The Ineffectiveness of “Volume Guarantee” Mode in Live-Streaming: A Nash Bargaining Analysis with Social Network Effects and Traffic Costs

Abstract

The unequal status between manufacturers and live-streamers often undermines supply chain profitability and social welfare. However, the “volume guarantee” commission mode, designed to mitigate this issue, has proven ineffective in practice. This paper adopts a Nash bargaining fairness framework to analyze this paradox, incorporating two defining features of live-streaming commerce: the social network effect and the streamer’s cost of purchasing public domain traffic. We develop a dynamic game model involving the platform, manufacturer, streamer, and consumers to examine commission mode selection and supply chain decision-making. Our analysis yields four key findings: (1) Under Nash bargaining fairness, the “volume guarantee” mode is invariably redundant, regardless of who sets the sales threshold. Bargaining power only influences profit distribution via commission rates without distorting optimal product pricing or traffic acquisition decisions. (2) The social network effect boosts product prices, traffic purchases, total profit, and social welfare, with its impact amplified by the streamer’s fanbase size. Thus, collaborating with top-streamers is advantageous for manufacturers. (3) While higher platform traffic costs do not affect the optimal product price, they reduce traffic purchase volume, thereby decreasing supply chain profits and social welfare. (4) To enhance social welfare, platforms can implement differentiated traffic pricing, offering discounts to top-streamers. This study provides critical managerial insights for designing fair contracts and fostering equitable cooperation in live-streaming ecosystems.

1. Introduction

Live-streaming is a new business model that has rapidly emerged in China and the world, and it has opened up new sales channels and lowered market entry barriers, allowing more small and medium-sized enterprises (SMEs) and individual entrepreneurs to participate in the market economy [1]. Through live-streaming platforms, producers can interact directly with consumers, shortening the product chain from production to sales, enhancing efficiency, and reducing costs [2]. Moreover, the format of live-streaming adds a new dimension to product presentation. With live commentary and physical demonstrations by hosts, it enhances the shopping experience for users and increases the conversion rate of purchases [3].

However, some issues beneath the surface of prosperity are gradually exposed, particularly the inequality between live-streamers and manufacturers [4,5]. In real life, top live-streamers with a large number of fans often gain higher cooperative status and bargaining power due to their market impact, while small and medium-sized enterprises find themselves at a disadvantage due to their lack of brand influence [6]. This phenomenon poses a severe threat to the healthy development of live-commerce and the socio-economic landscape [7]. For instance, top-live-streamers may demand lower prices or higher commissions from manufacturers, which can squeeze their profit margins [8]. Moreover, to gain higher status, some live-streamers engage in data fraud such as “brushing orders,” “buying followers,” and “manipulating reviews,” deceiving businesses and misleading consumers, resulting in a significant gap between actual live-commerce benefits and corporate expectations [9].

Many businesses enter the live-commerce field with high expectations for immediate sales results. However, the actual sales achieved through live-streaming often fall short of expectations [10]. The high costs and uncertain returns have led some companies to question the profitability of live-streaming. As a result, the “volume guarantee mode” has emerged in live-streaming, aiming to reduce manufacturers’ risks and increase cooperation enthusiasm by ensuring sales volume [11]. Despite this, the unequal status between manufacturers and live-streamers seems to make the “volume guarantee mode” unappealing to the market and manufacturers, instead leading to more chaos in live-commerce and exacerbating issues such as data fraud, unfair competition, and contract disputes [12].

To sum up, how manufacturers and live-streamers can make optimal decisions and design cooperation contracts to enhance social efficiency while ensuring mutual benefits has become a critical issue that urgently needs to be addressed for the development of live-streaming.

However, research on the power asymmetry between manufacturers and live-streamers has yielded valuable yet conflicting insights. Qi et al. [13] found manufacturers may sacrifice system efficiency with ordinary streamers but face profit loss with top influencers. Conversely, Ye et al. [4] showed seller preferences are threshold-dependent on industry bargaining power and compensation demands. Furthermore, these studies share a critical limitation: they overlook the defining features of live-streaming supply chains, particularly the social network effect and strategic public domain traffic costs. This omission likely contributes to their inconsistent findings and limits their ability to propose effective coordination mechanisms against issues like data fraud and contract disputes [12].

To address this gap, this article takes Nash bargaining fairness as the research background, considers the social network effect and traffic cost of live-streaming, and uses the Nash bargaining model to depict the game behaviors between platforms, manufacturers, live-streamers, and consumers. This article analyzes the optimal supply chain decisions under the “general” and the “volume guarantee” commission mode, and explores the impact of social network attributes, fanbase size, traffic costs on supply chain decisions (commission mode selection, price setting, traffic purchase), supply chain profit, and social welfare. Within this framework, the following three issues were mainly studied under Nash bargaining fairness:

⮚

(1) How do supply chain decisions, revenue, and social welfare depend on factors such as public traffic costs, social network effects, and fanbase-size of live-streamers?

⮚

(2) How efficient is the “volume guarantee” mode, and is it better than the “general” mode?

⮚

(3) How does unequal status (bargaining power) affect supply chain decision-making, profit, and social welfare?

By examining the above issues, our results yield novel insights into coordination mechanisms within live commerce supply chains.

• First, we demonstrate that under a fair profit-sharing scheme, manufacturers and streamers can achieve supply chain coordination without relying on external “volume guarantee” contracts. This outcome is unique to environments where long-term cooperation and fairness are valued, as the Nash bargaining solution endogenously aligns individual incentives with the collective optimum.
• Furthermore, our analysis highlights the critical role of the streamer’s social network. The size of the fanbase not only directly boosts demand but also amplifies the effectiveness of platform-based traffic, creating a multiplier effect on overall profits and social welfare.
• Additionally, we uncover a fundamental decoupling within the supply chain coordination mechanism: while the bargaining power determines the distribution of profits, it does not distort the operational decisions (pricing and traffic acquisition) that maximize the total supply chain value. This finding reveals that under Nash bargaining fairness, the pursuit of individual interest naturally aligns with collective efficiency, making complex “volume guarantee” contracts redundant for coordination.

The structure of next parts is arranged as follows. Section 2 explores the literature related to the article and the contributions of this article. Section 3 constructs the basic model based on bargaining game. Section 4 solves the game equilibrium of the basic model, and analyzes the influence of the social network effect and other factors on the equilibrium solution. Section 5 analyzes the basic model from two aspects: the difference in bargaining power and the maximization of social welfare. Section 6 summarizes the full text, the relevant management enlightenment, and the shortcomings.

