Contract Coordination of Fresh Agri-Product Supply Chain under O2O Model

Abstract

The development of the fresh-food e-commerce has led scholars to pay more attention to research on the agricultural product supply chain. This paper analyses the operation mode of the new O2O retail fresh supply chain by constructing the Stackelberg game model, takes the freshness and freshness-keeping efforts of fresh agricultural products as the main considerations, and analyses and compares the overall income of the supply chain under different decision-making modes. The results of the study show that: (1) In a centralized decision-making model, collaboration between suppliers and retailers online and offline can increase their respective revenues, and overall supply chain profits increase as the proportion of collaboration increases; (2) compared to decentralized decision-making, revenue-sharing contracts can increase the overall profitability of supply chain members and the supply chain when the revenue sharing factor is relatively high in the case of online and offline channel collaboration; (3) finally, the impact of revenue-sharing contracts on supply chain profitability is discussed through numerical analysis.

1. Introduction

Agricultural products are necessities of life and rigid consumer goods, and they are one of the most important means of social production [1]. Fresh produce occupies an important position among agricultural products because of its own characteristics. Fresh agricultural products normally include vegetables, fruits, livestock, poultry, seafood, eggs, milk and meat [2], which are essential for improving the quality of people’s diet. Fresh agricultural products have the characteristics of a high water content, a short shelf life, and a high level of perishability. With the improvement of people’s living standards, consumers’ requirements for the quality of fresh agricultural products have gradually increased. Therefore, it is important to study the fresh agricultural product supply chain.

The expanded demand for fresh agricultural products is both a challenge and an opportunity. One of the biggest challenges is to develop efficient supply chains for fresh produce that can meet the growing global demand for these products while managing the logistical and market complexities to meet demand at affordable prices and with as little waste as possible [3]. At present, the traditional single physical sales channel is no longer able to meet the diverse consumer needs, so online sales make up for the lack of physical sales. The emergence of online sales channels has a huge impact on physical sales. Specifically, online sales have significant advantages in terms of price and convenience, so some consumers prefer to buy the products they need online. Therefore, it is important to coordinate the interests of all members of the online and offline channels.

The application of high technology in all levels of society has promoted industrial transformation, and the logistics and supply chain, which are led by emerging high technology, continue to develop in the direction of intelligent digitalization. With the popularity of the internet, e-commerce has caused an upheaval in the traditional fresh produce market [4]. Therefore, fresh produce is no longer limited to offline channels. In 2020, the COVID-19 epidemic affected the rapid expansion of the market; the epidemic gave rise to the rapid development of same-city delivery, direct broadcast selling, social e-commerce, and other models, which also accelerated the pace of changing the way in which consumers buy. While traditional physical channels can no longer meet this changing consumer demand [5], online channels can meet these individual consumer preferences. By integrating online platforms and offline shops, O2O becomes an innovative e-commerce model that shares online and offline information [6]. O2O differs from omnichannel which is mainly considered from the retailer’s point of view, while the O2O model can be analysed both from the retailer’s point of view and from the supplier’s point of view. The omnichannel model mainly refers to a retailer’s comprehensive sales model from online and offline brick and mortar to meet the diverse needs of consumers, with the aim of providing a unified and consistent service and experience to consumers, such as Macy’s and Walmart in the US, and Suning and Gome in China [7]. The strength of the O2O model lies in the coordination of online and offline channels to provide a full range of services and enhance customer value [8]. Due to the perishable nature of fresh produce, it is important for companies to choose a suitable sales method. There are a variety of sales models in the fresh produce market, such as retail, direct sales, two-channel sales and O2O (online-to-offline) sales [4]. Dual-channel and O2O models are growing rapidly because they enable the diversification of online and offline buying and selling channels. Thus, fresh food e-commerce has risen rapidly. In recent years, the produce industry has seen a combination of direct online sales and brick-and-mortar retail stores. For example, Amazon acquired Whole Foods, a U.S. organic food chain, and Alibaba opened a grocery store called Hema Fresh Market and partnered with pop-up stores across China.

In summary, research on the fresh agricultural product supply chain under the internet environment, especially research on supply chain coordination and optimization and the supply chain model, is the focus of scholars at present. A few scholars have started to pay attention to the O2O model of fresh agricultural products, but no scholars have studied the synergistic mechanism of the O2O model of the fresh agricultural product supply chain. Based on the literature, this paper considers the characteristics of the combination of online and offline channels of the O2O supply chain and the freshness requirement of fresh agricultural products, combines the current digital development direction of the fresh agricultural supply chain, and uses the Stackelberg game model to build a dual-channel revenue-sharing contract model from the perspective of channel cooperation and profit maximization to coordinate the supply chain.

Based on the above analysis, this paper first builds a Stackelberg game model consisting of suppliers and retailers. Second, we construct and analyse the overall profit of the supply chain under the centralized and decentralized decision-making models. The results show that the overall profit of the supply chain under the centralized decision-making model is higher than that under the decentralized decision-making model. Then, we design a revenue-sharing contract model under the dual-channel supply chain, through which the revenue sharing ratio between suppliers and retailers can be calculated to coordinate the supply chain. Finally, through numerical analysis, we conclude that the revenue sharing coefficient of retailers is within a certain range and the revenue-sharing mechanism can enable companies in the dual-channel supply chain to better achieve the effective allocation of resources between channels and achieve a win-win situation.

Considering that suppliers and retailers make decisions in both decentralized and centralized decision-making situations, this study attempts to explain the following questions:

(1)

With centralized decision making, how does the cooperation and non-cooperation between line and offline affect the profitability of supply chain members and the overall profitability of the supply chain?

(2)

How can coordination mechanisms be designed to increase total profitability under decentralized decision-making and achieve improvements in the fresh produce supply chain?

