i.e., the supplier has higher cost control ability. On the contrary, when $\phi$ decreases, the supplier’s fresh-keeping service capacity becomes weaker, and the service level becomes lower, i.e., the retailer has higher cost control abilities.

If $\eta >1$, the supplier is averse to risk; if $0<\eta <1$, the supplier has a preference for risk; and if $\eta =1$, the supplier is neutral to risk. This paper only analyzes the situation in which the supplier, as the supply chain collaborator, has risk aversion and avoidance characteristics.

As the core member of the fresh produce supply chain, the retailer reflects the overall risk attitude of the supply chain, i.e., the supply chain has a risk-neutral attitude.

The process of forming the final pricing at the end of the fresh produce supply chain is part of a typical contingency pricing strategy. The final retail price is a function of the freshness level of fresh produce. At the end of the supply chain, the retailer determines the final retail price based on the freshness level of products provided by the supplier. Specifically, before the formal operation of the fresh produce supply chain, considering the overall freshness requirements for fresh produce, the retailer will first inform the supplier of a terminal retail price. Then, the supplier decides on its fresh-keeping service level according to the requirements and officially announces it to the other supply chain members. Only then will the real pricing decision be generated and reflected in the final pricing decision made by the retailer.

4. Analysis and Construction of the Revenue-Sharing Contract Model Based on Fresh-Keeping Service

4.1. Revenue-Sharing Contract Model under Decentralized Decision Making

According to the previous assumptions, the retailer is at the core position of the revenue-sharing contract model. Under the decentralized decision system, the process of the Stackelberg game among supply chain members is as follows: First, the retailer announces the retail price of fresh produce $p$ to the supplier. Then, the supplier determines its fresh-keeping service capacity coefficient $1-\phi$ according to the requirements. Next, according to the initial price decision made by the retailer, the supplier decides on the freshness level of fresh produce $t$ to maximize its interests. Finally, the retailer determines the final retail price of product $p$ according to the freshness level determined by the supplier.

The goal of the supplier’s decision [31]:

$$\mathrm{M}\mathrm{a}\mathrm{x}{U}_s^2=\phi (p-c)\left(\mu +\alpha t-\beta p\right)-\lambda t^2-2{\phi }^2{(p-c)}^2{\sigma }^2$$

(1)

The goal of the retailer’s decision [31]:

$$\mathrm{M}\mathrm{a}\mathrm{x}{U}_r^2=(1-\phi )(p-c)\left(\mu +\alpha t-\beta p\right)-{(1-\phi )}^2{(p-c)}^2{\sigma }^2$$

(2)

When ${\alpha }^2\phi -2\lambda \beta -2\lambda (1-\phi ){\sigma }^2<0$, the equilibrium result is as follows:

$$p_b^{*}=\frac{\lambda \mu -{\alpha }^2\phi c+c\lambda \beta +2\lambda c\left(1-\phi \right){\sigma }^2}{2\lambda \beta +2\lambda (1-\phi ){\sigma }^2-{\alpha }^2\phi }$$

(3)

On this basis, according to the previous analysis and the retail prices announced by the retailer in the fresh produce supply chain, the supplier determines the optimal freshness level of fresh produce. The level is suitable for the supplier and the supply chain. The details are as follows.

We solve Equation (1) for the second derivative of the freshness level of fresh produce. It is easy to find that the derivative is negative. Then, we solve its first derivative and make it equal to zero, and obtain the optimal freshness level of fresh produce:

$$t_b^{*}=\frac{\alpha \phi \left(p-c\right)}{2\lambda }$$

(4)

When the latest freshness level is brought into the optimal decision result in Equation (3), the first and second partial derivatives are obtained using Equations (5) and (6):

$$\begin{array}{l}\frac{\mathit{\partial }{U}_r^1}{\mathit{\partial }p}=\frac{\left(1-\phi \right)[{\alpha }^2\phi -2\lambda \beta -2\lambda (1-\phi ){\sigma }^2]}{\lambda }p\\ +\frac{\left(1-\phi \right)[\lambda \mu -{\alpha }^2\phi c+c\lambda \beta +2\lambda c\left(1-\phi \right){\sigma }^2]}{\lambda }\end{array}$$

(5)

$$\frac{{\mathit{\partial }}^2{U}_r^1}{\mathit{\partial }{p}^2}=\frac{\left(1-\phi \right)[{\alpha }^2\phi -2\lambda \beta -2\lambda (1-\phi ){\sigma }^2]}{\lambda }$$

(6)

When the result of Equation (6) is negative, i.e., ${\alpha }^2\phi -2\lambda \beta -2\lambda (1-\phi ){\sigma }^2<0$ is satisfied, the optimal retail price of the terminal supply chain $p_b^{*}$ is obtained by making Equation (5) equal to zero. The results are given by Equation (3). When the result is brought into Equation (4), the optimal freshness level can be obtained, and the results are given by Equation (7).