2. Literature Review

The rapid development of live-streaming commerce has sparked extensive academic research into its operational mechanisms and efficiency. As shown in

Table 1. Literature summary and this study’s contributions.

Research Stream	Representative Literature	Focus of Existing Studies	Research Gaps	Contributions of This Study
Social Network Effects in Live-streaming	Hu and Chaudhry [14]; Kim et al. [1]; Peng et al. [15]	Consumer engagement drivers; Platform paradigm amplified by network effects.	Lack of integration with bargaining fairness and joint supply chain decisions.	Integration of social network effects into a Nash bargaining framework to analyze joint decisions and enhance efficiency.
Supply Chain Decision-Making in Live-streaming	H. Liu and S. Liu [16]; Ji et al. [17]; Lin et al. [7]; Zhang et al. [5]	Platform fees; Pricing/quality decisions; Streamer effort/commission; Power structures.	Omission of explicit fairness mechanisms; Underexplored solutions for equitable cooperation.	Introduction of Nash bargaining to endogenize fairness. Demonstration of efficiency-equity separation (power affects distribution, not operations).
Cooperation Mode Selection in Live-streaming	Abhishek et al. [18]; Wang et al. [19]; Zhang and Xu [11]	Comparison of resale vs. agency selling; “Volume guarantee” mode analysis.	Inconsistent findings on commission structures; Lack of theoretical guidance for equitable contracts.	Proof of the ineffectiveness of the “volume guarantee” mode under fair bargaining, providing a theoretical explanation for market observations.

The specific research details are as follows.

, this study engages with three interconnected domains: the social network effect, supply chain decision-making, and cooperation mode selection in live-streaming. This review systematically examines the key literature across these fields to clarify current research progress and limitations, particularly focusing on how to optimize supply chain decisions and enhance efficiency and social welfare through fair cooperation contract design (especially commission models) in live-streaming commerce, while considering core characteristics such as social network effects and traffic costs. This foundation establishes the groundwork for our research and highlights its innovative value.

2.1. Social Network Effect in Live-Streaming

As e-commerce evolved from the traditional shelf model to a trust-based and people-centered content model, live-streaming commerce has been endowed with strong social attributes [20]. The key to its development lies in leveraging the emotional trust and social presence fostered through interaction. Live-streaming enhances the buyer’s shopping experience and trust via real-time interaction between buyers and sellers. Streamers and consumers establish quasi-social relationships during live sessions, positively influencing consumers’ purchase intention [21].

Existing scholarship has systematically validated the pivotal role of social network effects in enhancing live-streaming commerce. Foundational studies demonstrate that social ties, structural dynamics, and financial incentives collectively shape consumer engagement. Hu and Chaudhry [14] established a theoretical framework integrating stimulus–organism–response models and relationship marketing theory, revealing that emotional commitment mediates the impact of both social and structural ties on participation, whereas financial ties exert indirect influence through this emotional conduit. You et al. [22] subsequently extended this framework through a survey of 515 viewers, identifying direct positive correlations between social network effects and three dimensions of consumer support: emotional, instrumental, and economic. These findings have been further corroborated and refined in recent research by Zou and Fu [23], Shen and Wang [13] and Khan et al. [24], which collectively underscore the multidimensional mechanisms driving social influence in this domain.

Beyond foundational mechanisms, contemporary studies have explored advanced applications of social network dynamics. Kim et al. [1] conceptualized live-streaming as a platform paradigm amplified by network effects, positing that value creation follows a self-reinforcing feedback loop as streamers expand their follower networks and consumer bases synergistically. Fan et al. [25] analyzed manufacturer strategies, demonstrating that stronger social connections between streamers and consumers increase the ROI of collaborating with high-influence streamers. Peng et al. [15] introduced a multi-unit bargaining framework to examine manufacturer–influencer negotiations, revealing that social influence exacerbates competitive disparities through a Matthew effect—dominant influencers consolidate market power at the expense of weaker players, compelling manufacturers to adopt dynamic partnership strategies. These advancements illustrate how social network effects transition from participation drivers to strategic decision-making levers in live-streaming ecosystems.

2.2. Supply Chain Decision-Making in Live-Streaming

The inherent heterogeneity in streamer behavior underscores that live-streaming commerce does not uniformly benefit manufacturers [26], necessitating nuanced research into supply chain decision-making. Early scholarship primarily addressed platform-level strategies, revealing how profit-sharing mechanisms shape operational outcomes. For instance, H.Liu and S.Liu [16] demonstrated through Stackelberg game models that higher platform commission rates inadvertently suppress streamer effort, ultimately diminishing profits for all parties—a critical insight into incentive misalignment.

Concurrently, manufacturers’ operational choices emerged as a central research strand. Ji et al. [17] pioneered this domain by integrating social interaction effects into game-theoretic frameworks, establishing optimal pricing and quality decisions for celebrity-driven sales—a foundation this study formally extends. Subsequent work deepened this inquiry: Lin et al. [7] employed Nash bargaining to dissect manufacturer–streamer negotiations, uncovering a strategic dilemma where brands face lower profits in live sales compared to traditional channels; crucially, they found that under low product-availability scenarios, prioritizing higher commissions over price reductions maximizes manufacturer returns. Further complexity was revealed by Zhang et al. [5], whose analysis of power structures showed that product quality deteriorates under both streamer-dominated and manufacturer-dominated regimes when commission ratios rise, whereas centralized decision-making preserves quality—highlighting governance as a key variable.

Parallel research explored streamer-centric strategies, focusing on commission structures [6] and effort optimization [7]. These studies implicitly acknowledged the symbiotic manufacturer–streamer relationship, where commission rates and promotional effort directly influence pricing flexibility, sales volume, and joint profitability. Yet, this interdependence remained underexplored within a unified fairness framework—a gap this study addresses by embedding Nash bargaining as a core mechanism.

2.3. Cooperation Mode Selection in Live-Streaming

Supply chain decision-making aims to maximize profit in live-streaming e-commerce. How this profit should be distributed, and how distribution affects the decisions and behaviors of supply chain actors, represents a subsequent challenge [27].

Currently, similar to general online sales, the primary cooperation modes are resale mode and agency selling mode, and the resale mode involves the streamer purchasing products wholesale from the manufacturer [18]; the agency selling mode is commission-based, forming a principal–agent relationship [28]. Numerous studies have explored the choice between these modes, considering factors such as fixed service fees [29], discount pricing strategies [30], return policies [31], commission strategies [32], and the brand effect of manufacturers [26]. For example, Wang et al. [32] modeled a platform retail supply chain’s sales/marketing choices via Stackelberg game to derive live-streaming agreement equilibria, finding retail prices rise with commissions and advising suppliers to partner with high- or low-influence streamers (not moderate) when using live channels.