This paper makes two main contributions to the literatures:

(1)

As a single benefit-sharing model can no longer meet the benefit-sharing model of the O2O supply chain, this paper proposes a benefit-sharing contract model based on channel cooperation, which can achieve effective integration and allocation of resources between online and offline channels and better coordination of interests between supply chain members by taking advantage of the lower cost of retailer preservation efforts.

(2)

Compared with the traditional supply chain, the model of this paper shortens the circulation link to a certain extent, and the cooperation between online and offline channels can better realize the effective integration and allocation of resources, improve circulation efficiency and reduce logistics costs. The coordination of interests of supply chain members through game analysis and reasonable contract mechanisms can ensure the stability of the supply chain to a certain extent and has certain reference value for enterprise operations.

The rest of the article is organised as follows. In Section 2, we review the relevant literature. Section 3 describes the framework and model assumptions explored. Section 4 discusses the games of supply chain members under centralized and decentralized decision-making scenarios and the contractual coordination of dual-channel supply chains. Section 5 gives the analysis of numerical arithmetic examples and makes some recommendations. The conclusions and limitations of the study are given in Section 6.

2. Literature Review

2.1. Fresh Products Supply Chain Coordination

The study of fresh produce is crucial due to its perishable nature and its importance to human life. Current research on fresh produce is divided into the following categories: (1) cold chain transport and path optimization of fresh produce [9,10], with research methods being mainly algorithmic [11,12]. (2) Contractual coordination of fresh produce supply chains [13,14], with game theory as the main method [15,16]. (3) Fresh produce supply chain process optimization [17,18], the main method is empirical analysis [19,20]. However, this paper focuses on the contractual coordination of fresh produce supply chains.

The fresh produce supply chain can be seen as a process that starts with harvesting, with suppliers or other logistics service providers involved in freshness processing, packaging, and transportation, after which the produce is sold by retailers to end consumers. Compared with the general supply chain, the fresh produce supply chain requires more investment in distribution, processing, transportation and storage, with a focus on product freshness, quality and safety, and loss control. As fresh produce is an essential part of human life, an increasing number of scholars are paying attention to the problems in the fresh produce supply chain.

At present, methods of contractual coordination in fresh produce supply chains fall into the following categories: revenue-sharing contracts, freshness-keeping cost sharing contracts, investment cost sharing contracts, price discount contracts and option contracts. By collating the literature, we can find that some scholars have considered only one method of coordination when conducting contractual coordination of fresh produce supply chains. For example, in order to maintain the freshness of fresh produce during transportation, Cai et al. [21] use price discount contracts to coordinate producers and distributors, thus ensuring more revenue for both parties. Due to the negative effect of fairness concerns on freshness and utility, Yan et al. [22] find that revenue-sharing contract can achieve Pareto improvements in supply chains. Similarly, investment cost-sharing contracts not only reduce the amount of waste in the fresh produce supply chain, but also increase the profitability of supply chain members [23]. In times of demand uncertainty, Zhou et al. [24] develop option contracts to enable supply chain coordination and information sharing. In addition, option contracts can coordinate the supply chain in the event of disruption [25].

In addition, fresh produce supply chains have also been studied using two coordination methods. Hamed et al. [26] propose that option contracts can increase the profits of supply chain members and that revenue-sharing contracts can reduce the double marginalization effect, so they combine the two coordination mechanisms to coordinate the supply chain. Zhang and Gao [27] find that the dual alignment of revenue-sharing contracts and freshness-keeping cost-sharing contracts can increase the profits of both suppliers and retailers, enabling the alignment of dual-channel supply chains. A combination of revenue-sharing and investment cost-sharing contracts based on the investment decisions of supply chain members can lead to a win-win situation for both manufacturers and retailers [28]. The decentralized supply chain can be coordinated through freshness-keeping cost-sharing and revenue-sharing contracts [29]. At the same time, the decentralized fresh produce supply chain can also be coordinated through revenue-sharing contract and price discount contracts [30].

As shown in

Table 1. Summary of the related literature regarding fresh produce supply chain coordination.

Literature	Types of Contracts in the Supply Chain					Involves Online and Offline Cooperation
	Revenue-Sharing Contract	Fresh-Keeping Cost-Sharing Contract	Investment Cost-Sharing Contract	Price Discount Contract	Option Contract
Yan et al. [22]	√
Zhang and Gao [27]	√	√
Moon et al. [28]	√		√
Song and He [29]	√	√
Cai et al. [21]				√
Yan et al. [30]	√			√
Mohammadi et al. [23]			√
Zhou et al. [24]					√
Wan et al. [25]					√
Hamed et al. [26]	√				√
Our study	√					√

, we summarise the existing literatures on supply chain coordination for fresh produce. We can see that most studies of supply chain coordination have taken a revenue-sharing contract approach, but relatively few studies consider both online and offline collaboration. Therefore, our study introduces online and offline cooperation for fresh produce supply chain coordination on the basis of existing research.

2.2. O2O Model

With the growing integrated online and offline environment, O2O commerce has huge potential for growth [31]. Ding et al. [32] define O2O as “O2O, from online to offline, means that companies provide discounts, information and services through the Internet to attract consumers’ attention, allowing them to pay online and enjoy services offline, thus increasing consumer satisfaction and meeting personalized needs.” Many scholars have conducted research on the O2O model with different focuses.

Some studies on the O2O model have focused on price decisions, service quality and platform marketing. For example, based on the O2O model, Kong et al. [33] study the price decision problem between manufacturers and retailers in a closed-loop supply chain. Tang and Yang [34] study how different financing mechanisms affect pricing decisions in an O2O sales model when retailers are constrained by capital. Meanwhile, other scholars have studied the impact of O2O as an online platform, collaborative advertising strategies of supply chain members and online consumer reviews on the decisions of supply chain members [18,35]. In addition, to provide better service quality, Li et al. [36] investigate vehicle matching strategies in O2O freight platforms.