The optimal order quantity of fresh produce under the decentralized decision is given by Equation (8):

$$t_b^{*}=\frac{\alpha \phi \left(\mu -\beta \mathrm{c}\right)}{4\lambda \beta +2\lambda (1-\phi ){\sigma }^2-2{\alpha }^2\phi }$$

(7)

$$q_b^{*}=\frac{\left(\mu -c\beta \right)\left[\left(\lambda -2\right){\alpha }^2\phi +2\lambda \beta \right]}{4\lambda \beta +2\lambda (1-\phi ){\sigma }^2-2{\alpha }^2\phi }$$

(8)

Considering the revenue-sharing contract, the retailer also needs to propose the revenue-sharing coefficient at the beginning. Under the decentralized decision, the objective function of the supplier’s decision is

$$\mathrm{M}\mathrm{a}\mathrm{x}{U}_s^2=\left[\left(1-\theta +\phi \right)p-\phi c\right]\left(\mu +\alpha t-\beta p\right)-\lambda t^2-2{\left[\left(1-\theta +\phi \right)p-\phi c\right]}^2{\sigma }^2$$

(9)

The objective function of the retailer’s decision is

$$\begin{array}{c}\mathrm{Max}{U}_r^2=\left(\theta p-w\right)\left(a+\alpha t-\beta p\right)\\ =\left[\left(\theta -\phi \right)p-\left(1-\phi \right)c\right]\left(\mu +\alpha t-\beta p\right)-{\left[\left(\theta -\phi \right)p-\left(1-\phi \right)c\right]}^2{\sigma }^2\end{array}$$

(10)

According to the above process, the Stackelberg game is used to determine the requirements of a reverse solution. Based on the supplier’s optimal decision analysis, the optimal freshness level of fresh produce is determined according to the retail price.

A new value, $-2\lambda$, is obtained by solving the second derivative of Equation (9) regarding the freshness level of product. It is easy to find that its value is negative.

Now, the expected utility of the supplier is a quadratic function of the freshness level of fresh produce, which is a concave function. Therefore, there is a unique optimal solution to maximize the profit of the supplier. When the first derivative is equal to zero, the optimal freshness level can be obtained.

$$t_c^{*}=\frac{\left[\left(1-\theta +\phi \right)p-\phi c\right]\alpha }{2\lambda }$$

(11)

For $t_c^{*}>0$, $\left(1-\theta +\phi \right)p_c^{*}-\phi c>0$ must be met. With $0\le 1-\theta +\phi \le 2$, we obtain the following:

$$p_c^{*}>\frac{\phi c}{1-\theta +\phi }$$

(12)

Then, we bring the optimal freshness level $t_c^{*}$ into Equation (10) and solve the first and second partial derivatives for the retail price, and we obtain the following:

$$\begin{array}{l}\frac{\mathit{\partial }{U}_r^2}{\mathit{\partial }p}=\frac{\left(\theta -\phi \right)\left[{\alpha }^2\left(1-\theta +\phi \right)-2\lambda \beta -2\lambda (\theta -\phi ){\sigma }^2\right]}{\lambda }p\\ +\frac{\left(\theta -\phi \right)\left[2\mu \lambda -{\alpha }^2\phi c+{\alpha }^{2}c\left(1-\phi \right)\right]}{2\lambda }\\ +\frac{\left(1-\phi \right)c\left(2\lambda \beta -{\alpha }^2\right)+4\lambda c(\theta -\phi )\left(1-\phi \right){\sigma }^2}{2\lambda }\end{array}$$

(13)

$$\frac{{\mathit{\partial }}^2{U}_r^2}{\mathit{\partial }{p}^2}=\frac{\left(\theta -\phi \right)\left[{\alpha }^2\left(1-\theta +\phi \right)-2\lambda \beta -2\lambda (\theta -\phi ){\sigma }^2\right]}{\lambda }$$