On this basis, the “volume guarantee mode” studied in this paper is essentially a sub-model that influences the collaboration between manufacturers and streamers. Existing research often refers to it as the “target sales volume” or “sales commitment” [33]. Some studies suggest that this model can reduce streamers’ moral hazards and enhance the stability of collaboration through shared risk allocation and incentive compatibility [11,34]. For example, Zhang and Xu [11] investigated optimal proportional incentive contract design in live-streaming supply chains, demonstrating its superiority over linear contracts in achieving first-best solutions under threshold conditions and enhancing performance through commission-driven incentives.

2.4. Research Gap and Our Contribution

Despite significant progress, critical gaps remain in the existing literature:

(1)

While social network effects are recognized [1,15], their integration with bargaining fairness and its impact on joint supply chain decisions (pricing, traffic acquisition, commission bargaining) within the live-streaming context requires deeper exploration.

(2)

Research on supply chain decision-making has extensively analyzed factors like platform fees [16], manufacturer pricing/quality [5,17], and streamer effort/commission [7]. However, a key limitation is the frequent assumption of unequal power structures or the omission of explicit mechanisms to ensure fairness in the bargaining process between manufacturers and streamers. The inherent power imbalance is acknowledged as a problem [4,5], but solutions promoting equitable cooperation are underexplored.

(3)

Studies on cooperation mode selection, particularly regarding commission structures like the “volume guarantee” mode, reveal inconsistencies and practical challenges. For instance, conclusions about manufacturer preferences for influencers differ between Qi et al. [6] and Ye et al. [4], highlighting the difficulty in achieving fair decision-making. Existing research offers limited theoretical guidance for designing contracts that foster equitable cooperation between manufacturers and streamers, which is crucial for mitigating issues like data fraud and contract disputes [12].

Therefore, this study bridges these gaps by introducing Nash bargaining fairness as the core research background. We incorporate the essential characteristics of live-streaming—the social network effect and the streamer’s public domain traffic purchasing behavior—to construct a dynamic game model involving the platform, manufacturer, streamer, and consumer. Within this framework, we analyze commission mode selection (comparing the “general” mode and the “volume guarantee” mode) and supply chain decision-making under fairness constraints.

3. Base Model

3.1. Methodological Foundation: The Nash Bargaining Framework

Prior to formal model specification, let us justify the adoption of the Nash bargaining model as our core analytical framework. This choice is predicated on the model’s unique suitability for capturing the essence of manufacturer–streamer interactions in live-streaming commerce, which are characterized by cooperative dynamics under potential power asymmetry.

Traditional supply chain models, such as the Stackelberg game, often presuppose a clear leader–follower hierarchy. However, in live-streaming ecosystems, the power structure is more fluid and bilateral. Manufacturers possess product and brand authority, while streamers wield influence through their social capital and audience reach. This interdependence creates a scenario akin to a bilateral monopoly, where outcomes are determined through negotiation rather than unilateral decree. The Nash bargaining model is uniquely equipped for this context, as it does the following:

Endogenizes Fairness and Power Dynamics: It does not impose an exogenous power structure. Instead, it incorporates bargaining power parameters (e.g., manufacturer’s power e, streamer’s power 1 − e) that endogenously determine the profit-sharing rule, directly addressing the issue of unequal status highlighted in the introduction.

Focuses on Cooperative Surplus: The model’s primary objective is to maximize the joint profit (the “cooperative surplus”) before distributing it according to each party’s bargaining power. This aligns perfectly with the real-world goal of supply chain coordination, where both parties are incentivized to first expand the total pie.

Provides a Threat-Point-Based Rational Outcome: The solution is based on the parties’ disagreement points, which represent the outcomes if cooperation fails. This realistically reflects the optionality available to both manufacturers and streamers in a competitive market.

3.2. Game Players and Decision Order

This paper considers a manufacturer who wants to market his products with the help of a live-streamer, and they will carry out live-streaming on a third-party platform. Consider the most common sequential decision-making sequence at present, as follows.

First, the platform publishes the public domain traffic price and the manufacturer and the live-streamer make operational decisions according to their own and each other’s conditions, including the manufacturer’s decision on price and the live-streamer’s decision on traffic. Then, based on the two decisions and the expectation of the live-streaming effect, the manufacturer and the live-streamer haggle over the commission of the live-streaming. Finally, the live-streamer recommends the manufacturer’s products to consumers in the live room on the platform, and consumers decide whether to buy or not, as shown in Figure 1.

3.3. Research Assumptions

(1)

For the consumer

Different from the traditional sales channels, live-streaming with goods can enhance the emotional utility of the individual by establishing the emotional connection between the live-streamer and the audience. This is the source of social network effect, which helps to expand the influence of live-streaming. However, this emotional connection is often linked by consumers’ trust in the live-streamer, and there are great differences due to the personal characteristics of the live-streamers. In live-streaming, the size of the live-streamer’s fanbase is often the most intuitive representation of consumers’ trust in the live-streamer, which can well reflect the live-streamer’s influence in the social network. Therefore, we note the social network effect as $\lambda m$, $\lambda m<u$. Where $\lambda$ is the social network effect coefficient, and $m$ is the average online fanbase size of live-streamer. It should be noted that the larger the average online fanbase size, the larger the live-streamer’s fanbase size will generally be. Therefore, this paper also uses $m$ to characterize the live-streamer’s fanbase size.

Referring to the general expression of consumer utility, we assume the expected utility of consumers is:

$$u_c=v+\lambda m-p$$

(1)

where $v$ is the consumer’s estimate of the value of the product. Without losing the generality, we note $v~U\left(0,u\right)$, which is the fundamental embodiment of consumer heterogeneity, indicating that not all consumers entering the live studio have a desire to buy the product. In fact, this is a relatively common simplifying assumption in model construction, aimed at facilitating the derivation of analytical solutions for subsequent problems and their analysis. Many articles adopt this assumption [32]. In reality, consumers’ value estimates often follow more complex and even unobservable distributions, which we will continue to refine in future research. $\lambda m$ is the social network effect, which is an important manifestation of the difference between live-streaming and other channels; $p$ is the retail price of the product in the live-streaming.

Therefore, the market demand function is:

$$D=\left(m+n\right){\displaystyle \frac{\left(u-p+\lambda m\right)}{u}}$$

(2)

$n$ is the number of consumers attracted into the live room through purchasing public domain traffic. This demand function reflects the unique structure of live-streaming sales, where total demand is jointly composed of the anchor’s private traffic (online fans $m$) and paid public traffic $n$, and is modulated by the social network effect $\lambda m$ and product price $p$.

