Short Video Marketing or Live Streaming Marketing: Choice of Marketing Strategies for Retailers

Abstract

This study investigates retailers’ strategic choices between short video marketing (SVM) and live streaming marketing (LSM) in the social media era, with a focus on the synergistic effects and decision-making mechanisms of these two digital marketing models. Using game theory, we construct a game analysis model to analyze retailers’ optimal selection among three marketing strategies: S (sole implementation of SVM), L (sole implementation of LSM), and H (integration of both SVM and LSM). The findings reveal that retailers should make different strategic choices based on the different stages of development. In the early market entry phase, characterized by both a low mixed commission rate and a low slotting fee, the H strategy emerges as the optimal choice. As the market enters its growth phase, retailers should shift to the L strategy, driven by “influencer LSM”. When the market enters a mature stage, retailers should be more inclined to adopt the S strategy or the L strategy dominated by “merchant self-LSM”. These findings provide new theoretical insights into the dynamic selection mechanisms of digital marketing strategies while offering practical decision-making guidance for retailers in allocating marketing resources across different development stages. The conclusions have direct implications for optimizing corporate marketing mix strategies.

1. Introduction

1.1. Background and Motivation

With the rapid development of social media applications, short video marketing (SVM) and live streaming marketing (LSM) models have become favored by many users. According to the China Internet Network Information Center (CNNIC) statistics, by December 2024, the number of short video users in China had reached 1.04 billion, constituting 93.8% of total internet users. During the same period, the number of live streaming users had grown to 0.83 billion, representing 75.2% of the total internet users [1]. This vast user base has considerable business potential for SVM and LSM, making the two emerging sales models play an increasingly important role in promoting product sales and market promotion [2].

The SVM model, with its powerful editing and customization functions, provides retailers with flexible and diverse product promotion methods. Retailers can customize unique advertising styles through third-party intermediaries (such as Multichannel Networks, MCNs), ingeniously integrating products into the storylines of short videos, thereby attracting consumers’ attention and enhancing their interest in the products [3]. The “little yellow cart (shopping cart)” function embedded in these short videos has greatly simplified the user’s purchasing process, achieving an immediate conversion from content viewing to product purchase, and effectively promoting the improvement of sales efficiency. Meanwhile, the social attributes of short videos, such as the like, share, and comment mechanisms, have built a two-way interactive bridge between retailers and consumers, enhancing user stickiness and brand loyalty [4]. In contrast, the LSM model has attracted a large number of consumers’ attention with its real-time interactivity and the authenticity of product display. During the live broadcast, consumers can ask the host questions in real time, and the host stimulates consumers’ purchasing intention by answering the questions and showcasing the products [5]. This direct interaction method not only enhances consumers’ sense of participation and trust but also effectively boosts the conversion rate and sales volume of the product. Statistics indicate that the TikTok platform experienced an impressive 83% growth rate in SVM transaction volume during 2023 [6], while during Taobao’s “Double 11” promotion, multiple live streaming rooms surpassed RMB 100 million in transactions [7]. These fundamental differences imply that retailers need to make strategic choices among SVM, LSM, and hybrid strategies based on their specific needs. Notably, as both SVM and LSM become increasingly mature, hybrid strategies are emerging as a new trend. However, existing studies mostly analyze a single model and fail to reveal the dynamic selection mechanism among the three strategies [8]. It should be emphasized that consumer decision-making in these three sales models inherently involves large-scale group interactions (e.g., viral diffusion of short videos or real-time feedback in live streaming rooms) [9]. Their underlying mechanisms can be quantified through social network analysis methods—such as modeling influencer impact based on user trust relationships or integrating multi-source heterogeneous review data. This provides a micro-level behavioral foundation for constructing the retailer strategy game model in subsequent sections [10].

However, while SVM and LSM strategies create significant business opportunities for retailers, they also pose complex challenges at the level of strategy selection. In both of these emerging sales models, retailers need to pay a certain percentage of commission to the MCN. In line with Huang et al.’s [11] findings, the commission rate demanded by MCNs for the services of top-tier talent can increase to 50%. Notably, there are differences in the commission rates charged by MCNS under SVM and LSM. Moreover, LSM also involves additional cost expenditures such as slotting fees [12]. Based on whether the slotting fee is zero or not, LSM can be further subdivided into two types: merchant self-LSM and influencer LSM. Merchant self-LSM refers to retailers using their own employees to carry out live streaming activities. Its characteristic lies in the fact that the slotting fee is zero. Influencer LSM, on the other hand, involves retailers hiring outstanding hosts through the MCN for live promotion. Under this model, the slotting fee is indispensable. Therefore, in this study, we adopt the criterion of whether the slotting fee is zero to distinguish between these two LSMs. When retailers opt for a hybrid strategy, the resulting cost superposition issue (e.g., simultaneously bearing short video production costs and live streaming slotting fees) further complicates strategic decision-making. In view of this, retailers need to comprehensively evaluate the advantages, limitations, and cost structures of the two sales models in the decision-making process in order to carefully select the sales model combination that best suits their own development stage. In practical business operations, brands such as Florasis, Estée Lauder, and Lancôme adopt different marketing strategies at distinct stages of their development. Taking Florasis as an example, the brand strategically combined SVM with LSM strategies in its initial market entry phase to quickly penetrate the market. Subsequently, the company further expanded brand awareness through “influencer LSM”. As the brand matured, its marketing strategy gradually shifted toward “merchant self-LSM” or promoting products through SVM. This series of changes raises the following intriguing questions: What factors influence retailers’ choices of marketing strategies? Previous studies have focused primarily on single strategies, such as examining the marketing effectiveness of SVM [13,14] or LSM [15,16,17] separately. However, this study aims to comprehensively examine the SVM strategy, the LSM strategy, and the combined application of the two strategies to fill a gap in the current research. The two factors mentioned above—namely, the commission rate and the slotting fee—are common influencing factors. Therefore, we incorporated these factors into the research.

1.2. Research Questions and Major Findings

We consider a market scenario where a platform such as TikTok exists. Its huge user base and active SVM and LSM provide retailers with diversified sales channels. Depending on whether the retailer introduces an SVM strategy, an LSM strategy, or a combination of the two, the following three strategies are derived, (1) the SVM strategy (Strategy S), (2) the LSM strategy (Strategy L), and (3) the hybrid strategy (Strategy H), which involves simultaneously employing SVM and LSM strategies for consumer sales and product promotion. Meanwhile, we consider that there are three significant differences between SVM and LSM. Firstly, in terms of the degree of consumer sensitivity to these two formats, the SVM strategy demonstrate higher appeal and acceptance due to their embedded storylines, personalized styles of creators, and frequent browsing features (such as an average daily browsing time of over 50 min for American adults [18] and 85 min for Chinese users [19]). Therefore, in this study, we assume that the sensitivity of consumers to SVM is higher than LSM. Secondly, in terms of the mechanism for promoting demand growth, the SVM strategy focuses on increasing the number of videos to enhance the viewing probability and indirectly boost sales. The LSM strategy, on the other hand, relies more on the hosts’ real-time promotional activities, directly stimulating consumers’ purchasing intentions through interaction and product recommendations. Finally, in terms of the effect of stimulating demand growth, although SVM is widely popular due to its short nature (with an average duration of about 5 min) [20,21], in terms of stimulating purchases, it is slightly inferior to LSM [22]. LSM, with its longer duration, rich product introductions, and the interactive atmosphere and promotional strategies within the live-streaming room (such as time-limited flash sales and limited supply), can more effectively enhance consumers’ purchasing enthusiasm. Therefore, in this study, we hypothesize that in the SVM mode, the number of short videos has a positive driving effect on market demand, but the marginal utility is decreasing; while in the LSM mode, the promotional effort of streamers also has a positive pulling effect on market demand, and there is a positive correlation. Based on the above analysis, this study aims to deeply explore the following issues:

(1) What are the key factors that influence retailers’ strategic choices, and how do retailers rely on these factors to make decisions?