(14)

Let $A=\theta -\phi$, $B=1-\phi$

$$\begin{array}{l}\frac{\mathit{\partial }{U}_r^2}{\mathit{\partial }p}=\frac{A\left[{\alpha }^2\left(1-A\right)-2\lambda \beta -2\lambda A{\sigma }^2\right]}{\lambda }p\\ +\frac{A\left[2\mu \lambda +{\alpha }^{2}c\left(2B-1\right)\right]+Bc\left(2\lambda \beta -{\alpha }^2\right)+4\lambda cAB{\sigma }^2}{2\lambda }\end{array}$$

(15)

$$\frac{{\mathit{\partial }}^2{U}_r^2}{\mathit{\partial }{p}^2}=\frac{A\left[{\alpha }^2\left(1-A\right)-2\lambda \beta -2\lambda B{\sigma }^2\right]}{\lambda }$$

(16)

When Equation (16) is negative, i.e., $A\left[{\alpha }^2\left(1-A\right)-2\lambda \beta -2\lambda A{\sigma }^2\right]<0$, if Equation (15) is equal to zero, the optimal retail price $p_c^{*}$ can be obtained. The result is given by Equation (17).

By substituting the obtained results into Equation (11), the optimal freshness level of fresh produce $t_c^{*}$ can be obtained by Equation (18).

Meanwhile, the optimal order quantity $q_c^{*}$ under the decentralized decision of a revenue-sharing contract can be obtained by Equation (19).

In a revenue-sharing contract under the decentralized decision, when

$$A=\theta -\phi , B=1-\phi ;$$

$$A\left[{\alpha }^2\left(1-A\right)-2\lambda \beta -2\lambda A{\sigma }^2\right]<0;$$

$$A\left[2\lambda \left(\mu -2\beta c\right)+{\alpha }^{2}c\left(2B-2A+1\right)+4\lambda {\sigma }^{2}c\left(B-A\right)\right]+Bc\left(2\lambda \beta -{\alpha }^2\right)>0,$$

the equilibrium results are as follows:

$$p_c^{*}=\frac{A\left[2\mu \lambda +{\alpha }^{2}c\left(2B-1\right)\right]+Bc\left(2\lambda \beta -{\alpha }^2\right)+4\lambda cAB{\sigma }^2}{2A[2\lambda \beta +2\lambda A{\sigma }^2-{\alpha }^2(1-A)]}$$

(17)

$$t_c^{*}=\frac{A\left(1-A\right)\alpha \left[2\mu \lambda +{\alpha }^{2}c\left(2B-1\right)\right]+B\left(1-A\right)\alpha c\left(2\lambda \beta -{\alpha }^2\right)}{4\lambda A[2\lambda \beta +2\lambda A{\sigma }^2-{\alpha }^2(1-A)]}+\frac{\alpha c\left(1-B\right)}{2\lambda }$$

(18)

$$\begin{array}{l}{q}_c^{*}=\mu +\frac{A\left[{\alpha }^2(1-A)-2\lambda \beta \right]\left[2\mu \lambda +{\alpha }^{2}c\left(2B-1\right)\right]}{4A\lambda [2\lambda \beta +2\lambda A{\sigma }^2-{\alpha }^2(1-A)]}\\ +\frac{Bc\left[{\alpha }^2(1-A)-2\lambda \beta \right]\left(2\lambda \beta -{\alpha }^2\right)}{4A\lambda [2\lambda \beta +2\lambda A{\sigma }^2-{\alpha }^2(1-A)]}+\frac{{\alpha }^{2}c(B-1)}{2\lambda }\end{array}$$

(19)

For $p_c^{*}>c$, we obtain the following:

$$\frac{A\left[2\mu \lambda +{\alpha }^{2}c\left(2B-1\right)\right]+Bc\left(2\lambda \beta -{\alpha }^2\right)+4\lambda cAB{\sigma }^2}{2A[2\lambda \beta +2\lambda A{\sigma }^2-{\alpha }^2(1-A)]}>c,$$

Then,

$$A\left[2\lambda \left(\mu -2\beta c\right)+{\alpha }^{2}c\left(2B-2A+1\right)+4\lambda {\sigma }^{2}c\left(B-A\right)\right]+Bc\left(2\lambda \beta -{\alpha }^2\right)>0$$

4.2. Revenue-Sharing Contract Model under Collaborative Decision Making

Under collaborative decision making, the information among supply chain members is symmetric. It is assumed that the decision maker holds the information of the whole supply chain and can make decisions to maximize the overall utility of the supply chain. This is a state of absolute collaboration between the supplier and the demander.