(2)

For the manufacturer

The manufacturer can decide the price of the product independently, but the commission needs to bargain with the live-streamer. Suppose each unit of the product sold must pay a reward to the live-streamer $\beta$, that is, the commission after the bargaining. In fact, the manufacturer always needs to pay a fixed service fee to the live-streamer, which is recorded as $\alpha$. Therefore, the expected profit of the manufacturer is:

$${\pi }_m=D\left(p-\beta \right)-\alpha$$

(3)

This profit function reflects the actual revenue of manufacturers in live-streaming sales, where manufacturers pay a slotting fee while also needing to pay a certain percentage of sales as commission to the host. However, for the convenience of subsequent solving and analysis, we have simplified the form of the commission by directly assuming a fixed commission amount rather than as a proportion of the price. Nevertheless, this simplification does not affect our analysis of the results.

(3)

For the live-streamer

In live-streaming, the traffic in the live-streaming room is generally composed of online fans (private domain traffic) and platform traffic (public domain traffic, the number of consumers recommended by the platform to the live-streaming room). The live-streamer does not need to pay for the online fans, but the platform traffic needs to be paid to the platform. Using the general cost and utility transformation function form for reference [35], let the public domain flow cost be ${\displaystyle \frac{1}{2}}k{n}^2{c}_r$. Therefore, the expected profit of the live-streamer is:

$${\pi }_l=\alpha +D\beta -{\displaystyle \frac{1}{2}}k{n}^2{c}_r$$

(4)

It should be noted that ${\displaystyle \frac{1}{2}}k{n}^2$ represents the total amount of public domain traffic that needs to be purchased to convert $n$ units of consumers into the live room. This form is more common in previous studies that describe the cost problem [36,37], and it indicates that the cost of traffic conversion increases, that is, with the increase in conversion volume, consumers per conversion unit entering the live room need to buy more public domain traffic. $k$ is the conversion coefficient of public domain traffic. $c_r$ is the price of unit traffic, which is set by the platform. In reality, the platform often adopts intelligent recommendation, and will give priority to users who are interested in the product. The users recommended in advance are more likely to enter the live-streaming room, and it will gradually decrease in the next recommendations. This form is exactly the description of this realistic background.

3.4. Commission Mode

At present, there are three typical commission modes in the mainstream market, including fixed service fee add commission mode, pure commission mode, and volume guarantee mode. Among them, the fixed service fee adds commission mode, that is, the manufacturer must pay a commission per unit of sales to the live-streamer and the fixed service fee, $\alpha +D\beta$. The pure commission mode, i.e., $\alpha =0$, can be merged into the the fixed service fee add commission mode in our model.

However, in the volume guarantee mode, the manufacturer needs to pay a commission per unit of sales based on the pit fee only when the sales volume in the live-streaming delivery reaches the sales volume threshold, otherwise it only needs to pay the pit fee. Naote that the threshold is $M$. In this paper, the latter is referred to as “volume guarantee mode” for short, and the former is referred to as “general mode”.

4. Model Solving and Equilibrium Analysis

4.1. Under General Mode

Combined with the game order, the specific solution process is as follows: Firstly, a Nash bargaining model is constructed to solve the bargaining game between the manufacturer and the live-streamer to get the commission $\beta$. Obviously, if the negotiation breaks down, the retained profits of both parties are $\overline{{\pi }_m}=0$, $\overline{{\pi }_a}=0$. Therefore, the Nash bargaining model is:

$$\mathrm{\Pi }=\underset{\beta }{\max }\left({\pi }_m-\overline{{\pi }_m}\right) \ast \left({\pi }_l-\overline{{\pi }_l}\right)$$

$$\mathit{s.t.} {\pi }_m-\overline{{\pi }_m}>0; {\pi }_l-\overline{{\pi }_l}>0.$$

(5)

By solving the above model, we can get $\beta ={\displaystyle \frac{p}{2}}-{\displaystyle \frac{\alpha u}{\left(m+n\right)\left(u-p+\lambda m\right)}}+{\displaystyle \frac{k{n}^2{c}_{r}u}{4\left(m+n\right)\left(u-p+\lambda m\right)}}$. After substituting into Equation (3) and Equation (4), the partial derivatives of $n$ and $p$ are calculated, respectively, to obtain the reaction function, and the simultaneous reaction function can be used to obtain Lemma 1:

Lemma  1.

Under Nash bargaining fairness, considering the traffic cost (the price of platform traffic) and social network effect, the optimal decisions of the manufacturers and the live-streamers in the general mode are as follows:

$$p^{*}={\displaystyle \frac{u+\lambda m}{2}}; n^{*}={\displaystyle \frac{{\left(u+\lambda m\right)}^2}{4ku{c}_r}}; {\beta }^{*}={\displaystyle \frac{3\left(u+\lambda m\right)-4k{c}_{r}m}{8}}+{\displaystyle \frac{2\lambda \alpha }{u+\lambda m}}-{\displaystyle \frac{\left(4\alpha -k{c}_r{m}^2\right)\left(m{\lambda }^2+\lambda u+4ku{c}_r\right)}{2\left[{\left(u+\lambda m\right)}^2+4ku{c}_{r}m\right]}}.$$

(6)

Lemma 1 reveals the optimal decision of the manufacturer and the live-streamer when considering the public traffic cost and social network effect under the general mode. From Lemma 1, it can be seen intuitively that the traffic purchase decision of the live-streamer and the traffic cost of the platform cannot affect the optimal pricing of the manufacturer’s products. This is because the live-streamer’s traffic purchase behavior cannot directly affect the utility of consumers. Although the live-streamer increases the number of potential consumers through the purchase of traffic, thereby affecting sales and manufacturer’s profits, it cannot affect the manufacturer’s optimal price.

Then, to further analyze how the optimal decision depends on factors such as the cost of public traffic, the first-order partial derivatives $\lambda$, $c_r$, $\alpha$, and $m$ are obtained and analyzed, respectively, for the supply chain decision in Lemma 1. This leads us to Proposition 1:

Proposition  1.

Under Nash bargaining fairness in the general mode, considering the traffic cost (the price of platform traffic) and social network effect, the following occurs:

The manufacturer’s optimal product pricing$p^{*}$increases with$\lambda$and$m$, invariant with$c_r$and$\alpha$;

The live-streamer’s optimal traffic purchase volume$n^{*}$increases with$\lambda$and$m$, and decreases with$c_r$, invariant with$\alpha$;

The optimal commission through bargaining${\beta }^{*}$increases with$\lambda$and$m$, and decreases with$c_r$and$\alpha$.