However, some studies have also focused on the issue of contractual coordination in the O2O model. Under O2O deterministic and stochastic demand, Govindan and Malomfalean [37] investigate how different coordination mechanisms affect the total profitability of the supply chain. Under the O2O sales model, Qiu et al. [38] find that wholesale price, cost- sharing and two-part tariff contracts can effectively coordinate a fragmented supply chain and increase supply chain profitability. Pei et al. [39] develop a new coordination mechanism (i.e., manufacturers offer discounts to offline consumers while offering volume discounts to retailers) for mitigating competition in the O2O channel. In addition, Yang and Tang [4] study the optimal price decisions of suppliers and retailers under different sales models, and find that the O2O model can lead to higher supply chain profits under coordination. Based on the O2O platform, a new two-stage risk-sharing contract can effectively coordinate the supply chain and lead to higher supply chain profits [40].

As shown in

Table 2. Summary of the related literature regarding O2O.

Literature	Research Content of O2O Model				Involves Freshness and Freshness-Keeping
	Pricing Decision	Service Quality	Platform Marketing	Contract Coordination
Kong et al. [33]	√
Tang and Yang [34]	√
Li [18]			√
Li et al. [35]			√
Li et al. [36]		√	√
Govindan and Malomfalean [37]				√
Qiu et al. [38]				√
Pei et al. [39]				√
Yang and Tang [4]	√			√
Yang and Peng [40]			√	√
Our study				√	√

, we summarise the main literatures on the different research foci in the O2O model. We can see that there is a relatively large literature on contractual coordination, but relatively little on contractual coordination in fresh supply chains that considers both freshness and freshness-keeping. Therefore, our study focuses on freshness and freshness-keeping as the main factors for contractual coordination between suppliers and retailers under decentralised decision making, which helps to bridge the gap in fresh produce supply chains.

2.3. Research Gaps

In summary, previous literature lacks research on fresh product supply chains in terms of both freshness and freshness-keeping effort, and there is a relative lack of literature that considers both online and offline channel cooperation under a revenue-sharing contractual coordination mechanism. Therefore, our model differs from previous studies in two aspects: First, this paper considers freshness of fresh agricultural product and freshness-keeping effort, and constructs a Stackelberg game model to analyse the profits of suppliers and retailers under centralised and decentralised decision making. Unlike the existing literature, we also consider the cooperation between suppliers and retailers in the online and offline channels under the centralised decision model. Second, in order to alleviate channel conflicts and increase the overall profitability of the supply chain, the revenue-sharing contract we designed also considers the cooperation between online and offline channels of supply chain members, which is of theoretical value in improving the current research related to fresh product supply chains.

3. Model Formulation

3.1. Assumptions and Variables

The supplier distributes the fresh produce with a unit production cost of c to the retailer at the wholesale price of ω and sells it to consumers online at the price of ${\mathrm{p}}_{\mathrm{m}}$; the retailer uses the price ${\mathrm{p}}_{\mathrm{r}}$ in traditional channels according to ω and ${\mathrm{p}}_{\mathrm{m}}$ to sell to consumers. Among them, the supplier’s online channel and the retailer’s cooperation orders account for $\lambda$ (0$\le \lambda \le$1), and the noncooperative part account for (1 − $\lambda$). When $\lambda$ = 0, the online channel has no cooperative relationship between the two parties.

(1) If both the supplier and the retailer want to maximize their own interests, then both are risk-neutral and fully rational decision-makers, where the supplier is the dominant party, and the retailer is the follower.

(2) According to the literature [13], this paper assumes that the freshness function of fresh agricultural products is $\theta \left(t\right)=1-\frac{t^2}{T^2},$ 0 ≤ t ≤ T, where t is the product sales lead time determined by the supplier according to the retailer’s demand and T is the life cycle of fresh agricultural products. From $\frac{\mathrm{d}\theta }{\mathrm{d}t}=-\frac{2t}{T^2}\le 0,$ $\frac{{\mathrm{d}}^2\theta }{\mathrm{d}{t}^2}=-\frac{2}{T^2}\le 0$, the freshness accelerates and decreases with time.

(3) If same freshness is the same, then the supplier’s online channel preservation effort cost is higher than the retailer’s traditional channel cost ($C_m>C_r$); thus, the supplier’s online channel is willing to cooperate with the retailer. According to the literature [41], suppose the unit preservation cost function is $C\left(\theta \right)=\frac{k{\theta }^2}{2}$; that is, the unit preservation cost is a strictly increasing function of $\theta$. Among them, k > 0 is the influence coefficient of the freshness preservation effort level on the cost; thus, the freshness preservation cost of the supplier $C_m=\frac{k_m{\theta }^2}{2}$, the freshness preservation cost of the retailer $C_r=\frac{k_r{\theta }^2}{2}$, and 0 < $k_r$< $k_m$.

(4) In the fresh produce distribution process, only the freshness cost is considered for ease of calculation. To make retailers willing to purchase products from suppliers through wholesale prices rather than through suppliers’ online direct sales channels, $\omega <p_m$ must be satisfied.

(5) The demand function is established according to the literature [42], assuming that the retailer’s demand is equal to the ordered quantity, without considering the inventory and the loss in transportation. The consumer demand in the offline channel is $Q_r$, and the demand in the online channel is $Q_m$:

$$\{\begin{array}{c}{Q}_r=\mu \chi -\frac{\alpha }{\theta }{p}_r+\frac{\beta }{\theta }{p}_m\\ Q_m=\left(1-\mu \right)\chi -\frac{\alpha }{\theta }{p}_m+\frac{\beta }{\theta }{p}_r\end{array}$$

(1)

where $\chi$ is the total market demand; $\mu \left(0<\mu <1\right)$ is the market share of consumer demand in the traditional channel under the dual channel; $\alpha (\alpha >0)$ represents the elasticity index of market demand with respect to price; $\beta (\alpha >\beta >0)$ represents the cross-price elasticity index between channels; and $\theta$ is the freshness of the product when the consumer receives it. Both channels have their own loyal customers, $Q_r>0$ and $Q_m>0$.