(2) How do different strategic choices affect retailers’ pricing and demand?

(3) How does consumer sensitivity to different strategies impact the effectiveness of retailers’ strategic implementation?

Our primary findings are summarized as follows. First, as the development stage of the market changes, retailers should be flexible in adjusting their marketing strategies accordingly. In the initial market entry stage, characterized by a low mixed commission rate and a low slotting fee, retailers should embrace Strategy H and leverage the use of widespread marketing and brand-building approaches to swiftly penetrate the market. As the market enters its growth stage with an increasing mixed commission rate but a relatively low slotting fee, particularly when the LMS commission rate for influencers remains low, retailers should utilize primarily Strategy L centered on “influencer LSM”. During the market maturity stage, when both commission rates and the slotting fee are high, particularly if the slotting fee increases significantly, retailers should consider adopting Strategy S or L focused on “merchant self-LSM”. Second, Strategy H is more effective than the other two strategies in helping retailers increase product selling prices while simultaneously boosting market demand and consumer surplus. In contrast, the impact of Strategy L and Strategy S on retailers’ pricing power and market demand depends on certain conditions, specifically, which commission rate is higher and offers a greater incentive advantage. Third, during the implementation of the strategy, retailers must prioritize consumers’ sensitivity to SVM and LSM promotions, as the higher the sensitivity, the greater the pricing space for products and the increase in market demand. In addition, collaborating with influencers boasting a substantial fan base can significantly bolster the effectiveness of LMS promotions, thereby attracting more potential consumers and broadening market share.

1.3. Contribution Statement and Structure of This Paper

This study contributes to both the academic literature and practical operations in three key aspects. First, we deepen and extend existing research on short video marketing (SVM) and live streaming marketing (LSM). By constructing a game-theoretic model between retailers and multichannel networks (MCNs), we employ Nash equilibrium and mathematical analysis to derive optimal strategies for both parties while uncovering how parameter variations influence equilibrium outcomes. To our knowledge, this paper is the first to comprehensively analyze strategic choices between SVM, LSM, and hybrid strategies. By systematically examining the conditions under which these strategies succeed, we provide retailers with an actionable decision-making framework, thereby addressing a critical gap in the literature. Second, through mathematical properties of the optimal solutions (e.g., monotonicity, concavity/convexity), we reveal how retailers should select and adapt their strategies across different market development phases. Specifically, we clarify how commission rates and slotting fees dynamically shape the optimal choice of SVM, LSM, or a hybrid approach during the initial entry, growth, and maturity stages. These findings empower retailers to align their strategies with market evolution for maximum profitability. Finally, we demonstrate the pivotal role of consumer sensitivity in strategy effectiveness. Our results show that as consumer responsiveness to SVM or LSM promotions grows, retailers can raise prices, expand demand, and boost profits—regardless of the chosen strategy. This offers actionable insights for tailoring strategies to observed consumer behavior.

The subsequent sections of this paper are organized as follows. Section 2 provides a comprehensive review of the relevant literature. In Section 3, we construct a model and examine the equilibrium prices, market demand levels, and profits under different strategies. In Section 4, a comprehensive comparison is conducted, and optimal strategic recommendations and managerial insights are provided for retailers. Some extensions are explored in Section 5. Finally, Section 6 outlines the conclusions, discusses the managerial implications and limitations of this study, and suggests potential directions for future research.

2. Literature Review

The focus of our research aligns closely with the following two primary areas of the literature: (i) short video marketing and (ii) live streaming marketing.

2.1. Short Video Marketing

Recent research on short video marketing (SVM) has systematically examined its mechanisms through three primary lenses. First, content design elements have been shown to significantly drive engagement, with Xiao et al. [4] identifying four core components (content matching, information relevance, storytelling, and emotional expression) that enhance consumer interactions on TikTok, while also demonstrating the moderating role of posting timing. Complementing this, Li and Zhang [23] and Li et al. [24] established the importance of sensory appeal and interactivity in creating emotional connections and behavioral responses. Second, creator influence has been highlighted by Yuan et al. [25], who found that exceptional creators generate universal effectiveness regardless of product price, particularly with substantial content exposure. Third, studies have explored behavioral outcomes, revealing SVM’s dual impact on purchase intentions [26,27], viewer engagement [28], and sharing motivations [29,30], while also identifying potential adverse effects [31,32] that require consideration.

In summary, while these studies have provided valuable insights into SVM, they have primarily focused on empirically examining its underlying mechanisms. Although some scholars, such as Chen [33] and Lu et al. [34], have begun exploring the implications of adopting SVM strategies through modeling approaches, they have overlooked the interactive relationship between SVM and LSM. This study effectively addresses this research gap by developing an integrated model of SVM and LSM that simultaneously incorporates commission rate differentials.

2.2. Live Streaming Marketing

The robust development of live streaming marketing (LSM) has fundamentally transformed consumer shopping habits and prompted brands to refine their marketing strategies. Existing research on LSM can be categorized into two main streams: (1) streamer characteristics and consumer behavior; (2) strategic adoption and mode selection.

With respect to the streamer characteristics and consumer behavior, scholars have extensively examined how streamer attributes influence purchase decisions. Zhang et al. [35] highlight the importance of interactive methods (e.g., entertainment- and information-based engagement) in enhancing user engagement metrics. Wang et al. [36] further classify streamer–viewer interactions into relationship-oriented and transaction-oriented types, demonstrating the dominance of relationship-focused engagement. Additionally, studies explore the impact of streamer personality traits [37] and content topics [38] on consumer behavior.

With respect to the strategic adoption and mode selection, research on LSM adoption examines key influencing factors, such as spillover effects in competitive markets [39] and the role of product quality [15,40]. Meanwhile, mode selection studies compare merchant self-LSM and influencer LSM, identifying optimal strategies based on commission rates [12,16], product characteristics [41], and platform dynamics [42]. Some studies caution against universal LSM adoption, noting potential profit risks for manufacturers [43].

Despite these contributions, the existing literature has rarely investigated the synergistic optimization between SVM and LSM, nor has it adequately addressed the strategic selection mechanisms under different commission structures. This study bridges these gaps by integrating marginal utility analysis with Nash equilibrium methods, thereby providing retailers with a more comprehensive perspective for marketing channel decision-making.

2.3. Literature Review Summary

We summarize the differences between our work and the related research work in

Table 1. Differences between our work and the related research work.