Proposition 1 reveals how the optimal decisions of manufacturers and live-streamers depend on $\lambda$, $c_r$, $\alpha$, and $m$ when considering the public domain traffic cost and social network effect. As shown in Figure 2, we can find the manufacturer’s optimal price, the live-streamer’s optimal traffic, and the optimal commission increase with the increase of $\lambda$ and $m$.

A particularly counterintuitive finding emerges regarding the live-streamer’s traffic acquisition behavior: rather than decreasing as the follower base expands, it increases. This observation contradicts the conventional intuition that a larger follower base naturally reduces the need for external traffic purchases. The underlying rationale is that a larger follower base enhances the live-streamer’s social influence and amplifies the social network effects experienced by consumers. Consequently, the live-streamer becomes more effective at converting potential consumers attracted by public domain traffic into actual buyers. This increased conversion efficiency justifies a higher willingness to pay for each unit of public domain traffic, leading to greater traffic investment.

Furthermore, the analysis reveals that the optimal commission rate decreases as the public traffic cost ($cᵣ$) or a fixed service fee (which can be interpreted as enhancing the manufacturer’s bargaining power) increases. This occurs because the manufacturer and the live-streamer are assumed to have equal bargaining power. While purchasing public domain traffic can strengthen the live-streamer’s bargaining position and push for a higher commission, an increase in traffic cost dampens this effort and weakens their bargaining power, thereby reducing the commission. Similarly, a higher fixed service fee enhances the manufacturer’s bargaining power, which also leads to a lower commission rate.

Next, we further analyze the impact of $\lambda$, $c_r$, $\alpha$, and $m$ on supply chain profits. By substituting the optimal decision in Lemma 1 into Equations (3) and (4), we can obtain Lemma 2:

Lemma  2.

Under Nash bargaining fairness, considering the traffic cost (the price of platform traffic) and the social network effect, the supply chain profits in the general mode are, respectively:

$${\pi }_m^{*}={\displaystyle \frac{m{\left(u+\lambda m\right)}^2}{8u}}+{\displaystyle \frac{{\left(u+\lambda m\right)}^4}{64{k{c}_{r}u}^2}}; {\pi }_l^{*}={\displaystyle \frac{m{\left(u+\lambda m\right)}^2}{8u}}+{\displaystyle \frac{{\left(u+\lambda m\right)}^4}{64{k{c}_{r}u}^2}}.$$

Lemma 2 gives the optimal profit of the manufacturer and the live-streamer considering the traffic cost and social network effect. Through the simple analysis of Lemma 2, it can be seen that when the manufacturer and the live-streamer haggle over the commission, their optimal profits are the same. This is because we have considered the equal bargaining power of the two.

Further, based on Lemma 2, the partial derivatives of $\lambda$, $c_r$, and $m$ are obtained for ${\pi }_m^{*}$ and ${\pi }_l^{*}$, respectively, and Proposition 2 is obtained as follows:

Proposition  2.

Under Nash bargaining fairness, considering the traffic cost (the price of platform traffic) and the social network effect in the general mode, both${\pi }_m^{*}$and${\pi }_l^{*}$increase with$\lambda$and$m$decreases with$c_r$and is invariant with$\alpha$.

Proposition 2 reveals the impact of traffic cost, fanbase size, and the social network effect on the profits of manufacturers and live-streamers. As shown in Figure 3, we find that the social network effect can effectively improve the profits of the supply chain, and this effect can increase with the increase in the number of live-streamer fans. That means that compared with the general live-streamers, the head live-streamers with larger fanbases do have greater advantages. Under the Nash bargaining fairness, the cooperation between the manufacturer and the head live-streamer can better obtain profits.

Furthermore, a more significant finding emerges from our examination: the profits of the manufacturer and the live-streamer are invariant to the service fee $\alpha$. The underlying rationale is that the bargaining process between the manufacturer and the live-streamer regarding the commission rate is, in essence, a negotiation over how to split the total revenue generated by the live-streaming activity. Within this framework, the service fee $\alpha$ acts merely as a transfer price or an internal accounting tool that recalibrates the baseline for commission negotiation. Consequently, it does not influence the final equilibrium of the profit-sharing arrangement, which is determined primarily by the relative bargaining power of the two parties.

4.2. Under the Volume Guarantee Mode

Under the volume guarantee mode, the utility of consumers will not change, and the total demand of consumers in the live-streaming delivery is still as shown in Formula (2) above.

Different from the general mode, if the sales volume of the live-streamer cannot reach the target threshold $M$, the manufacturer does not need to pay a commission $\beta$ to the live-streamer. At this mode, the expected profits of the manufacturer and live-streamer are:

$${\pi }_m^M=\left\{\begin{array}{c}Dp-\alpha D<M\\ D\left(p-\beta \right)-\alpha D\ge M\end{array}\right.$$

(7)

$${\pi }_l^M=\left\{\begin{array}{c}\alpha -{\displaystyle \frac{1}{2}}k{n}^2{c}_r D<M\\ \alpha +D\beta -{\displaystyle \frac{1}{2}}k{n}^2{c}_r D\ge M\end{array}\right.$$

(8)

The retained profits are the same as the general mode. Moreover, in reality, the target threshold $M$ is often determined by the live-streamer. In order to analyze this problem more comprehensively, this paper will consider three scenarios: ‘joint decision’, ‘live-streamer commitment’, and ‘manufacturer decision’.

4.2.1. Joint Decision

First, the bargaining model is constructed as follows:

$${\mathrm{\Pi }}^M=\underset{\beta ,\mathit{M}}{\max }\left({\pi }_m^M-\overline{{\pi }_m}\right)\times \left({\pi }_l^M-\overline{{\pi }_l}\right)$$

$$\mathit{s.t.} {\pi }_m^M-\overline{{\pi }_m}>0; {\pi }_a^M-\overline{{\pi }_a}>0$$

(9)

By solving the model, the following results can be obtained:

$$\beta ={\displaystyle \frac{p}{2}}-{\displaystyle \frac{\alpha u}{\left(m+n\right)\left(u-p+\lambda m\right)}}+{\displaystyle \frac{k{n}^2{c}_{r}u}{4\left(m+n\right)\left(u-p+\lambda m\right)}}, D\ge M$$

(10)

Comparing Equations (9) and (5), we can find that they are exactly the same. So, the subsequent solving process is also the same as that in the general mode. Therefore, we have Lemma 3 as follows.

Lemma  3.