3.2. Notations and Definitions

In this paper, we use superscript “$c$” for centralized decision-making, superscript “$d$” for decentralized decision-making, superscript “$s$” for decision-making under revenue sharing, superscript “$*$” for optimal decision, subscript “$r$” for retailer, and subscript “$m$” for supplier.

$Q_m$: Consumer demand on online channels;

$Q_r$: Consumer demand in traditional channels; retailers’ order volume;

$\chi$: Total market demand;

$\mu$: Market share of consumer demand in traditional channels under dual channels;

$\omega$: Retailer’s wholesale price;

$p_r$: Retailer’s selling price;

$p_m$: Direct price from suppliers online;

$c$: Cost of the supplier;

$\theta$: Freshness of products;

$\lambda$: Percentage of supplier working with retailer in online channels;

$\alpha$: Elasticity of market demand with respect to price index;

$\beta$: Cross-price elasticity index between channels;

$C_m$: Supplier’s preservation costs;

$C_r$: Retailer’s preservation costs;

${\phi }_r$: Percentage of shared revenue for retailer in the traditional channel;

${\phi }_m$: Percentage of shared revenue for supplier in the online channel which is part of the cooperation.

3.3. Model Framework

The model structure of this paper is shown in Figure 1. Based on relevant theoretical research and using the O2O platform as an auxiliary system, we choose the Stackelberg game approach to analyse the coordinated decision-making relationship between the two parties.

4. Model Analysis

4.1. Supply Chain Centralised Decision Model Construction and Analysis

In the centralized decision-making model, suppliers and retailers are considered as a whole. Based on the above assumptions, the total profit of the supply chain can be obtained as follows:

$${\pi }^c=\left(p_r-c-C_r\right)Q_r+\left[p_m-\left(1-\lambda \right)\left(c+C_m\right)-\lambda \left(c+C_r\right)\right]Q_m$$

(2)

where $\left(p_r-c-C_r\right)Q_r$ is the total profit of the traditional channel and $\left[p_m-\left(1-\lambda \right)\left(c+C_m\right)-\lambda (c+C_r)\right]Q_m$ is the total profit of the online channel.

Proposition 1.

When centralised decision-making takes place, the retailer’s optimal retail price and the supplier’s best online direct selling price are$p_r^{c^{*}}$and$p_m^{c^{*}}$. Consumer demand in the traditional and online channels respectively are$Q_r^{c*}$and$Q_m^{c*}$. Total supply chain profit at this time is ${\pi }^{c*}$.

$$\{\begin{array}{c}{p}_r^{c^{*}}=\frac{\left({\alpha }^2-{\beta }^2\right)\left(c+C_r\right)+\left(1-\mu \right)\chi \beta \theta +\mu \chi \alpha \theta }{2\left({\alpha }^2-{\beta }^2\right)}\\ p_m^{c^{*}}=\frac{\left({\alpha }^2-{\beta }^2\right)\left[\left(1-\lambda \right)C_m+\lambda C_r+c\right]+\left(1-\mu \right)\chi \alpha \theta +\mu \chi \beta \theta }{2\left({\alpha }^2-{\beta }^2\right)}\\ Q_r^{c*}=\frac{\beta \left[\left(1-\lambda \right)C_m+\lambda C_r+c\right]+\mu \chi \theta -\alpha \left(c+C_r\right)}{2\theta }\\ Q_m^{c*}=\frac{\beta \left(c+C_r\right)+\left(1-\mu \right)\chi \theta -\alpha \left[\left(1-\lambda \right)C_m+\lambda C_r+c\right]}{2\theta }\\ {\pi }^{c*}=\frac{\left[\left(1-\mu \right)\chi \beta \theta +\mu \chi \alpha \theta -\left({\alpha }^2-{\beta }^2\right)\left(c+C_r\right)\right][\beta \left[\left(1-\lambda \right)C_m+\lambda C_r+c\right]+\mu \chi \theta -\alpha \left(c+C_r\right)\text{}}}{4\left({\alpha }^2-{\beta }^2\right)\theta }+\\ \frac{\left\{\left(1-\mu \right)\chi \alpha \theta +\mu \chi \beta \theta -\left({\alpha }^2-{\beta }^2\right)\left[\left(1-\lambda \right)C_m+\lambda C_r+c\right]\right\}[\beta \left(c+C_r\right)+\left(1-\mu \right)\chi \theta -\alpha \left[\left(1-\lambda \right)C_m+\lambda C_r+c\right]\text{}}}{4\left({\alpha }^2-{\beta }^2\right)\theta }\end{array}$$

(3)

Proof. 

See Appendix A. □

Conclusion 1:

Cooperation between online and offline channels of agricultural products can increase the total profit of the supply chain; the following proves the necessity of cooperation between online channels and traditional channels.