Study	SVM	LSM	Dual-Channel	Influencer Contract Design	Cost Structure	Demand Form	Competitive Setting
Wang et al. [15]	×	√	×	×	×	Nonlinear	×
Chen et al. [42]	×	√	√	×	Commission + slotting fee	Linear	×
Lu et al. [17]	×	√	√	×	Commission + slotting fee	Stochastic	Dual-channel competition
Wang et al. [44]	×	√	√	√	Commission	×	Dual-channel competition
Lu et al. [34]	√	×	√	×	Commission	Stochastic	Dual-channel competition
He et al. [14]	√	×	√	×	Commission + slotting fee	Linear	Dual-channel competition
Lei et al. [45]	×	×	√	×	×	Linear	Co-opetition
Zhang et al. [12]	×	√	×	√	Commission + slotting fee	Linear	×
Yang et al. [46]	×	√	×	√	Commission + slotting fee	Nonlinear demand	×
Our work	√	√	√	×	Commission + slotting fee	Nonlinear	×

, thereby highlighting the research gaps and underscoring our contributions.

In summary, two key streams emerge from the literature: (1) SVM, primarily focusing on content dissemination and user engagement mechanisms; and (2) LSM, mainly examining how promotional efforts influence consumer purchasing behavior, determinants of live streaming sales performance, and modality selection in live commerce. However, few studies have explored the interaction between these factors, particularly in the context of marketing synergy strategies and dynamic market evolution. This study contributes to the literature by (1) expanding the research perspective from single-strategy analysis to combinatorial optimization of SVM and LSM strategies, with live streaming modalities categorized by the presence or absence of slotting fees (merchant self-LSM or influencer LSM); (2) developing stage-adaptive decision rules based on commission rate thresholds, translated into more intuitive market development phases (initial entry, growth, and maturity stages); and (3) simulating retailers’ pricing decisions, market share dynamics, and profitability variations under different channel mix strategies (S/L/H strategy), providing actionable guidance for marketing strategy selection.

3. Model

3.1. Problem Description

In this paper, we consider the existence of a short video and live streaming platform in the market, such as TikTok, which operates primarily as a social platform centered on short videos and live streaming commerce and boasts a substantial user base.

To address the research question of whether retailers should adopt an SVM model, an LSM model, or a combination of both, we outline the following three strategies for retailers, (1) a SVM strategy (Strategy S); (2) a LSM strategy (Strategy L); and (3) a hybrid strategy (Strategy H), which involves simultaneously employing both SVM and LSM strategies for product sales and promotion to consumers. The symbols and definitions presented in

Table 2. Basic symbols.

Symbol	Meaning
Decision variables
$p$	Sales price of a retailer’s product
$\alpha$	Number of short videos, measured as the total count obtained from the platform’s backend data
$s$	Promotional effort of LSM, measured as the proportion of time spent on product demonstrations
Parameters
$Q$	Potential market size
$f$	Production cost per unit of short video
$\theta$	Sensitivity coefficient of consumers to SVM
$\beta$	Sensitivity coefficient of consumers to LSM
$k_S$	Commission rate for SVM
$k_L$	Commission rate for LSM
$k_H=k_L+k_S$	Mixed commission rate
$T$	Slotting fee
Superscripts
S	Short video strategy
L	Live streaming strategy
H	Hybrid strategy
Subscripts
r	Retailer
m	MCN

are adopted in this paper. The three strategies, Strategies S, L, and H, are subsequently discussed.

In summary, this study is grounded in a game framework, assuming that the retailer and MCN form a Nash equilibrium in their strategic interactions. Both parties simultaneously optimize their decisions with profit maximization as their objective, where the concavity of profit functions ensures the uniqueness of equilibrium solutions. The model further assumes symmetric decision-making power between the parties and risk-neutral preferences, indicating their exclusive focus on expected payoffs without risk considerations.

3.2. Strategy S

When retailer adopts Strategy S, their operating model relies on short video channel. By cooperating with short video creators, the retailer delivers short videos on a large scale to achieve the growth of product exposure. To enhance the operationality of the analysis, we make the following assumption: with respect to consumer demand, based on the empirical research of Lodish et al. [47] and Schmidt and Eisend [48], advertising number has a significant positive driving effect on consumer demand, but its marginal utility follows the law of diminishing returns, that is, as the expenditure increases, the additional demand gradually weakens. Furthermore, referring to the research framework of He et al. [13] and Siqin et al. [49] (that is, the price dimension follows the classical law of demand), the market demand function under Strategy S is constructed as follows:

$$D^S=Q-p+\theta \sqrt{\alpha }$$

(1)

where $Q$ represents the potential market size, and the unit sales price $p$ has a linear negative relationship with demand, reflecting consumers’ immediate sensitivity to price changes; the number of short videos $\alpha$, on the other hand, exerts a positive driving effect through a square root function form ($\sqrt{\alpha }$), indicating diminishing marginal utility. That is, the initial increase in short video placements can efficiently activate demand, but as the number continues to grow, the additional demand generated by each unit of short video placement gradually diminishes; $\theta$ embodies the sensitivity coefficient of consumers to SVM, not only reflecting the matching efficiency between short videos and market demand but also determining the intensity of the demand pull from short video placements.

With respect to the retailer and MCN profit, the retailer and MCN form a revenue-sharing cooperation mechanism under Strategy S. Specifically, the retailer must pay a commission to the MCN responsible for hiring short video creators to produce short videos, with the commission rate designated as $k_S$. Meanwhile, the revenue structure of the MCN presents two characteristics: on the one hand, the MCN earns sales commission through the commission rate $k_S$; on the other hand, they need to bear the production costs that are linearly related to the number of short videos, including creators’ revenue sharing, content production, and equipment loss, etc. [21]. Assume that the production cost of a single short video is $f$. Based on the above mechanism design, the retailer, and the MCN profit functions under Strategy S are constructed as follows:

$${\pi }_r^S(p)=(1-k_S)p{D}^S$$

(2)

$${\pi }_m^S(\alpha )=k_{S}p{D}^S-f\alpha$$

(3)

The game sequence for Strategy S is as follows. First, the retailer chooses the optimal strategy. Second, if Strategy S is subsequently chosen, then since the retailer and the MCN possess equal power, they proceed to engage in a Nash equilibrium game (due to the sufficient number of retailers and MCNs engaged in short video and live streaming commerce in the market, we postulate that retailers and MCNs possess equal decision-making authority and engage in a Nash equilibrium game). In this game, the retailer determines the sales price $p$, whereas the MCN concurrently determines the number of advertising short videos $\alpha$.

Through the equilibrium solution, the optimal results under Strategy S can be obtained; these results are summarized in Lemma 1. See the Appendix A, Appendix B, Appendix C and Appendix D for all mathematical proofs.

Lemma 1.

Under Strategy S, the optimal sales price for the retailer is $p^S={\displaystyle \frac{2fQ}{4f-k_S{\theta }^2}}$, and the optimal number of short videos is ${\alpha }^S={\displaystyle \frac{{k_S}^2{Q}^2{\theta }^2}{{(4f-k_S{\theta }^2)}^2}}$.

On the basis of Lemma 1, the retailer’s market demand is $D^S={\displaystyle \frac{2fQ}{4f-k_s{\theta }^2}}$, and the retailer’s profit is ${\mathrm{\Pi }}_r^S={\displaystyle \frac{4(1-k_S)f^2{Q}^2}{{(4f-k_S{\theta }^2)}^2}}$ under Strategy S.

Corollary 1.

Under Strategy S, the sensitivity analysis of each parameter is as follows:

(1) ${\displaystyle \frac{\partial p^S}{\partial \theta }}={\displaystyle \frac{\partial D^S}{\partial \theta }}>0$; ${\displaystyle \frac{\partial p^S}{\partial f}}={\displaystyle \frac{\partial D^S}{\partial f}}>0$; ${\displaystyle \frac{\partial p^S}{\partial k_S}}={\displaystyle \frac{\partial D^S}{\partial k_S}}>0$.