Under the volume guarantee mode, if the target threshold$M$is jointly decided by the manufacturer and the live-streamer, the optimal decision of the manufacturer and the live-streamer is the same as that in the general mode, and$M$is set as$M\le {\displaystyle \frac{1}{u}}\left[m+{\displaystyle \frac{{\left(u+\lambda m\right)}^2}{4ku{c}_r}}\right]{\displaystyle \frac{u+\lambda m}{2}}$.

Lemma 3 indicates that if the sales target threshold $M$ is determined through bargaining, then adopting the volume guarantee mode under Nash bargaining fairness becomes ineffective. This conclusion can be attributed to two main reasons. First, although the target threshold is bargained upon, the bargaining process occurs after the initial decision-making. To ensure cooperation, the negotiated threshold cannot exceed the live-streamer’s capability; otherwise, an agreement would not be reached. Second, since the threshold must be achievable for the live-streamer, the bargaining process essentially mirrors that of the general commission mode. As a result, both the manufacturer and the live-streamer will naturally make the same optimal decisions as they would in the absence of a volume guarantee agreement

4.2.2. Live-Streamer Commitment

In order to maximize profit, the live-streamer must promise the sales target threshold on the basis of ensuring that their own target can be achieved. Under this premise, the transformation of the equilibrium solving problem is exactly the same as the general model. Thus, we have Lemma 4 as follows.

Lemma  4.

Under the volume guarantee mode, if the target threshold$M$is promised by the live-streamer, the optimal decision of the manufacturer and the live-streamer is the same as that in the general mode, and$M$is set as $M\le {\displaystyle \frac{1}{u}}\left[m+{\displaystyle \frac{{\left(u+\lambda m\right)}^2}{4ku{c}_r}}\right]{\displaystyle \frac{u+\lambda m}{2}}$.

Lemma 4 presents the manufacturer’s and the live-streamer’s decisions and the corresponding threshold under the volume guarantee mode, where the live-streamer commits to a sales threshold. The results in Lemma 4 align closely with those in Lemma 3. Furthermore, when the live-streamer provides a sales guarantee, the guaranteed amount itself does not materially influence the decisions within the supply chain.

4.2.3. Manufacturer Decision

Firstly, manufacturers should set sales thresholds to incentivize live-streamers to purchase traffic, thereby boosting product sales and profits, rather than establishing excessively high thresholds to prevent them from earning commissions. Such behavior could lead live-streamers to withdraw from cooperation. Therefore, during commission negotiations, it is crucial for manufacturers to ensure that the threshold $D$ is greater than or equal to a minimum value $M$ to maintain cooperation viability.

Secondly, any agreement must guarantee that the live-streamer profits from the arrangement, with their earnings not falling below the profit generated when no traffic is purchased. This can be expressed mathematically as ${\pi }_l^M\left(D\left[n\left(M\right)\right]\right)\ge {\pi }_l^M\left(D\left[0\right]\right)$, which represents the live-streamer’s profit under a sales threshold $D$ with traffic purchases, and ${\pi }_l^M\left(0\right)$ denotes the profit without such purchases. We disregard scenarios where profit is compared to zero sales volume because live-streamers’ followers constitute inherent resources; if traffic purchases result in lower profits, the effort becomes unprofitable and discourages further cooperation.

Thus, based on the bargaining model analysis, the decision-making problems for the manufacturer and the live-streamer are as follows:

$$\underset{P,M}{\max }{\pi }_m^M=D\left[n\left(M\right)\right]\left(p-\beta \right)-\alpha$$

$$\begin{gathered} \mathit{s.t.} \beta ={\displaystyle \frac{p}{2}}-{\displaystyle \frac{\alpha u}{\left(m+n\right)\left(u-p+\lambda m\right)}}+{\displaystyle \frac{k{n}^2{c}_{r}u}{4\left(m+n\right)\left(u-p+\lambda m\right)}}; {\pi }_l^M=\alpha +D\left[n\left(M\right)\right]\beta -{\displaystyle \frac{1}{2}}k{n}^2{c}_r; \\ D\left[n\left(M\right)\right]\ge D;{\pi }_l^M\left(D\left[n\left(M\right)\right]\right)\ge {\pi }_l^M\left(D\left[0\right]\right) \end{gathered}$$

(11)

Solving the above problems according to KKT conditions, Lemma 5 can be obtained as follows:

Lemma  5.

Under the volume guarantee mode, if the target threshold$M$is decided by the manufacturer, the optimal decision of the manufacturer and the live-streamer is the same as that in the general mode and$M$is set as follows: $M\le {\displaystyle \frac{1}{u}}\left[m+{\displaystyle \frac{{\left(u+\lambda m\right)}^2}{4ku{c}_r}}\right]{\displaystyle \frac{u+\lambda m}{2}}$.

The conclusion of Lemma 5 presents a result that diverges from initial expectations. Specifically, the sales target threshold set by the manufacturer remains unable to influence the manufacturer’s optimal pricing, the live-streamer’s traffic acquisition volume, or the commission rates. Although a higher threshold may incentivize the live-streamer to purchase more public domain traffic—which can increase sales volume—this action does not ultimately enhance the live-streamer’s profit. Moreover, as the volume of acquired public-domain traffic rises, the corresponding commission payable to the live-streamer also increases, which in turn reduces the manufacturer’s profit. Consequently, under Nash bargaining, the equilibrium outcome converges to that of the conventional mode, implying that the “volume guarantee” mechanism fails to alter the strategic decisions or improve the profits of either part.

Integrating Lemma 3, Lemma 4, and Lemma 5, we have Proposition 3:

Proposition  3.

Under Nash bargaining fairness, considering the traffic cost and social network effect, it is meaningless to adopt the volume guarantee mode.

Proposition 3 offers an explanation, grounded in traffic cost and social network effects, for the widespread industry resistance to the “volume guarantee” commission mode. It demonstrates that regardless of which party—the manufacturer, the live-streamer, or both—sets the sales target threshold, the ultimate decisions on pricing, traffic acquisition, and the commission structure remain identical to those in a general mode without any threshold. This indicates that the contract designed under Nash bargaining fairness is already optimized; manufacturers and live-streamers cannot find a superior alternative to this arrangement.

The underlying reason for this phenomenon lies in the supply chain contract framework based on Nash bargaining fairness. Within this framework, the manufacturer and the live-streamer engage in a bargaining process to distribute the total profits generated from the live-stream. The only avenue for either party to increase its own profit is to first expand the overall profit pie created by the collaboration. Consequently, maximizing total supply chain profit becomes a common goal for both parties. To achieve this, they will consciously adopt optimal decisions without needing the external imposition of a sales threshold $M$.