Proof: 

When $\lambda \to 0$, that is, the total profit of the supply chain when there is no partnership between online channels and traditional channels:

$$\begin{gathered} {\pi }_{\lambda =0}^{c*}=\frac{\left[\left(1-\mu \right)\chi \beta \theta +\mu \chi \alpha \theta -\left({\alpha }^2-{\beta }^2\right)\left(c+C_r\right)\right]\left[\beta \left(c+C_m\right)+\mu \chi \theta -\alpha \left(c+C_r\right)\right]}{4\left({\alpha }^2-{\beta }^2\right)\theta }+ \\ \frac{\left[\left(1-\mu \right)\chi \alpha \theta +\mu \chi \beta \theta -\left({\alpha }^2-{\beta }^2\right)\left(c+C_m\right)\right]\left[\beta \left(c+C_r\right)+\left(1-\mu \right)\chi \theta -\alpha \left(c+C_m\right)\right]}{4\left({\alpha }^2-{\beta }^2\right)\theta } \end{gathered}$$

Total profit differential of the dual-channel supply chain in the case of centralised decision-making:

$$\mathsf{\Delta }{\pi }^{c*}={\pi }^{c*}-{\pi }_{\lambda =0}^{c*}$$

(4)

After sorting to obtain

$$\begin{gathered} \mathsf{\Delta }{\pi }^{c*}=\frac{\lambda \left(C_m-C_r\right)\left\{\left(c+C_r\right)+\left(1-\mu \right)\chi \theta -\alpha \left(c+C_m\right)\right\}+\left\{\beta \left(c+C_r\right)+\left(1-\mu \right)\chi \theta -\alpha \left[\left(1-\lambda \right)C_m+\lambda C_r+c\right]\right\}}{4\theta } \\ \because C_m>C_{r },Q_m^{c*}=\frac{\beta \left(c+C_r\right)+\left(1-\mu \right)\chi \theta -\alpha \left[\left(1-\lambda \right)C_m+\lambda C_r+c\right]}{2\theta }>0 \\ \therefore \mathsf{\Delta }{\pi }^{c*}>0 \end{gathered}$$

(5)

That is, cooperation between the dual channels leads to increased supply chain benefits, and the increased profits increase with $\lambda$.

By comparing the total profit of the supply chain under centralised decision making when online and offline cooperate and when they do not, we can find that in the centralized decision-making model of supply chain members, the overall profit level of supply chain members when online and offline cooperate (i.e., $\lambda >0$) is always higher than the overall profit of supply chain members when online and offline do not cooperate (i.e., $\lambda =0$). This suggests that the share of orders from suppliers working with retailers in the online channel further optimizes the dual-channel supply chain under centralized decision making. At the same time, the management insight is that it is necessary to further increase the profitability of all members of the dual-channel supply chain by means of order cooperation in the traditional centralised decision coordination model. □

4.2. Supply Chain Decentralisation Decision Model Construction and Analysis

The cooperative part of the product order is compensated by the retailer for providing freshness measures and conducting offline for distribution by the supplier. The supplier’s revenue comes from direct sales in online channels and wholesale in traditional channels, and the retailer’s revenue is from offline physical sales.

Overall supply chain revenue is as follows:

$${\pi }^d={\pi }_m^d+{\pi }_r^d$$

(6)

Retailer profits are as follows:

$${\pi }_r^d=\left(p_r-\omega -C_r\right)Q_r$$

(7)

Supplier profits are as follows:

$${\pi }_m^d=\left[p_m-\left(1-\lambda \right)\left(c+C_m\right)-\lambda \left(c+C_r\right)\right]Q_m+\left(\omega -c\right)Q_r$$

(8)

The solution of the equilibrium price in the Stackelberg game model can be solved by the inverse induction method. In this paper, the supplier is the dominant party, and the retailer is the follower; in the second stage of the analysis, the retailer sets the offline sales price $p_r$ from the pursuit of maximizing its own profit ${\pi }_r^d$ through the wholesale price $\omega$ set by the supplier and the pricing $p_m$ in the online channel.

Proposition 2:

Under decentralised decision-making, the supplier also sets the online sales price$p_m$that maximizes the profit${\pi }_m^d$based on the predicted retailer pricing$p_r$. According to the calculation, the optimal online direct selling price and the optimal wholesale price of the supplier can be obtained as$p_m^{d^{*}}$and${\omega }^{d^{*}}$. Optimal traditional retail price is$p_r^{d^{*}}$. The profits for retailers and suppliers respectively are${\pi }_r^{d^{*}}$and${\pi }_m^{d^{*}}$. Total supply chain profit under decentralised decision-making is${\pi }^{d^{*}}$.

$$\{\begin{array}{c}{p}_m^{d^{*}}=\frac{\left({\alpha }^2-{\beta }^2\right)\left[c+\lambda C_r+\left(1-\lambda \right)C_m\right]\chi \alpha \theta +\mu \chi \beta \theta }{2\left({\alpha }^2-{\beta }^2\right)}\\ {\omega }^{d^{*}}=\frac{\left({\alpha }^2-{\beta }^2\right)\left(c-C_r\right)+\left(1-\mu \right)\chi \beta \theta +\mu \chi \alpha \theta }{2\left({\alpha }^2-{\beta }^2\right)}\\ p_r^{d^{*}}=\frac{\left({\alpha }^2-{\beta }^2\right)\left\{\alpha \left(c+C_r\right)+\beta \left[\left(1-\lambda \right)C_m+\lambda C_r+c\right]\right\}+\left(3{\alpha }^2-{\beta }^2\right)\mu \chi \theta +\left(1-\mu \right)\chi \alpha \beta \theta }{4\alpha \left({\alpha }^2-{\beta }^2\right)}\\ {\pi }_r^{d^{*}}=\frac{{\left\{\mu \chi \theta +\beta \left[\left(1-\lambda \right)C_m+\lambda C_r+c\right]-\alpha \left(c+C_r\right)\right\}}^2}{16\alpha \theta }\\ {\pi }_m^{d^{*}}=\frac{\alpha \left[\left({\beta }^2-{\alpha }^2\right)\left(c+C_r\right)+\left(1-\mu \right)\chi \beta \theta +\mu \chi \alpha \theta \right]\left\{\mu \chi \theta +\beta \left[\left(1-\lambda \right)C_m+\lambda C_r+c\right]-\alpha \left(c+C_r\right)\right\}}{8\alpha \theta \left({\alpha }^2-{\beta }^2\right)}+\\ \frac{\left\{\left({\beta }^2-{\alpha }^2\right)\left[\left(1-\lambda \right)C_m+\lambda C_r+c\right]+\left(1-\mu \right)\chi \alpha \theta +\mu \chi \beta \theta \right\}\left\{\left({\beta }^2-2{\alpha }^2\right)\left[\left(1-\lambda \right)C_m+\lambda C_r+c\right]+\alpha \beta \left(c+C_r\right)+2\left(1-\mu \right)\chi \alpha \theta +\mu \chi \beta \theta \right\}}{8\alpha \theta \left({\alpha }^2-{\beta }^2\right)}\\ {\pi }^{d^{*}}=\frac{\alpha \left[\left({\beta }^2-{\alpha }^2\right)\left(c+C_r\right)+\left(1-\mu \right)\chi \beta \theta +\mu \chi \alpha \theta \right]\left\{\mu \chi \theta +\beta \left[\left(1-\lambda \right)C_m+\lambda C_r+c\right]-\alpha \left(c+C_r\right)\right\}}{8\alpha \theta \left({\alpha }^2-{\beta }^2\right)}+\\ \frac{\left\{\left({\beta }^2-{\alpha }^2\right)\left[\left(1-\lambda \right)C_m+\lambda C_r+c\right]+\left(1-\mu \right)\chi \alpha \theta +\mu \chi \beta \theta \right\}\left\{\left({\beta }^2-2{\alpha }^2\right)\left[\left(1-\lambda \right)C_m+\lambda C_r+c\right]+\alpha \beta \left(c+C_r\right)+2\left(1-\mu \right)\chi \alpha \theta +\mu \chi \beta \theta \right\}}{8\alpha \theta \left({\alpha }^2-{\beta }^2\right)}+\\ \frac{{\left\{\mu \chi \theta +\beta \left[\left(1-\lambda \right)C_m+\lambda C_r+c\right]-\alpha \left(c+C_r\right)\right\}}^2}{16\alpha \theta }\end{array}$$