The retailer’s profit is

$${\Pi }_r^1=\left(\theta p-w\right)\left(a+\alpha t-\beta p\right)=\left[\left(\theta -\phi \right)p-\left(1-\phi \right)c\right]\left(a+\alpha t-\beta p\right)$$

(20)

The retailer has a neutral risk attitude at this point, so the expected utility is equal to its expected profit:

$$\begin{array}{l}{U}_r^1=\left(\theta p-w\right)\left(a+\alpha t-\beta p\right)=E({\Pi }_r^1)-Var({\Pi }_r^1)\\ =\left[\left(\theta -\phi \right)p-\left(1-\phi \right)c\right]\left(\mu +\alpha t-\beta p\right)-{\left[\left(\theta -\phi \right)p-\left(1-\phi \right)c\right]}^2{\sigma }^2\end{array}$$

(21)

The supplier’s profit is

$$\begin{array}{l}{\Pi }_s^1=\left(1-\theta \right)p\left(a+\alpha t-\beta p\right)+\left(w-c\right)\left(a+\alpha t-\beta p\right)-\lambda t^2\\ =\left[\left(1-\theta +\phi \right)p-\phi c\right]\left(a+\alpha t-\beta p\right)-\lambda t^2\end{array}$$

(22)

The expected mean variance utility function of the supplier is

$$\begin{array}{l}{U}_s^1=E({\Pi }_s^1)-\eta Var({\Pi }_s^1)\\ =\left[\left(1-\theta +\phi \right)p-\phi c\right]\left(\mu +\alpha t-\beta p\right)-\lambda t^2-2{\left[\left(1-\theta +\phi \right)p-\phi c\right]}^2{\sigma }^2\end{array}$$

(23)

Under the revenue-sharing contract, when the supplier has a risk-averse attitude, the overall expected utility of the supply chain is

$$\begin{array}{l}{U}_{sc}^1=\left(p-c\right)\left(\mu +\alpha t-\beta p\right)-\lambda t^2-2{\left[\left(1-\theta +\phi \right)p-\phi c\right]}^2{\sigma }^2\\ -{\left[\left(\theta -\phi \right)p-\left(1-\phi \right)c\right]}^2{\sigma }^2\end{array}$$

(24)

We solve the first partial derivative of Equation (24) concerning the retail price and freshness level, respectively:

$$\begin{array}{l}\frac{\mathit{\partial }{U}_{sc}^1}{\mathit{\partial }p}=-2[\beta +2{\left(1-\theta +\phi \right)}^2{\sigma }^2+{(\theta -\phi )}^2{\sigma }^2]p\\ +\mu +\alpha t+\beta c+4c\phi \left(1-\theta +\phi \right){\sigma }^2+2(\theta -\phi )(1-\phi ){\sigma }^2\end{array}$$

(25)

$$\frac{\mathit{\partial }{U}_{sc}^1}{\mathit{\partial }t}=\alpha \left(p-c\right)-2\lambda t$$

(26)

The Hessian matrix is as follows:

$$H\left(p,t\right)=\left[\begin{array}{cc}\frac{{\mathit{\partial }}^2{U}_{sc}^1}{\mathit{\partial }{p}^2}& \frac{{\mathit{\partial }}^2{U}_{sc}^1}{\mathit{\partial }p\mathit{\partial }t}\\ \frac{{\mathit{\partial }}^2{U}_{sc}^1}{\mathit{\partial }t\mathit{\partial }p}& \frac{{\mathit{\partial }}^2{U}_{sc}^1}{\mathit{\partial }{t}^2}\end{array}\right]=\left[\begin{array}{cc}-2[\beta +2{\left(1-\theta +\phi \right)}^2{\sigma }^2+{(\theta -\phi )}^2{\sigma }^2]& \alpha \\ \alpha & -2\lambda \end{array}\right]$$

(27)

If $4\lambda [\beta +2{\left(1-\theta +\phi \right)}^2{\sigma }^2+{(\theta -\phi )}^2{\sigma }^2]-{\alpha }^2>0$, the Hessian matrix is negative definite. Thus, an optimal solution exists and is unique.