This leads to the counterintuitive conclusion that the “volume guarantee” clause is essentially redundant under Nash bargaining fairness. The mutual desire to maximize the total profit pool naturally aligns their incentives, making additional restrictive thresholds unnecessary for ensuring efficient outcomes.

5. The Expansion: Bargaining Power and Social Welfare

5.1. Heterogeneous Bargaining Power

In the above, we assume that the manufacturer and the live-streamer are absolutely fair, that is, they have the same negotiation ability. In reality, there are also many negotiators with different negotiation abilities to some extent. Therefore, we further explore the impact of bargaining power on supply chain decision-making.

Referring to relevant research [38], assuming that the bargaining power of the manufacturer is $e$, $0<e<1$, and the bargaining power of the live-streamer is 1 − e, then the bargaining model considering heterogeneous bargaining power is:

$$\mathrm{\Pi }=\underset{\beta }{\max }{\left({\pi }_m-\overline{{\pi }_m}\right)}^e \ast {\left({\pi }_l-\overline{{\pi }_l}\right)}^{1-e}$$

$$\mathbf{s.t.} {\pi }_m-\overline{{\pi }_m}>0; {\pi }_l-\overline{{\pi }_l}>0$$

(12)

We can use the above solving method to carry out the above analysis of Equation (10) to obtain Proposition 4.

Proposition  4.

Under Nash bargaining fairness, when considering the platform traffic cost and social network effect, the bargaining power does not affect the supply chain decision, but only positively affects the profit of each entity by affecting the commission.

As shown in Proposition 4, under the Nash bargaining fairness framework, although the bargaining power of manufacturers and live-streamers is not equal, their optimal decisions (such as quality investment and effort level) and intrinsic capabilities exhibit consistency. This indicates that power asymmetry does not distort the operational efficiency of the supply chain, and the total benefits of cooperation are maximized. The difference in bargaining power only redistributes the cooperative surplus through adjustments in transfer payments such as commission ratios. The party with greater power (whether the manufacturer or the live-streamer) can leverage their influence to set more favorable distribution terms during negotiations, thereby securing a larger profit share.

This finding indirectly confirms the validity of Proposition 3: under the fairness criterion of Nash bargaining, individual rationality of members and collective rationality of the supply chain achieve synergy. Both parties will spontaneously adopt strategies that maximize the total system profit, while power merely acts as the “rule-maker” for distribution, without affecting the efficiency foundation of “value creation.”

5.2. Social Welfare and Platform

The Nash bargaining can promote the equilibrium decision of maximizing social welfare under the premise of a two decision makers’ game. How related factors affect social welfare and how to further improve social welfare are also important issues of concern to society.

First, the optimal solution is substituted into the platform profit ${\pi }_p^{*}={\displaystyle \frac{{\left(u+\lambda m\right)}^4}{32{k{c}_{r}u}^2}}$. Then, we find the sum of consumer utility. Since the consumer value estimation $v~U\left(0,u\right)$, we only need to know the upper limit of the utility of a single consumer, and take $v=u$, $n=n^{*}, p= p^{*}$ into Formula (1). So, $\overline{U}={\displaystyle \frac{u+\lambda m}{2}}$. Therefore, according to the nature of uniform distribution, the total utility of consumers is: $U^{*}={\displaystyle \frac{0+\overline{U}}{2}}D={\displaystyle \frac{m{\left(u+\lambda m\right)}^2}{8u}}+{\displaystyle \frac{{\left(u+\lambda m\right)}^4}{32{k{c}_{r}u}^2}}$. Finally, Nash’s social welfare under fair bargaining is as follows:

$$SW=U^{*}+{\pi }_p^{*}+{\pi }_m^{*}+{\pi }_l^{*}={\displaystyle \frac{3m{\left(u+\lambda m\right)}^2}{8u}}+{\displaystyle \frac{3{\left(u+\lambda m\right)}^4}{32{k{c}_{r}u}^2}}.$$

Taking the partial derivatives of $\lambda$, $c_r$, and $m$ of $SW$, we can obtain Proposition 5 as follows.

Proposition  5.

Under Nash bargaining fairness, considering the traffic cost and the social network effect,$SW$increases with$\lambda$and$m$, decreases with$c_r$, and is invariant with$\alpha$.

Proposition 5 demonstrates that under Nash bargaining fairness, the social network effects, platform traffic pricing, and live-streamer fanbase size collectively influence social welfare. As illustrated in Figure 4, in live-streaming contexts, social network effects contribute positively to social welfare—an effect that is further amplified by a larger fanbase. In contrast, an increase in public traffic price exerts a negative impact on social welfare. These findings are intuitively explainable: the social network effect generates additional utility for consumers beyond the product itself, thereby enhancing consumer experience and elevating overall social welfare. On the other hand, raising traffic prices, while potentially boosting platform profits, discourages live-streamers from purchasing public domain traffic. This reduction in traffic acquisition ultimately diminishes the profits of both streamers and manufacturers, leading to a net decline in social welfare.

From a managerial perspective, these results highlight the importance of leveraging social connectivity and audience engagement as strategic assets. Platforms should prioritize fostering network effects and supporting streamers in expanding their follower base, rather than relying predominantly on monetizing public traffic. Such an approach not only aligns the incentives of all participants but also promotes sustained welfare growth in the digital ecosystem.

To enhance social welfare, two primary avenues can be pursued: strengthening the social network effect and reducing the price of public traffic. Consequently, for manufacturers, collaborating with live-streamers who possess a larger fanbase can lead to greater profitability and generate increased social welfare. For live-streamers, the focus should be on rigorous product selection, ensuring the quality of the live-streaming content, and thereby improving the social network effect coefficient. For the platform, providing support to live-streamers with substantial followings and offering lower public traffic prices can also contribute to social welfare enhancement. Based on the above, we have Corollary 1:

Corollary  1.

Under the framework of Nash bargaining fairness, if manufacturers choose to collaborate with influencers who have a larger fanbase and greater social influence, or if the platform can differentially lower the public traffic costs for these influencers, it can enhance social welfare.

Corollary 1 proposes a differential pricing strategy for platforms that yields counterintuitive yet impactful managerial insights. Contrary to the conventional wisdom of offering support to smaller entities [6], it suggests that platforms should set lower traffic prices for live-streamers with large fanbases and higher prices for those with smaller followings, after securing a baseline profit. The rationale is that the social network effect amplifies with the size of a streamer’s audience; a larger follower base enhances trust and generates stronger network externalities, which is a key driver for improving social welfare. Intentionally reducing the traffic cost for major streamers amplifies their drainage volume, creating a positive feedback loop that augments overall welfare.