(9)

Proof. 

See Appendix B. □

Conclusion 2:

There is a double marginal effect in the dual-channel supply chain system, and a contractual mechanism needs to be designed for coordination to achieve Pareto optimality.

Proof: 

Retailer’s optimal zero price: $p_r^{c^{*}}=\frac{\left({\alpha }^2-{\beta }^2\right)\left(c+C_r\right)+\left(1-\mu \right)\chi \beta \theta +\mu \chi \alpha \theta }{2\left({\alpha }^2-{\beta }^2\right)}\ne p_r^{d^{*}}=\frac{\mu \chi \theta +\beta p_m^{d*}+\alpha \left({\omega }^{d*}+C_r\right)}{2\alpha }$ □

Supplier online best direct price: $p_m^{c^{*}}=p_m^{d^{*}}=\frac{\left({\alpha }^2-{\beta }^2\right)\left[\left(1-\lambda \right)C_m+\lambda C_r+c\right]+\left(1-\mu \right)\chi \alpha \theta +\mu \chi \beta \theta }{2\left({\alpha }^2-{\beta }^2\right)}$.

Total supply chain profit differential between centralised and decentralised decision-making:

$$\mathsf{\Delta }{\pi }^{*}={\pi }^{c*}-{\pi }^{d*}>0$$

(10)

Bringing Equations (A13) and (7) into the equation gives

$$\mathsf{\Delta }{\pi }^{*}=\frac{{\left\{\mu \chi \theta +\beta \left[\left(1-\lambda \right)C_m+\lambda C_r+c\right]-\alpha \left(c+C_r\right)\right\}}^2}{16\alpha \theta }={\pi }_r^{d^{*}}$$

(11)

The overall profit of the supply chain is higher in the centralised decision-making model than in the decentralised decision-making model.

By comparing the total supply chain profit under decentralised decision making with the total supply chain profit under centralised decision making, we can see that the total profit of the supply chain members under decentralised decision making is always lower than the total profit under centralised decision making. This is due to the reduction in the overall profit level of the supply chain because of the individual decisions of the dual-channel supply chain members. Therefore, we can achieve a Pareto improvement in the profitability of all members through contractual design.

4.3. Dual-Channel Supply Chain Contract Coordination

In the case of dual-channel cooperation, the supply chain can increase the revenue gained. However, in the case of decentralised decision-making, both suppliers and retailers make decisions from the perspective of maximizing their own interests, which results in lower overall benefits than those found in the case of centralised decision-making.

The revenue-sharing contract model designed in this paper is as follows. The traditional channel retailer buys fresh produce with unit cost c from the supplier at a lower price $\omega$ and promises to return the gained revenue to the supplier in the proportion of ${\phi }_r$, the online channel supplier returns part of the revenue from cooperation with the retailer to the retailer in the proportion of ${\phi }_m$.

Proposition 3:

Under contractual coordination, the supplier’s profit and the retailer’s profit can be calculated as${\pi }_m^{s^{*}}$and${\pi }_r^{s^{*}}$. And the total supply chain profit at this point is${\pi }^{s^{*}}$.

$$\{\begin{array}{c}{\pi }_m^{s^{*}}=\left\{\left[1-\frac{\left(1-{\phi }_{\mathrm{r}}\right)\mathrm{B}}{\mathrm{A}+\mathrm{B}}\right]p_m^{s^{*}}-\left[\left(1-\lambda \right){\mathrm{C}}_{\mathrm{m}}+\lambda {\mathrm{C}}_{\mathrm{r}}+\mathrm{c}\right]\right\}{\mathrm{Q}}_{\mathrm{m}}^{*}+\left[{\phi }_{\mathrm{r}}{p}_r^{s^{*}}+\frac{\left(1-{\phi }_{\mathrm{r}}\right)\mathrm{M}}{\mathrm{A}+\mathrm{B}}-\left(\mathrm{c}+{\mathrm{C}}_{\mathrm{r}}\right)\right]{\mathrm{Q}}_{\mathrm{r}}^{*}\\ {\pi }_r^{s^{*}}=\left[\left(1-{\phi }_r\right)p_r^{s^{*}}-\frac{\left(1-{\phi }_{\mathrm{r}}\right)\mathrm{M}}{\mathrm{A}+\mathrm{B}}\right]{\mathrm{Q}}_{\mathrm{r}}^{*}+\frac{\left(1-{\phi }_{\mathrm{r}}\right)\mathrm{B}}{\mathrm{A}+\mathrm{B}}{p}_m^{s^{*}}{\mathrm{Q}}_{\mathrm{m}}^{*}\\ {\pi }^{s^{*}}={\pi }_r^{s^{*}}+{\pi }_m^{s^{*}}\end{array}$$

(12)

Proof. 