(2) ${\displaystyle \frac{\partial {\alpha }^S}{\partial \theta }}>0$; ${\displaystyle \frac{\partial {\alpha }^S}{\partial f}}<0$; ${\displaystyle \frac{\partial {\alpha }^S}{\partial k_S}}>0$.

(3) ${\displaystyle \frac{\partial {\mathrm{\Pi }}_r^S}{\partial \theta }}>0$; ${\displaystyle \frac{\partial {\mathrm{\Pi }}_r^S}{\partial f}}<0$. When $k_S<2-{\displaystyle \frac{4f}{{\theta }^2}}$, ${\displaystyle \frac{\partial {\mathrm{\Pi }}_r^S}{\partial k_S}}>0$; when $k_S>2-{\displaystyle \frac{4f}{{\theta }^2}}$, ${\displaystyle \frac{\partial {\mathrm{\Pi }}_r^S}{\partial k_S}}<0$.

Corollary 1 shows that the sales price, the number of short videos, market demand, and the retailer’s profit increase with increasing consumer sensitivity to SVM. The number of short videos and the retailer’s profit decrease with the increasing production costs of short videos. These results are intuitive. Furthermore, as the commission rate $k_S$ increases, the sales price, number of short videos, and market demand increase. This finding suggests that despite the retailer adopting an increasing price strategy to offset cost increases, market demand remains resilient to such price hikes because of the increased number of advertising short videos (indicating intensified promotional effort). In essence, the positive impact of Strategy S on market demand surpasses the adverse effects of elevated prices and costs. Nonetheless, when $k_S$ reaches a certain threshold, the retailer’s profit begins to decline. In fact, this is a new finding. In fact, this is a novel discovery. Previous studies have shown that a low commission rate makes retailers more inclined to use the SVM mode [14,50]. However, our findings reveal that only when the commission rate is lower than a certain threshold do retailers’ profits increase as the commission rate increases. The reason might be that only when the commission rate is at a relatively low level can a higher commission be exchanged for better traffic support, advertising exposure or data services from the platform, thereby driving up sales and offsetting the commission cost. Consequently, retailers should aim to maximize consumer sensitivity to SVM and choose an appropriate commission rate to increase the profits of implementing Strategy S, rather than blindly increasing the number of short videos.

3.3. Strategy L

For Strategy L, we make the following assumption: with respect to consumer demand, when retailer adopts the L strategy, its operation model relies on the LVM and organically combines product display, interactive Q&A, and promotion incentives through the real-time shopping guide service with the contracted streamer team in the live streaming process, which directly triggers the purchase demand of consumers. According to the study of live streaming shopping behavior by [51], consumers form instantaneous purchase decisions during the process of watching live streaming, and their demand is not only linearly affected by the potential market size $Q$ and price $p$, but also driven by the streamer’s promotional effort level, represent as $s$. Based on the quantitative studies of streamer performance in [52,53,54], we assume that the market demand is positively correlated with the streamer’s promotional effort level.

$$D^L=Q-p+\beta s$$

(4)

where $\beta$ represents the sensitivity coefficient of consumers to LSM, it not only represents the matching efficiency between the professional ability of the streamer and the market demand, but also determines the intensity of the demand-pull caused by the input of the promotional effort level of each streamer.

With respect to the retailer and MCN profit, the retailer must pay a commission to the MCN responsible for hiring the influencer streamer for live streaming, with the commission rate designated as $k_L$ under Strategy L. Furthermore, a slotting fee, $T$, for live streaming is incurred, which specifically refers to the slotting fee of products within the live streaming room [12,55]. To ensure a comprehensive analysis, different scenarios, $T=0$ and $T\ne 0$, are considered. Specifically, $T=0$ implies that the retailer employs its own staff for LSM, termed “merchant self-LSM”, whereas $T\ne 0$ implies that the retailer engages influencers for LSM through an MCN, termed “influencer LSM”. The specific marketing format of live streaming is contingent on the value of $T$. The MCN profit consists of the following three parts: the slotting fee, the sales commission, and the live streaming promotional costs. Moreover, as the intensity of LSM promotional efforts escalates, the promotional cost also increases, exhibiting a marginal cost escalation trend. This cost can be approximated as a convex function of the LSM promotional effort, i.e., as ${\displaystyle \frac{1}{2}}{s}^2$ [56,57,58]. Based on the above mechanism design, the retailer and MCN profit functions under Strategy L are constructed as follows:

$${\pi }_r^L(p)=(1-k_L)p{D}^L-T$$

(5)

$${\pi }_m^L(s)=k_{L}p{D}^L+T-{\displaystyle \frac{1}{2}}{s}^2$$

(6)

The game sequence for Strategy L is similar to that of Strategy S. First, the retailer selects the optimal strategy. Second, if Strategy L is subsequently selected, then the retailer determines the sales price $p$. Concurrently, the MCN determines the promotional effort $s$ of the LSM.

Through the equilibrium solution, the optimal results under Strategy L can be obtained; these results are summarized in Lemma 2.

Lemma 2.

Under Strategy L, the retailer’s optimal selling price is $p^L={\displaystyle \frac{Q}{2-k_L{\beta }^2}}$, and the optimal LSM promotional effort is $s^L={\displaystyle \frac{k_L\beta Q}{2-k_L{\beta }^2}}$.

On the basis of Lemma 2, the retailer’s market demand is $D^L={\displaystyle \frac{Q}{2-k_L{\beta }^2}}$, and the retailer’s profit is ${\mathrm{\Pi }}_r^L={\displaystyle \frac{(1-k_L)Q^2}{{(2-k_L{\beta }^2)}^2}}-T$ under Strategy L.

Corollary 2.

Under Strategy L, the sensitivity analysis of each parameter is as follows:

(1) ${\displaystyle \frac{\partial p^L}{\partial \beta }}={\displaystyle \frac{\partial D^L}{\partial \beta }}>0$; ${\displaystyle \frac{\partial p^L}{\partial k_L}}={\displaystyle \frac{\partial D^L}{\partial k_L}}>0$.

(2) ${\displaystyle \frac{\partial s^L}{\partial \beta }}>0$; ${\displaystyle \frac{\partial s^L}{\partial k_L}}>0$.

(3) ${\displaystyle \frac{\partial {\mathrm{\Pi }}_r^L}{\partial \beta }}>0$; ${\displaystyle \frac{\partial {\mathrm{\Pi }}_r^L}{\partial k_L}}<0$.

Through Corollary 2, it is evident that the sales price, market demand, promotional effort of LSM, and retailer’s profit increase with increasing consumer sensitivity to LSM. However, in contrast to Strategy S, although the sales price, promotional effort, and market demand increase with increasing commission rate $k_L$, the retailer’s profit paradoxically decreases. This observation underscores the potential for retailers to adopt a “merchant self-LSM” approach as Strategy L to mitigate costs. By doing so, retailers may need to adjust their pricing strategies to remain competitive and attract a larger consumer base, but ultimately, this approach can facilitate the attainment of greater profits through reduced commission expenses and optimized marketing efforts.