Let us make Equations (25) and (26) equal to zero and establish two equations simultaneously:

$$\begin{array}{l}-2[\beta +2{\left(1-\theta +\phi \right)}^2{\sigma }^2+{(\theta -\phi )}^2{\sigma }^2]p\\ +\mu +\alpha t+\beta c+4c\phi \left(1-\theta +\phi \right){\sigma }^2+2c(\theta -\phi )(1-\phi ){\sigma }^2=0\end{array}$$

(28)

$$\alpha \left(p-c\right)-2\lambda t=0$$

(29)

Then, Equations (28) and (29) are combined to obtain Equations (30) and (31). We know $p_a^{*}>c$, so

$$\frac{2\lambda [\mu +\beta c+4c\phi \left(1-\theta +\phi \right){\sigma }^2+2c(\theta -\phi )(1-\phi ){\sigma }^2]-{\alpha }^{2}c}{4\lambda [\beta +2{(1-\theta +\phi )}^2{\sigma }^2+{(\theta -\phi )}^2{\sigma }^2]-{\alpha }^2}>c,$$

then $\mu -\beta c>0$.

Under the revenue-sharing contract with collaborative decision making, if $4\lambda [\beta +2{\left(1-\theta +\phi \right)}^2{\sigma }^2+{(\theta -\phi )}^2{\sigma }^2]-{\alpha }^2>0$ and $\mu -\beta c>0$, there is a unique optimal retail price and an optimal freshness level. The results are as follows:

$$p_a^{*}=\frac{2\lambda [\mu +\beta c+4c\phi \left(1-A\right){\sigma }^2+2cAB{\sigma }^2]-{\alpha }^{2}c}{4\lambda [\beta +2{(1-A)}^2{\sigma }^2+A^2{\sigma }^2]-{\alpha }^2}$$

(30)

$$t_a^{*}=\frac{\alpha (\mu -\beta c)}{4\lambda [\beta +2{\left(1-A\right)}^2{\sigma }^2+A^2{\sigma }^2]-{\alpha }^2}$$

(31)

Taking $p_a^{*}$ and $t_a^{*}$ under cooperative decision making with a revenue-sharing contract into the optimal order quantity function, we obtain the following:

$$q_a^{*}=d_a^{*}=\frac{2\lambda \beta (\mu -\beta c)+8\lambda \left(1-A\right){\sigma }^2\left[\mu \left(1-A\right)-\beta c\phi \right]}{{[4\lambda [\beta +2{\left(1-A\right)}^2{\sigma }^2+B^2{\sigma }^2]-{\alpha }^2]}^2}$$

(32)

4.3. Supply Chain Coordination Conditions Considering Fresh-Keeping Service and Revenue Sharing

To coordinate the supply chain using a revenue-sharing contract, the retailer and the supplier must design their respective parameters, including the revenue-sharing coefficient and the fresh-keeping service capacity. By analyzing these two parameters, the total expected utility of the supply chain under decentralized decision making with a revenue-sharing contract is equal or close to the total expected utility of the supply chain under collaborative decision making.

Under the specific parameter values, further coordination of the supply chain can be achieved by adjusting the coefficient of the revenue-sharing contract. One or more equilibrium solutions can be achieved through supply chain coordination. As the proportion of income sharing meets the conditions, the following results are obtained:

$$\begin{array}{l}\left[\left(1-A\right)p_{d1}^{*}-\phi c\right]q_{d1}^{*}-\lambda {(t_{d1}^{*})}^2-2{\left[\left(1-A\right)p_{d1}^{*}-\phi c\right]}^2{\sigma }^2>\\ \phi (p_{b1}^{*}-c)q_{b1}^{*}-\lambda {(t_{b1}^{*})}^2-2{\phi }^2{\left(p_{b1}^{*}-c\right)}^2{\sigma }^2\end{array}$$

(33)

$$\left[A{p}_{d1}^{*}-Bc\right]q_{d1}^{*}-{\left[A{p}_{d1}^{*}-Bc\right]}^2{\sigma }^2>B(p_{b1}^{*}-c)q_{b1}^{*}-B^2{\left(p_{b1}^{*}-c\right)}^2{\sigma }^2$$

(34)

We assume that $p_{d1}^{*}=p_c^{*}$ and $q_{d1}^{*}=t_c^{*}$. The fresh-keeping service capacity $\phi$ and the income-sharing coefficient $\theta$ are unknown, while other parameters are known. We can obtain the parameter expressions of the fresh-keeping service capacity $\phi$ and income-sharing coefficient $\theta$ and find that the number of solutions to the two parameters is uncertain.