See Appendix C. □

Conclusion 3:

There exists${\phi }_r\in (0,1)$such that${\pi }^{s*}={\pi }^{c*}>{\pi }^{d*}$and the supply chain reaches coordination.

Proof: 

Let the functions $F\left({\phi }_r\right)={\pi }^{c*}-{\pi }^{s*}$ and $p_m^{s^{*}}=p_m^{c^{*}},p_r^{s^{*}}=p_r^{c^{*}}$; then, $F\left({\phi }_r\right)$ can be expressed as follows:

$$F\left({\phi }_r\right)=\lambda {\phi }_m{p}_m^{s^{*}}{Q}_m^{*}-{\phi }_r{p}_r^{s^{*}}{Q}_r^{*}=\frac{B{p}_m^{s^{*}}{Q}_m^{*}}{A+B}-{\phi }_r\left(\frac{B{p}_m^{s^{*}}{Q}_m^{*}}{A+B}+p_r^{s^{*}}{Q}_r^{*}\right)$$

(13)

From Equation (13), $F\left({\phi }_r\right)$ is a monotonic continuous function on ${\phi }_r\left(0<{\phi }_r<1\right)$ and

$$\{\begin{array}{c}\underset{{\phi }_r\to 0}{\lim }F\left({\phi }_r\right)>0\\ \underset{{\phi }_r\to 1}{\lim }F\left({\phi }_r\right)<0\end{array}$$

(14)

There must exist ${\phi }_r\in (0,1)$ such that $F\left({\phi }_r\right)=0$ and the revenue-sharing contract can be effectively coordinated to obtain the overall profit of the supply chain.

Through the design of the revenue sharing contract, we can find that when we control the revenue sharing factor within a certain range, the contract can effectively coordinate the total profit of the supply chain members. This leads to an improvement in supply chain profitability and retailer profitability under decentralised decision-making. The management insight we can draw is that suppliers and retailers should make full use of the contractual mechanism to coordinate the profits of all parties. □

5. Numerical Simulation Analysis

The main purpose of this case study is to demonstrate that revenue-sharing contract in the case of online and offline partnerships can coordinate the fresh product supply chain. Applying the theoretical results to actual enterprises can solve practical problems in fresh product supply chain coordination and further improve the profitability of supply chain enterprise members and the stability of the supply chain.

Shanghai Yihe Agricultural Products Technology Development Co., Ltd. is a comprehensive service provider for the distribution of global quality fresh product. The company has established a deep sales channel covering the whole country in China, providing direct terminal sales and commercial distribution services for fresh product according to different customer types.

Sales data for the company’s fresh produce were obtained for this study. Based on consumer feedback, we know that the acceptable freshness level for consumers is no less than 0.5. The effect of freshness on profitability is linear and stable when the freshness level of fresh product is higher than 0.5. The relevant parameters are shown in

Table 3. Supply chain benefit distribution parameters.

Project	$\mathit{\chi }$	$\mathit{\mu }$	$\mathit{\alpha }$	$\mathit{\beta }$	$\mathit{\lambda }$	${\mathit{C}}_{\mathit{r}}$	${\mathit{C}}_{\mathit{m}}$	c
Assignment	100	0.6	0.7	0.3	0.3	15	20	10

:

Based on the company’s actual data, we find that consumers do not choose to buy fresh product when $\theta <0.5$; thus, the value $\theta \in [0.5, 1)$ is used for the analysis.

According to Figure 2, we can more intuitively see that as the freshness of fresh produce $\theta$ increases, the profits of suppliers, retailers and the overall supply chain increase, and the overall supply chain revenue ${\pi }^{c*}>{\pi }^{d*}$.

Taking $\theta =0.5$, it is obtained from the theorem that ${\omega }^s,{\phi }_m$ can all be expressed linearly by ${\phi }_r$, and the gain sharing factor ${\phi }_r$ can be taken only to express. The optimal decision under different ${\phi }_r$ values is calculated by using Excel software. The calculations and the graphs based on the results are obtained in

Table 4. Effect of the revenue-sharing factor on supply chain profit.

${\mathit{\phi }}_{\mathit{r}}$	${\mathit{\pi }}_{\mathit{r}}^{\mathit{s}*}$	${\mathit{\pi }}_{\mathit{r}}^{\mathit{d}*}$	$\mathbf{\Delta }{\mathit{\pi }}_{\mathit{r}}^{\mathit{s}}$	${\mathit{\pi }}_{\mathit{m}}^{\mathit{s}*}$	${\mathit{\pi }}_{\mathit{m}}^{\mathit{d}*}$	$\mathbf{\Delta }{\mathit{\pi }}_{\mathit{m}}^{\mathit{s}}$
0.1	423.952	79.125	344.827	132.836	398.537	−265.701
0.2	376.846	79.125	297.721	179.942	398.537	−218.595
0.3	329.740	79.125	250.615	227.047	398.537	−171.489
0.4	282.635	79.125	203.510	274.153	398.537	−124.384
0.5	235.529	79.125	156.404	321.259	398.537	−77.178
0.6	188.423	79.125	109.298	368.365	398.537	−30.172
0.7	141.317	79.125	62.192	415.470	398.537	16.934
0.748	118.669	79.125	39.544	438.081	398.537	39.544
0.8	94.212	79.125	15.087	462.576	398.537	64.039
0.9	47.106	79.125	−32.019	509.682	398.537	111.145

and Figure 3, where $\mathsf{\Delta }{\pi }_r^s={\pi }_r^{s*}-{\pi }_r^d$, $\mathsf{\Delta }{\pi }_m^s={\pi }_m^{s*}-{\pi }_m^d$

According to Figure 3, when ${\phi }_r\in (0.664,0.832)$, the benefits of supply chain node members increase further under contractual coordination than when the decision is decentralised and the benefit sharing mechanism enables suppliers and retailers to reach a win–win situation.