3.4. Strategy H

For Strategy H, we make the following assumption: with respect to consumer demand, when the retailer adopts the H strategy, its operation mode is reflected in the dual-wheel-drive mechanism of SVM and LSM. On the one hand, it builds high-frequency exposure touchpoints by cooperating with short video creators, and on the other hand, it relies on the MCN’s streamer team to create a high-converting interactive scenario. Therefore, the market demand is subject to the combined influence of SVM and LSM, expressed as follows:

$$D^H=Q-p+\theta \sqrt{\alpha }+\beta s$$

(7)

Compared to Strategy S ($D^S=Q-p+\theta \sqrt{\alpha }$), the Strategy H incorporates a linear driving effect ($\beta s$) through the promotional effort $s$ of LSM, which compensates for the diminishing marginal utility of the number of short videos $\alpha$. This dual mechanism maintains long-term brand exposure while significantly enhancing immediate conversion capabilities. In contrast to Strategy L ($D^L=Q-p+\beta s$), Strategy H preserves the square root effect of short videos ($\theta \sqrt{\alpha }$). This strategic design prevents the potential cost inefficiencies arising from sole reliance on streamer efforts, thereby achieving more comprehensive market coverage with greater stability.

With respect to the retailer and MCN profit, the retailer and MCN also form a revenue-sharing partnership under Strategy H. Specifically, the retailer not only needs to pay the MCN a mixed commission $k_H=k_S+k_L$ on its sales revenue at the commission rate, but also needs to pay a slotting fee. The MCN’s revenue consists of four components: slotting fee, mixed commission, short video production cost, and live streaming promotion cost. Notably, LSM within Strategy H is further divided into the following two categories, “merchant self-LSM” and “influencer LSM”, each of which is characterized by different slotting fees, denoted as $T=0$ and $T\ne 0$, respectively. Similarly, if the retailer adopts the method of “merchant self-LSM”, then the slotting fee is zero, i.e., then $T=0$, and the MCN is actually equivalent to its own staff. Based on the above mechanism design, the retailer and MCN profit functions under Strategy H are constructed as follows:

$${\pi }_r^H(p)=(1-k_H)p{D}^H-T$$

(8)

$${\pi }_m^H(\alpha ,s)=k_{H}p{D}^H+T-f\alpha -{\displaystyle \frac{1}{2}}{s}^2$$

(9)

The game sequence for Strategy H is similar to that of Strategies L and S. First, the retailer selects the optimal sales strategy. Second, if Strategy H is subsequently selected, then the retailer determines the sales price $p$. Concurrently, the MCN determines the number of short videos $\alpha$ (i.e., the promotional effort of SVM) and the promotional effort of LSM $s$.

Through the equilibrium solution, the optimal results under Strategy H can be obtained; these results are summarized in Lemma 3.

Lemma 3.

Under Strategy H, the retailer’s optimal selling price is $p^H={\displaystyle \frac{2fQ}{2f\left(2-(k_L+k_S){\beta }^2\right)-(k_L+k_S){\theta }^2}}$, the optimal number of short videos is ${\alpha }^H={\displaystyle \frac{{(k_L+k_S)}^2{\theta }^2{Q}^2}{{\left(2f\left(2-(k_L+k_S){\beta }^2\right)-(k_L+k_S){\theta }^2\right)}^2}}$

, and the optimal live streaming promotional effort is $s^H={\displaystyle \frac{2\beta f(k_L+k_S)Q}{2f\left(2-(k_L+k_S){\beta }^2\right)-(k_L+k_S){\theta }^2}}$.

According to Lemma 3, the retailer’s market demand is $D^H={\displaystyle \frac{Q\left(2f+(k_L+k_S){\theta }^2(\theta Q-1)\right)}{2f\left(2-(k_L+k_S){\beta }^2\right)-(k_L+k_S){\theta }^2}}$, and the retailer’s profit is ${\mathrm{\Pi }}_r^H={\displaystyle \frac{2f{Q}^2\left(1-(k_L+k_S)\right)\left(2f+(k_L+k_S){\theta }^2(\theta Q-1)\right)}{{\left(2f\left(2-(k_L+k_S){\beta }^2\right)-(k_L+k_S){\theta }^2\right)}^2}}-T$ under Strategy H.

Corollary 3.

Under Strategy H, the sensitivity analysis of each parameter is as follows:

(1) ${\displaystyle \frac{\partial p^H}{\partial \theta }}>0$; ${\displaystyle \frac{\partial p^H}{\partial \beta }}>0$; ${\displaystyle \frac{\partial p^H}{\partial k_S}}={\displaystyle \frac{\partial p^H}{\partial k_L}}>0$.

(2) ${\displaystyle \frac{\partial {\alpha }^H}{\partial \theta }}>0$; ${\displaystyle \frac{\partial {\alpha }^H}{\partial \beta }}>0$; ${\displaystyle \frac{\partial {\alpha }^H}{\partial k_S}}={\displaystyle \frac{\partial {\alpha }^H}{\partial k_L}}>0$.

(3) ${\displaystyle \frac{\partial s^H}{\partial \theta }}>0$; ${\displaystyle \frac{\partial s^H}{\partial \beta }}>0$; ${\displaystyle \frac{\partial s^H}{\partial k_S}}={\displaystyle \frac{\partial s^H}{\partial k_L}}>0$.

(4) ${\displaystyle \frac{\partial D^H}{\partial \theta }}>0$; ${\displaystyle \frac{\partial D^H}{\partial \beta }}>0$; ${\displaystyle \frac{\partial D^H}{\partial k_S}}={\displaystyle \frac{\partial D^H}{\partial k_L}}>0$.

(5) ${\displaystyle \frac{\partial {\mathrm{\Pi }}_r^H}{\partial \theta }}>0$; ${\displaystyle \frac{\partial {\mathrm{\Pi }}_r^H}{\partial \beta }}>0$; when $k_L+k_S<{\displaystyle \frac{4f}{2{\beta }^{2}f+{\theta }^2(4Q\theta -5)}}$, ${\displaystyle \frac{\partial {\mathrm{\Pi }}_r^H}{\partial k_S}}={\displaystyle \frac{\partial {\mathrm{\Pi }}_r^H}{\partial k_L}}>0$; and when $k_L+k_S>{\displaystyle \frac{4f}{2{\beta }^{2}f+{\theta }^2(4Q\theta -5)}}$, ${\displaystyle \frac{\partial {\mathrm{\Pi }}_r^H}{\partial k_S}}={\displaystyle \frac{\partial {\mathrm{\Pi }}_r^H}{\partial k_L}}<0$.

Similarly to Corollary 1, Corollary 3 demonstrates that as consumers exhibit greater sensitivity to SVM and LSM, the sales price, market demand, and retailer’s profit increase. With an increase in the mixed commission rate, the sales price, market demand, LSM promotional effort, and the number of short videos (i.e., the SVM promotional effort) all exhibit an upward trend. In contrast, the retailer’s profit initially increases and subsequently decreases as the mixed commission rate continues to increase. By synthesizing the insights from Corollaries 1 and 2, it becomes evident that the SVM commission rate positively influences the retailer’s profit. In contrast, the commission rate for LSM exerts a negative influence. When the positive impact outweighs the negative impact, the retailer’s profit increases. These findings offer several strategic implications for retailers operating under Strategy H. First, retailers should strive to enhance consumer sensitivity to SVM and LSM by investing in high-quality content and engaging influencers with large followings. By doing so, retailers can capitalize on the positive effects of increased sensitivity on the sales price, market demand, and, ultimately, profit. Second, retailers must carefully manage the mixed commission rate to optimize profits. While an increase in the mixed commission rate may initially drive up the sales price, market demand, and promotional effort, it is crucial to recognize that the profit will eventually decline as the commission rate continues to rise. Therefore, retailers should aim to find the optimal balance between commission rates for SVM and LSM that maximizes profits. Third, retailers should leverage the positive impact of SVM commission rates on profits, which can be achieved by seeking short video creators with large fan bases to maximize brand exposure and sales growth at a relatively low SVM commission rate. By implementing these strategies, retailers can effectively navigate the complex dynamics of SVM and LSM commission rates to increase their overall profitability under Strategy H.