To coordinate the fresh produce supply chain, we need to effectively improve the expected utility of both the supplier and retailer under the revenue-sharing contract. Based on this, the relationship satisfied by Equations (33) and (34) can be obtained.

If some values of $\phi$ and $\theta$ do not meet the above conditions, the fresh produce supply chain cannot be coordinated. However, we can still explore the values of $\phi$ and $\theta$ when the supply chain is closest to coordination.

5. Case Study

In order to illustrate and verify the feasibility of the revenue-sharing contract model for the fresh produce supply chain considering the preservation investment, this paper performs a numerical analysis based on an actual case.

5.1. Data Processing and Computing

5.1.1. Example Background: Fresh Networking Project

On 11 April 2012, the Wuhan Information Industry Office launched the Fresh Networking project. The Fresh Networking project uses Justeasy Agricultural Technology Co., Ltd., a fresh e-commerce retail enterprise, as the core company. Based on the business-to-customer (B2C) model, a new integrated development platform for the circulation of fresh produce is constructed. The platform is dedicated to building a low-carbon, environmentally friendly, and healthy fresh produce supply chain to form a scientific and efficient industrial process, including a set of specific activities: sorting, processing, warehousing, sales, distribution, and after sales [34].

Fresh Networking uses internet technology and an e-commerce platform to realize the operational mode of direct production and marketing. It integrates the cross-industry activities of fresh produce (production, procurement, warehousing, processing, and distribution) into a scientific and efficient production process, as well as quality monitoring, food safety, diet, and other factors, to build a fresh agricultural product industry chain based on aspects of health, sustainability, low-carbon environmental protection, efficient operation, and quality assurance.

Supply chain members consider the e-commerce platform as the carrier and retailers as the core member enterprises. They jointly build online supermarkets and production and circulation systems for fresh produce. Functioning in a coordinated operational mode, the Fresh Networking project reduces the intermediate links of the fresh produce supply chain, saves processing time, and maximally responds to the freshness preservation and quality needs of consumers [35].

5.1.2. Data Collection

Taking the Fresh Networking project as an example, we analyzed the coordination problem of a two-echelon supply chain composed of Justeasy (the core retail and processing enterprise) and its supplier LW. The data collected for the empirical analysis of the contract model are shown in

Table 4. Basic data of case analysis (RMB).

Parameter	$\mathit{\mu }$	$\mathit{\alpha }$	$\mathit{\beta }$	c	$\mathit{\eta }$
Value	100	2	2	5	2
Parameter	σ	σ2	$\mathit{\theta }$	$\mathit{\phi }$	$\mathit{\lambda }$
Value	2	4	0.8	0.3	2

.

The model sets the retailer’s revenue-sharing coefficient $\theta$ and the supplier’s fresh-keeping service capability $\phi$. We improved the retailer-led supply chain contract mechanism. Then, based on the model mentioned before and its derivation process, it is easy to see that the supply chain’s total expected utility is always larger under collaborative decision making than under decentralized decision making.

5.1.3. Numerical Analysis

In the revenue-sharing contact, the values of the total expected utility of the fresh agricultural product supply chain under decentralized and collaborative decision making are gradually approached by adjusting $\theta$ and $\phi$ [16,17].

There has already been research conducted on the fresh produce supply chain revenue-sharing model considering fresh-keeping service and case data from the Fresh Networking project. Through numerical calculation of the retailer’s revenue-sharing coefficient $\theta$ and the supplier’s fresh-keeping service capability $\phi$, we can deeply analyze the influence trend of these two parameters ($\theta$ and $\phi$) on the three parameters.

As shown in Figure 2, Figure 3, Figure 4, Figure 5, Figure 6 and Figure 7, we used Matlab 2015a to analyze the numerical changes in the optimal retail price, freshness level, and order quantity concerning $\theta$ and $\phi$. Then, we analyzed the results and graphs to discuss the impact of the model on retail price, freshness level, and order quantity.