6. Discussion and Managerial Insights

6.1. Discussion

In this paper, we construct a Stackelberg game model of cooperative and non-cooperative supply chains based on O2O platforms and explore the optimal decisions and overall supply chain benefits of suppliers and retailers in fresh produce supply chains under centralised and decentralised decision making. Our study finds that the revenue-sharing contract is effective in aligning the interests of supply chain members and the total supply chain benefits under cooperation between online and offline channels.

In the centralised decision-making model, suppliers and retailers share profits as a whole. Based on the calculations, we find that the cooperation of suppliers and retailers in both online and offline channels significantly increased the total profit of the supply chain, thus demonstrating the necessity of cooperation between online and traditional offline channels. And the greater the proportion of suppliers’ online channels working with retailers, the higher the total supply chain profit will be. In the decentralised decision model, there is a double marginal effect in the supply chain system as suppliers and retailers seek to maximize their respective profits, thus resulting in lower total supply chain profits under decentralised decisions than under centralised decisions, indicating that contractual coordination is necessary.

By designing a revenue sharing contract mechanism, we obtain a revenue sharing factor that can effectively coordinate the interests of each supply chain member and the total profit of the supply chain. The study shows that when there is a conflict of interests among supply chain members, the contract coordination mechanism can be designed to coordinate the overall profit of the supply chain. In addition, cooperation between supply chain members is important for the overall profitability and stability of the supply chain. Therefore, it is necessary to explore effective contractual coordination strategies under the O2O model to ensure that the profits of each supply chain member are maximized and to improve the efficiency of channel operations.

In previous studies, most scholars have used revenue-sharing contracts to study the coordination of fresh produce supply chains [22,26,27,28,29,30], but few have included both online and offline cooperation in the coordination. Therefore, the revenue-sharing contract designed in this paper coordinates the dual-channel system by considering that the retailer incentivizes the supplier’s online channel to cooperate with it with the advantage of lower freshness cost, while the supplier takes out part of the increased profit share from online cooperation to incentivize the retailer to cooperate with it. This type of coordination further illustrates the extent to which cooperative ratios can more effectively align the interests of supply chain members and reduce channel competition. In addition, most scholars use the contractual coordination method to study the supply chain in the research on O2O model [4,37,38,39,40], but there are relatively few studies that focus on freshness and freshness-keeping in the coordination of the supply chain. Therefore, with the O2O sales model in mind, we take freshness and freshness-keeping as the main factors to study the O2O contractual coordination in the fresh produce supply chain, thus illustrating the role of freshness in what range can better meet consumer demand. Therefore, this paper fills the gap in the fresh produce supply chain field related research gap.

6.2. Managerial Insights

In light of the consumer habits of consumers who shop through O2O platforms in their lives, we considered a supply chain model where suppliers and retailers sell online versus offline channels. We then propose revenue sharing contracts that mitigate profit loss in non-cooperative situations. The results of the study provide some management insights for supply chain members.

This paper has been able to make the members of the supply chain aware that competition between channels leads them to make decisions with only their own profit maximization in mind, thus making it impossible to maximize overall supply chain profit. Therefore, these findings remind suppliers and retailers that they can increase their respective margins through a collaborative model of online and offline channels. And contract coordination among supply chain members can effectively reduce the loss of fresh produce, improve freshness and win consumer acceptance. At last, contractual coordination between members can also effectively mitigate channel competition, improve the overall profitability of supply chain members and the supply chain, and maintain the stability of the supply chain.

7. Conclusions

This paper constructs a dual-channel fresh produce supply chain model consisting of a supplier and a retailer with the assistance of an O2O platform. The optimal decision problem in a Stackelberg game model based on freshness and freshness-keeping effort is investigated. It was found that: (1) In the centralised decision model, cooperation between online and offline can improve the efficiency of the supply chain and can be more profitable when the cooperation ratio is higher, so cooperation between channels is necessary. (2) The overall profit of the supply chain under the centralised decision-making model is better than that of the supply chain under decentralised decision-making. Therefore, we designed a revenue-sharing contract to coordinate the cooperation between online and offline channels of the supplier and retailer, and when the revenue-sharing factor is relatively high, the revenue of the supply chain members is better than the profit when the decision-making is decentralised, and the effective coordination between online and offline channels is achieved. Compared with the actual production transactions, this paper also has some limitations in the model analysis. First, this paper only considers the game decision of the second-level supply chain strips, while most of the real-life supply chains are multilevel. Second, the model is built on the basis of information sharing to anticipate the order demand quantity change function, and the impact of total shortage is not considered. Third, for the convenience of calculation, the real-life logistics cost and product loss are not considered to be included in the model. In addition, the dominance of decision-making also changes as the proportion of revenue sharing is jointly determined by the real participants in mutual discussion. In our subsequent research, we hope to investigate production optimization at the front end of the supply chain under demand uncertainty based on the literature [43,44,45,46], and thus coordinate more supply chain members, and to continue to improve in related theoretical studies.