4. Strategy Comparison and Analysis

4.1. Analysis of the Impact of Different Strategies on the Retailer’s Pricing and Demand

Through a comprehensive comparison and analysis of sales prices, market demand, and promotional efforts across different strategies, we formulate Propositions 1 and 2 below.

Proposition 1.

The comparison results of sales prices and market demand under the three strategies are as follows:

(1) When $k_L/k_S>{\displaystyle \frac{{\theta }^2}{2f{\beta }^2}}$, $p^H>p^L>p^S$ and $D^H>D^L>D^S$.

(2) When $k_L/k_S<{\displaystyle \frac{{\theta }^2}{2f{\beta }^2}}$, $p^H>p^S>p^L$ and $D^H>D^S>D^L$.

(3) ${\displaystyle \frac{\partial (k_L/k_S)}{\partial {\theta }^2}}>0$, and ${\displaystyle \frac{\partial (k_L/k_S)}{\partial {\beta }^2}}<0$.

Proposition 1 reveals the specific impact of the three strategies on sales prices and market demand. First, under any circumstances, retailers that adopt Strategy H consistently achieve higher market demand while increasing the sales price. This phenomenon can be interpreted from the following two perspectives: on the one hand, the mixed method (combination of SVM and LSM) of product display provides consumers with a deeper understanding, effectively mitigating return risks and subsequently augmenting consumer purchase intentions, which grants the retailer an opportunity to augment profit margins by setting a higher selling price, and on the other hand, due to increased operating costs, the retailer is inclined to increase the sales price to sustain profitability, e.g., after adopting Strategy H, the makeup brand Florasis significantly surpassed its product sales prices and market demand levels from the implementation of Strategy L or S.

Second, the ratio of the commission rate ($k_L/k_S$) plays a pivotal role in determining the comparison of the sales prices and market demand under Strategy L versus Strategy S. Specifically, when this ratio is high, both the sales price and market demand under Strategy L surpass those under Strategy S. Conversely, when the ratio is low, the opposite holds true. This finding is different from previous studies, such as Liu et al. [16] and Pan et al. [59], which indicated that higher selling prices would suppress market demand. The underlying reason for this conclusion stems primarily from the incentive effect of a high commission rate ratio. When the commission rate for LSM is higher than that for SVM, this implies that the streamer can garner greater revenue. This incentive effect may motivate the streamer to more vigorously promote and sell products, further increasing the sales price and market demand. Taking the apparel brand Semir as an example, upon the rise in the commission rate for LSM, the brand intensified its LSM efforts. Following a collaboration with the renowned film and television star and streamer Jia Nailiang, the brand further partnered with the couple Batu & Bogu, increasing sales volume while maintaining increased prices.

Third, the threshold of the aforementioned ratio is influenced by consumer sensitivity to SVM and LSM. A higher sensitivity to SVM results in a higher threshold, whereas a higher sensitivity to LSM leads to a lower threshold. Therefore, to refine pricing strategies and marketing methods, retailers must adopt highly consumer-centric strategies and must actively listen to and analyze consumer feedback across different marketing channels. This approach enables them to tailor their marketing messages and pricing structures to better resonate with specific consumer segments, ultimately stimulating stronger purchase intentions.

Proposition 2.

The comparison results of promotional efforts under the three strategies are as follows: ${\alpha }^H>{\alpha }^S$ and $s^H>s^L$.

Proposition 2 reveals the specific impacts of the three strategies on promotional efforts. According to this proposition, the promotional effort of Strategy H surpasses that of both Strategies S and L. The superiority of Strategy H in terms of promotional effort is attributed primarily to its diverse display methods and interactivity. By integrating SVM with LSM, Strategy H not only presents product features and advantages in an interspersed storyline and intuitive manner but also facilitates real-time responses to consumers’ questions, significantly enhancing the effectiveness of promotion. Therefore, retailers should harness the advantages of mixed media to fortify product promotion and marketing endeavors. Through the amalgamation of SVM and LSM, enterprises can forge a richer, more three-dimensional product image, ultimately augmenting consumer purchase intention and loyalty.

4.2. Analysis of the Retailer’s Strategic Choices

Proposition 3.

When the slotting fee $T=0$, the retailer’s comprehensive equilibrium strategy selection is as follows:

(1) When $k_H\in \{k_H<k_{HL}\cap k_H<k_{HS}\}$, if $k_L<k_{LS}$, then ${\mathrm{\Pi }}_r^H>{\mathrm{\Pi }}_r^L>{\mathrm{\Pi }}_r^S$, whereas if $k_L>k_{LS}$, then ${\mathrm{\Pi }}_r^H>{\mathrm{\Pi }}_r^S>{\mathrm{\Pi }}_r^L$.

(2) Otherwise, if $k_L<k_{LS}$, then ${\mathrm{\Pi }}_r^L>{\mathrm{\Pi }}_r^S>{\mathrm{\Pi }}_r^H$, and if $k_L>k_{LS}$, then ${\mathrm{\Pi }}_r^S>{\mathrm{\Pi }}_r^L>{\mathrm{\Pi }}_r^H$. Among them, $k_H=k_L+k_S$.

All parameter values for numerical analysis align with real-world physical meanings and model assumptions. According to TikTok’s 2023 White Paper, production costs for a single apparel short video typically range RMB 500–2000, while marginal costs for live streaming sessions (including venue and host fees) reach RMB 5000–20,000. Based on industry profit margin analysis, we normalized parameters as follows: potential market size $Q=100$ (representing 1% sampling of 120 million active users), short video production cost $f=0.15$ (derived from the ratio of 1000 RMB/video to RMB 300 average order value). To avoid extreme results (such as explosive growth in sales along with the promotion volume) caused by excessive sensitivity coefficients in numerical simulation, we assume that the sensitivity coefficients are all less than 1. Furthermore, based on the previous text, we assume that the sensitivity coefficient of consumers to SVM is higher than that to LSM. Following Zhang et al. [12] and He et al. [14], we assume $\beta =0.15$ and $\theta =0.16$ based on the above principle, we set the parameters to $Q=100; \beta =0.15; \theta =0.16; f=0.15$. Therefore, we obtain Figure 1, which graphically presents the result of Proposition 3.

Proposition 3 reveals that when the slotting fee is zero or the form of LSM is “merchant self-LSM”, the retailer’s strategic choice is determined mainly by the level of the respective commission rate. Specifically, when the mixed commission rate ($k_H$) is at a low level, the retailer tends to choose Strategy H because it can maximize sales performance while controlling costs. However, when the mixed commission rate increases, the choice becomes more complicated. If the LSM commission rate ($k_L$) is relatively low, then the retailer is more inclined to choose the Strategy L strategy. In contrast, the retailer is more inclined to choose Strategy S to reduce the high LSM commission cost.