5.2. Result Analysis

5.2.1. Characteristics of Revenue-Sharing Contract on Effectiveness of Supply Chain Coordination

As shown in Figure 8, the total expected utility of the supply chain is always greater in collaborative decision making than in decentralized decision making. The supply chain cannot achieve coordination under this parameter design. However, with the collaborative change in the two coefficients, under some specific values, the total expected utility of the supply chain is very close between decentralized and collaborative decision making. Therefore, the revenue-sharing contract can effectively coordinate the supply chain.

5.2.2. Analysis of the Influence of Optimal Retail Price on Supply Chain Coordination

(1) The following can be seen from Figure 2 (decentralized decision making):

The optimal retail price increases with an increase in the retailer’s revenue-sharing coefficient. The hypothesis that the retailer is in the core leading position in the supply chain can also be verified. When the supplier’s fresh-keeping service capacity reaches 0.755, the change trend in the optimal retail price is the most obvious. At this point, the maximum value can be obtained.

Meanwhile, the optimal retail price shows an upward trend with increased retailer’s fresh-keeping service capacity. Overall, it shows a trend of first increasing and then decreasing. When the retailer’s revenue-sharing coefficient increases, the fresh-keeping serviceability increases, the supplier’s bargaining power strengthens, and the overall cost of the supply chain increases. As the core leader of the supply chain, the retailer uses contingency strategies to improve the optimal retail price. According to the above analysis, the market demand is related to the freshness level of fresh produce and the retail market price at the end of the supply chain. Additionally, it is affected by the uncertain terminal market size. The retail price increases with continuous improvement of the supplier’s bargaining power for fresh-keeping services. Then, the impact of market size is greater than that of the supplier’s fresh-keeping service. Hence, the retail price falls instead of rising.

(2) The following can be seen from Figure 3 (collaborative decision making):

Overall, the change trend in the optimal retail price is the same as that under decentralized decision making, i.e., it gradually increases with an increasing retailer’s revenue-sharing coefficient. Meanwhile, under collaborative decision making, the supplier’s fresh-keeping service represents the service level of the whole supply chain. The retailer’s retail price will gradually decrease with an improved fresh-keeping service capacity. When the fresh-keeping service capacity improves, the retailer’s ability to control costs is improved. Therefore, the retail price can be effectively controlled. Furthermore, the market competitiveness of agricultural products in the supply chain can be strengthened.

5.2.3. Analysis of the Impact of Supply Chain Coordination on the Optimal Freshness Level

(1) The following can be seen from Figure 4 (decentralized decision making):

The impact of fresh-keeping service and revenue sharing on the optimal freshness level of fresh produce is the same as that of retail price. When the supplier’s fresh-keeping service capacity reaches about 0.65, the change trend in the optimal retail price is the most obvious. At this point, the maximum value can be obtained.

(2) The following can be seen from Figure 5 (collaborative decision making):

In general, the optimal freshness level will not change too much as the revenue-sharing coefficient changes; it will be basically unchanged, i.e., there is no direct relationship between the optimal freshness level of the supply chain and the revenue-sharing coefficient under collaborative decision making, which is in line with the actual situation. At the same time, the optimal freshness level will fluctuate and rise to about 0.75 with an increased freshness service capacity. After a stable period, the optimal freshness level will fluctuate and fall. The reasons are as follows. When the fresh-keeping service capacity is enhanced, the freshness level of fresh produce becomes higher. However, the freshness level will affect the market size and retail price. Therefore, the freshness level of fresh produce in the supply chain decreases. The low freshness level will make the retailer reduce wholesale prices, resulting in lower retail prices.

5.2.4. Analysis of the Influence of Supply Chain Coordination on Optimal Order Quantity

(1) The following can be seen from Figure 6 (decentralized decision making):

In general, the optimal order quantity increases with the retailer’s revenue-sharing coefficient. However, the optimal order quantity will gradually decrease when the revenue-sharing coefficient increases to about 0.9. The reasons are as follows. The revenue-sharing coefficient is too large, affecting suppliers’ supply intention and the market’s scale. This further leads to a decline in order quantity. In addition, the optimal order quantity will gradually increase with the improved fresh-keeping service capacity. When the fresh-keeping service capacity increases to about 0.9, the optimal order quantity will gradually decrease as the service capacity further improves. The reason is that a too strong fresh-keeping service capacity will lead to high input costs, affecting the retailer’s willingness to order. Then, the order quantity will decrease.