This strategy finds empirical support in the collaboration between platforms and top streamers. For instance, the live-streaming ecosystem around Xinba and the Xinxuan Group demonstrates the massive sales volume and market influence that head streamers can command, effectively creating significant social network effects. Similarly, the success of Hongdou Men’s Wear partnering with top Taobao streamer Lie’er Baobei, which set a record for a single live-streaming event, underscores the powerful synergy between established brands and streamers with substantial reach. These cases illustrate that platforms, by potentially offering favorable traffic conditions to such high-impact streamers, can maximize the value generated for the entire ecosystem. The case of Huaihua’s suitcase industry further shows how specialized live-streaming events featuring influential hosts can rapidly boost sales for a regional manufacturing cluster, highlighting the practical application of leveraging large-fanbase streamers for widespread economic impact.

6. Conclusions

Live-streaming has emerged as a dominant force in e-commerce, fundamentally reshaping supply chain dynamics by integrating social networks with retail operations. This paper investigates how this integrated model influences the decisions of manufacturers, streamers, and platforms by developing a four-party game-theoretic framework that incorporates Nash bargaining to capture fairness concerns.

6.1. Summary of Main Findings

Our investigation yields several pivotal findings that challenge conventional wisdom and provide fresh perspectives on operational decision-making in this dynamic domain.

First, and most significantly, the study proves the inherent ineffectiveness of the “volume guarantee” mode under conditions of bargaining fairness. As established in Proposition 3, regardless of whether the sales threshold is set by the manufacturer, the live-streamer, or through joint negotiation, the resulting equilibrium decisions—optimal product pricing, traffic acquisition volume, and supply chain profit—are identical to those under the simpler “general” commission mode. This fundamental finding provides a theoretical explanation for the industry’s observed resistance to such clauses. The reason is that Nash bargaining fairness inherently aligns the individual objectives of both parties with the goal of maximizing the total pie (joint profit). Any attempt to artificially constrain the outcome with a sales threshold becomes redundant, as both parties are already intrinsically motivated to implement system-optimal decisions.

Second, the research quantifies the dual role of the social network effect, moving beyond qualitative descriptions. Proposition 1 reveals that the social network effect coefficient and the streamer’s fanbase size act as complementary drivers that simultaneously increase the optimal product price, the optimal traffic purchase volume, and the commission rate. This positive correlation between fanbase size and traffic acquisition is particularly noteworthy, countering the intuition that larger fanbases reduce the need for paid traffic. Instead, as Proposition 2 clarifies, a larger fanbase amplifies the marginal utility of each unit of acquired traffic by enhancing consumer trust and social utility, thereby making additional traffic investments more profitable. This creates a virtuous cycle where influential streamers can justify both higher prices and greater marketing expenditures.

Third, the model delineates the distinct impact of bargaining power from operational decisions. A key insight from Proposition 4 is that while the relative bargaining power of the manufacturer and streamer directly affects the profit distribution by influencing the final commission, it has no bearing on the optimal product price or traffic volume. This separation indicates that the efficiency of the supply chain (the size of the total profit) is determined by market factors and the social network, while the equity of the collaboration (the division of profit) is determined by bargaining power. This finding underscores that fostering operational efficiency and ensuring fair profit distribution are two separate, addressable challenges.

Finally, the analysis provides clear guidance for social welfare optimization. Proposition 5 establishes that social welfare increases with the social network effect and fanbase size but decreases with the platform’s traffic price. This leads to a critical managerial implication, formulated in Corollary 1: platforms can enhance overall welfare by implementing a differentiated traffic pricing strategy. Reducing traffic costs for top streamers with large fanbases encourages higher traffic acquisition, which in turn boosts transactions, supply chain profits, and consumer surplus, creating a more vibrant and efficient ecosystem.

6.2. Theoretical and Practical Implications

The findings carry significant implications for both theory and practice. Theoretically, this study contributes to the literature on supply chain coordination by introducing Nash bargaining fairness as a viable and efficient alternative to traditional incentive contracts in settings characterized by intense human interaction and relationship-building, such as live-streaming. It bridges the gap between operations management and social commerce theories by formally modeling the economic value of trust and social influence.

From a practical standpoint, our results offer clear guidance for various stakeholders:

For Manufacturers: The pursuit of partnerships should prioritize streamers with a large and engaged fanbase, as the social network effect is a quantifiable driver of profitability. More importantly, managers can simplify contract negotiations by moving away from difficult-to-enforce “volume guarantee” clauses and focusing instead on establishing a fair bargaining process for profit-sharing. For example, a small cosmetics brand could prioritize collaboration with a beauty streamer who has a highly engaged community of 1 million followers, even if the commission rate is higher, because the amplified social network effect can justify the cost and lead to higher total profit than partnering with ten streamers each having 100,000 followers.

For Platforms: The pricing of public domain traffic should not be uniform. To enhance overall ecosystem health and social welfare, platforms should adopt a differentiated pricing strategy, offering preferential rates to high-impact streamers. This encourages greater traffic acquisition, boosts transaction volumes, and creates a virtuous cycle of growth. A platform could implement a tiered traffic pricing system. For instance, streamers with over 10 million followers might receive a 20% discount on public traffic purchases, incentivizing them to drive more transactions and enhance overall ecosystem vitality, as indirectly supported by the exclusive partnerships between platforms and top streamers like Xinba.

For Policymakers: Encouraging transparency and fairness in manufacturer–streamer collaborations can help mitigate the risks of data fraud and contract disputes, promoting the sustainable development of the live-streaming industry. For instance, a public certification system for streamers’ follower authenticity and sales performance data could be developed. This would reduce information asymmetry between manufacturers and streamers, mitigate risks of data fraud, and create a more trustworthy ecosystem where Nash bargaining fairness can be effectively achieved.

6.3. Limitations and Future Research Directions

While this study provides valuable insights, it is not without limitations, which also present opportunities for future research.

First, the model operates under the assumption of perfect information. Future work could incorporate information asymmetry, such as the streamer’s private knowledge about their true conversion capability or the prevalence of “fake followers,” which would introduce signaling and screening dynamics.

Second, the model focuses on a single manufacturer–streamer dyad. Extending the analysis to a competitive setting with multiple streamers or manufacturers would yield insights into market structure and competitive strategy.

Finally, the social network effect is modeled primarily as a convergence mechanism. Investigating products where consumer behavior is driven by a desire for differentiation (e.g., luxury goods) would require integrating divergence effects into the model. Empirical validation of the proposed model using real-world transaction data from leading platforms represents another promising research avenue.