Let us take the brands Florasis, Estée Lauder, and Lancôme as examples. During the nascent stages of SVM and LSM, these brands opted for Strategy H because of the relatively low mixed commission rate. This strategy involves combining LSM and SVM for product sales and promotion with the aim of swiftly enhancing brand recognition and sales volume. SVM can be utilized by companies to exhibit the unique characteristics and application effects of their products, effectively capturing the attention of potential consumers. Concurrently, LSM can facilitate real-time interaction with consumers to address their inquiries and offer personalized recommendations and purchase advice. For example, Florasis initiated a short video on TikTok for the “Unloading Makeup to Reveal Face Painting” challenge that garnered significant user participation and sharing. Additionally, the brand leveraged live streaming on the platform to demonstrate various classical makeup imitation tutorials and promote related products, ultimately boosting sales performance and brand visibility. These endeavors underscore the advantages and efficacy of Strategy H in marketing. However, with the rise in the mixed commission rate, Florasis has curtailed its investment in SVM promotions on these platforms and, upon reaching the mature stage of the market, has adopted primarily the “merchant self-LSM” model for product sales and promotion.

Proposition 4.

When the slotting fee $T\ne 0$, the retailer’s comprehensive equilibrium strategy selection is as follows:

(1) When $k_H<k_{HL}$, if $0<T<T_{LS}$, then ${\mathrm{\Pi }}_r^H>{\mathrm{\Pi }}_r^L>{\mathrm{\Pi }}_r^S$; if $T_{LS}\le T\le T_{HS}$, then ${\mathrm{\Pi }}_r^H>{\mathrm{\Pi }}_r^S>{\mathrm{\Pi }}_r^L$; and if $T>T_{HS}$, then ${\mathrm{\Pi }}_r^S>{\mathrm{\Pi }}_r^H>{\mathrm{\Pi }}_r^L$.

(2) When $k_H>k_{HL}$, if $0<T<T_{HS}$, then ${\mathrm{\Pi }}_r^L>{\mathrm{\Pi }}_r^H>{\mathrm{\Pi }}_r^S$; if $T_{HS}\le T\le T_{LS}$, then ${\mathrm{\Pi }}_r^L>{\mathrm{\Pi }}_r^S>{\mathrm{\Pi }}_r^H$; and if $T>T_{LS}$, then ${\mathrm{\Pi }}_r^S>{\mathrm{\Pi }}_r^L>{\mathrm{\Pi }}_r^H$. Among them, $k_H=k_S+k_L$.

Figure 2 visually depicts the conclusions of Proposition 4 and highlights that the retailer’s strategic choices are influenced primarily by the mixed commission rate ($k_H$) and the slotting fee ($T$) when the slotting fee is nonzero or the LSM approach is “influencer LSM”. Specifically, the retailer’s preferences are manifested as follows.

(i) When both the mixed commission rate and the slotting fee are relatively low, the retailer is more inclined to choose Strategy H. This preference is driven primarily by the economic incentives provided by low costs, which facilitate broader market expansion through Strategy H. The dual stimulation of SVM and LSM often enhances consumers’ willingness to accept a slightly increasing sales price, as noted in Proposition 1. For example, during its initial market entry, the Florasis brand opted for Strategy H because of the relatively low commission rate and slotting fee associated with influencer LSM, such as those offered by Li Jiaqi. This approach involved initiating the “Unloading Makeup to Reveal Face Painting” short video challenge on the TikTok platform and conducting live sales on the Taobao live streaming platform with Li Jiaqi.

(ii) When the mixed commission rate is relatively high and the slotting fee is relatively low, the retailer tends to favor Strategy L. This is different from the research of Zhang and Zhang [60], who found that even if both the commission rate and the slotting fee are relatively high, brand owners may still adopt the “influencer LSM” strategy. We show that when the slotting fee is low, the “influencer LSM” strategy is better for retailers. This preference arises from the advantages of LSM, including strong real-time interaction, an authentic and intuitive product display, and high-level transaction conversion efficiency. These advantages make LSM an ideal choice in such scenarios. Additionally, the reduced cost burden associated with a lower slotting fee further contributes to this preference. For example, the apparel brand Semir has achieved stable sales growth through frequent collaborations with influencers such as Jia Nailiang and the couple Batu & Bogu for LSM.

(iii) When both the mixed commission rate and, in particular, the slotting fee are relatively high, the retailer is more inclined to adopt Strategy S and to rely on SVM for market promotion. This discovery is different from previous studies, such as He et al. [14] and He et al. [13]. These studies suggested that it is advisable to adopt SVM when the slotting fee and the commission rate are both low. However, considering that retailers can choose between SVM and LSM modes, we find that even when both the mixed commission rate and the slotting fee are high, retailers may still opt for the SVM mode. This preference arises because, despite the strong interactivity of LSM, the increased costs associated with higher slotting fees and LSM commission rates make SVM more attractive due to its cost effectiveness. For example, the Qumei Home Furnishing brand successfully attracted a substantial number of target customers by inviting renowned bloggers on the RED platform to create high-quality short videos.

Upon synthesizing Propositions 3 and 4, it becomes evident that the commission rates and the slotting fees are crucial factors influencing retailers’ strategic decisions. Retailers must flexibly adjust their marketing strategies on the basis of different combinations of marketing costs (slotting fee and commission rates). First, during the early phase of market entry or when both the mixed commission rate and the slotting fee are relatively low, retailers should fully capitalize on Strategy H and harness the dual benefits of SVM and LSM to maximize market expansion and promptly secure market share. Second, as the market transitions into its growth phase or when the mixed commission rate is high but the slotting fee is relatively low, particularly when the slotting fee for influencers is low, retailers should pivot toward Strategy L centered on “influencer LSM”. At this time, low-cost LSM through influencers is particularly advantageous. Finally, when the market reaches maturity or when both the mixed commission rate and the slotting fee are high, especially when the slotting fee is relatively high, retailers should increasingly rely on Strategy S or L, which focuses on “merchant self-LSM”, leveraging SVM and merchant self-LSM for market promotion to mitigate marketing costs while sustaining brand exposure and market share.

5. Expansion

5.1. Consumer Surplus and Social Welfare

In this subsection, we analyze consumer surplus and social welfare. According to Ye et al. [58], we express consumer surplus under Strategies S, L, and H as $C{S}^S={\displaystyle {\int }_{p_{\min }^S}^{p_{\max }^S}{D}^S}d{p}^S={\displaystyle \frac{{(D^S)}^2}{2}}$,$C{S}^L={\displaystyle {\int }_{p_{\min }^L}^{p_{\max }^L}{D}^L}d{p}^L={\displaystyle \frac{{(D^L)}^2}{2}}$, and $C{S}^H={\displaystyle {\int }_{p_{\min }^H}^{p_{\max }^H}{D}^H}d{p}^H={\displaystyle \frac{{(D^H)}^2}{2}}$, respectively. The corresponding social welfare levels under these strategies are denoted as $U^S={\prod }_r^S+{\prod }_m^S+C{S}^S$, $U^L={\prod }_r^L+{\prod }_m^L+C{S}^L$ and $U^H={\prod }_r^H+{\prod }_m^H+C{S}^H$.

Proposition 5.

The comparison results of consumer surplus under the three strategies are as follows:

(1) When $k_L/k_S>{\displaystyle \frac{{\theta }^2}{2f{\beta }^2}}$, $C{S}^H>C{S}^L>C{S}^S$.

(2) When $k_L/k_S<{\displaystyle \frac{{\theta }^2}{2f{\beta }^2}}$, $C{S}^H>C{S}^S>C{S}^L$.