(2) The following can be seen from Figure 7 (collaborative decision making):

In general, the retailer leads the fresh produce supply chain in the collaborative decision-making system. The retailer’s ability to control the supply chain strengthens when the revenue-sharing coefficient increases. The fluctuations in the optimal order quantity tend to be stable. In addition, with an increased fresh-keeping service capacity under collaborative decision making, the optimal order quantity becomes more stable. The reason is that the supply chain stability improves when the fresh-keeping service capacity improves, and this stabilizes the optimal order quantity.

To sum up, for decentralized decision making, the optimal retail price, freshness level, and order quantity of fresh produce in the supply chain gradually increase with the retailer’s revenue-sharing coefficient and the supplier’s fresh-keeping service capacity. Under collaborative decision making, the retail price increases with an increased revenue-sharing coefficient and decreases gradually with an improved fresh-keeping serviceability. The level of preservation will fluctuate with improved preservation serviceability. The order quantity will stabilize with enhanced revenue sharing and fresh-keeping service. In addition, due to the influence of market demand, scale, and other factors, the change trend will fluctuate.

6. Conclusions and Prospects

This research focuses on two issues:

(1) With the rapid development of fresh produce e-commerce and community fresh produce marketing, as the main investors of capital, retailers have gradually occupied a core position in the supply chain. On this basis, is it possible to establish a revenue-sharing contract model centered on retailers to reduce the impact of uncertainties such as market size on the supply chain, and at the same time reasonably distribute the revenue of the supply system and reduce or eliminate the resulting conflict of interest among members? Can a supply chain system with fresh produce retailers at the core be specifically coordinated through a contractual relationship among members?

(2) In practice, the market size of fresh produce is generally uncertain. How can the long production cycle, short sales cycle, and fresh-keeping service of fresh agricultural products and the strong cooperation goals among supply chain members be combined through the use of a revenue-sharing contract model to establish fresh-keeping service requirements as a way to effectively collaborate and coordinate the supply chain?

Addressing these two main problems, the supply chain contract model constructed in this paper is composed of the core retailers and suppliers of fresh produce. As the main investor and leader of the supply chain, the retailer has a strong ability to resist risk and generally has the right to determine the contract, price, and quantity of products in the supply chain; compared to the retailer, the supplier is smaller in scale and has a weak bargaining power. Thus, when choosing to accept or reject the contract, uncertain market demand will bring risks to the supplier’s decision making, and as a risk-averse follower member of the supply chain, the supplier will decide to supply the operation level of parameters related to the freshness of agricultural products in the chain.

Considering the core position of the retailers, the uncertainty of market size, and the requirements of a fresh-keeping service, we studied a revenue-sharing contract model of a fresh produce supply chain. Based on an actual case, we adjusted the fresh-keeping service parameters and revenue-sharing coefficient to analyze the coordination level of the supply chain. We also analyzed the change trend in the optimal retail price, freshness level, and order volume caused by changes in both the fresh-keeping service and revenue-sharing coefficient. This research provides a theoretical and empirical basis for a retailer-led fresh produce supply chain contract model. Moreover, it can help in optimizing the management mode and operation efficiency of the supply chain.

There are some limitations to this paper, highlighting directions for further research.

(1) The basic construction of coordinating the supply chain around the fresh food platform is still in progress. At present, there is a lack of extensive cases and practical data to carry out relevant empirical research, especially using a model built for integrated online and offline modes. With the full development of circulation in community fresh marketing and fresh e-commerce, the coordination and quality control of the supply chain in practice can be more fully verified in empirical data by studying a combined online and offline mode. This is an important direction for development in the future.

(2) Considering the fresh-keeping service and the core position of the retailer, we analyzed the impact of different factors on system coordination in a two-echelon fresh produce supply chain. However, our research contents and objects are limited to a relatively stable, risk-neutral environment. For future research and analysis, scholars can continue to explore the matching issues between research methods and the fresh produce supply chain. At the same time, more advanced mathematical models and methods can be used to further study the risk of potential market demand expansion and a random environment. In addition, it would also be worthwhile to study a multi-task and multi-level fresh produce supply chain with a fresh-keeping service.