Figure 3 graphically illustrates the findings of Proposition 5. First, Proposition 5 demonstrates that Strategy H consistently enables consumers to attain greater consumer surplus, attributed to the dual exhibition of SVM and LSM, which empowers consumers with a deeper comprehension of products, effectively mitigating return risks and subsequently augmenting consumer surplus. Second, the ratio of commission rates between LSM and SVM has a decisive influence on the comparison of consumer surplus between Strategies L and S. Specifically, when this ratio is higher, consumer surplus under Strategy L surpasses that under Strategy S. Conversely, when this ratio is lower, the opposite holds true. This phenomenon aligns with the principle outlined in Proposition 1, suggesting that a higher commission rate for LSM than for SVM can more effectively incentivize streamers and prompt them to promote and showcase products with greater vigor. Consequently, in scenarios with a higher ratio, consumers can obtain more surplus under Strategy L than under Strategy S.

Given the complexity of the results, a comparison of social welfare under the three strategies is exhibited through numerical analysis, as depicted in Figure 4. A comparison of Figure 3 and Figure 4 shows that social welfare and consumer surplus demonstrate strikingly similar outcomes. This observation implies that the magnitude of social welfare is influenced predominantly by consumer surplus rather than by supply chain profits. Consequently, augmenting consumer surplus holds substantial importance for elevating social welfare. Furthermore, the congruence between social welfare and consumer surplus underscores the critical role of consumer-centric strategies in fostering overall societal benefits. Policies and practices aimed at enhancing consumer surplus are likely to have positive spillover effects on social welfare. Therefore, enterprises and policymakers should prioritize strategies that empower consumers, such as providing detailed product information through multiple channels (e.g., SVM and LSM channels), ensuring transparency in product specifications, and facilitating easy return policies. These measures not only bolster consumer trust and satisfaction but also contribute to the elevation of social welfare.

5.2. Impact of Fan Quantity

In this subsection, we contemplate the impact of the fan quantity of short video creators and live streamers on the retailer’s strategic choices. The fan quantity of short video creators is denoted as $x_1$, and the fan quantity of live streamers is denoted as $x_2$. On the basis of observations from commercial practices, live streamers typically possess a vast fan base. For example, the renowned top live streamer Li Jiaqi boasts an astonishing 160 million fans across all platforms, and with a single LSM, on average, more than 2 million viewers. Short video creators generally have fewer fans than live streamers, marking a notable distinction between the two. Hence, we assume that $x_1<x_2$. Another significant difference lies in the fact that, despite having a larger fan base, live streamers usually do not focus on promoting a specific product during an LSM. Taking the prominent live streamer Li Jiaqi as an example, during an LSM, he may sell multiple beauty products and potentially venture into other unrelated product categories. In contrast, short video creators tend to concentrate on promoting a specific product within a single video they post. This disparity leads to different purchase rates among fans. In this subsection, we assume that the purchase rate of short video creator fans is ${\gamma }_1$ and that the purchase rate of live streamer fans is ${\gamma }_2$. On this basis, we can derive the demand functions under Strategies S and L as $D^{S^{\prime }}=Q-p+\theta \sqrt{\alpha }+{\gamma }_1{x}_1$ and $D^{L^{\prime }}=Q-p+\beta s+{\gamma }_2{x}_2$, respectively. When the retailer adopts Strategy H, which involves the use of both SVM and LSM for sales, the construction of the demand function requires special consideration of potential overlap between the fan bases of short video creators and that of live streamers. In other words, some consumers may simultaneously be followers of both the short video creator and the live streamer. To more accurately reflect this market situation, we assume that under Strategy H, the actual purchase volume generated by fans of both is $b({\gamma }_1{x}_1+{\gamma }_2{x}_2)$, where $b$ represents the nonoverlapping fan ratio and $0<b<1$. Consequently, we can derive the demand function under Strategy H as $D^{H^{\prime }}=Q-p+\theta \sqrt{\alpha }+\beta s+b({\gamma }_1{x}_1+{\gamma }_2{x}_2)$. Other assumptions remain consistent with the main model. Owing to the complexity of the results, we similarly present the outcomes through numerical analysis. The key parameters are assigned values of $b=0.8; x_1=40; x_2=70; {\gamma }_1=0.5; {\gamma }_2=0.5$, whereas the other parameters remain consistent with those of the main model. In Figure 5, the gray area represents Strategy S, the pink area represents Strategy H, and the white area represents Strategy L. To more intuitively illustrate the impact of fan quantity on the retailer’s optimal strategic choices, we outline the optimal strategic regions from the main model via dashed lines. The areas enclosed by these dashed lines carry the same implications as those in Figure 1.

Figure 5 illustrates the impact of fan quantity on the retailer’s optimal strategic choices. Upon reviewing Proposition 3, it is observed that while the results in Figure 5 align with the main model’s outcomes, significant disparities exist in the applicable scope of different strategies. Specifically, when accounting for the fan quantity of short video creators and live streamers, the applicability of Strategy L notably expands, whereas those of Strategies H and S contract. This phenomenon suggests that for retailers, leveraging the fan base of influencers becomes a strategic imperative, particularly in saturated markets where differentiation is crucial. The expansion of the applicability of Strategy L underscores the efficacy of aligning with influencers who possess a significant following because this partnership can amplify product reach and enhance brand visibility. Conversely, the contraction of the scope of Strategies H and S indicates that SVM may be a less influencer-focused strategy than LSM.

From a managerial perspective, retailers should prioritize partnerships with short video creators and live streamers who possess a large fan base. Such collaboration can significantly enhance market penetration and consumer engagement. However, in light of the varying applicability of strategies on the basis of influencer fan quantity, retailers must remain agile in their strategic approach. This includes the readiness to shift to influencer-centric strategies when the market context dictates the need for such a shift. While leveraging influencer fan quantity is important, retailers should also ensure that their overall marketing strategy integrates complementary tactics. This holistic approach can optimize market impact and customer acquisition. For example, as the main model indicates, partnering with short video creators with a substantial fan base and adopting Strategy S can lead to unexpected commercial success when new products are introduced or expanded into new markets.

6. Results

In this study, we examine a market scenario where a prominent short video and live streaming platform, akin to TikTok or RED, exists. The platform focuses primarily on SVM and LSM, boasts a substantial user base, and engages in commerce through short videos and live streaming. The ecosystem of this platform includes consumers (users), a third-party intermediary MCN (that supplies live streamers and short video creators), and a brand retailer. To provide strategic insights for retailers, we construct and analyze the following three distinct scenarios: S (sole implementation of SVM), L (sole implementation of LSM), and H (integration of both SVM and LSM). The main findings are as follows:

(1) The retailer’s strategic choices should vary across different stages of market development. Specifically, in the initial market entry phase or when both the mixed commission rate and the slotting fee are maintained at lower levels, retailers should actively adopt Strategy H. During the market growth phase or when the mixed commission rate is high but the slotting fee is relatively low, particularly when the commission rate for LSM influencers remains low, retailers should prioritize Strategy L centered on “influencer LSM” in a timely manner. In the market maturity phase, or when both the mixed commission rate and the slotting fee escalate to higher levels, especially with a notable rise in the slotting fee, retailers should lean more toward adopting Strategy S or L focused on “merchant self-LSM”.

(2) In terms of the relationship between price and demand, Strategy H, which integrates the advantages of SVM and LSM, is always more beneficial than are the other two strategies for retailers because it enables them to increase product selling prices while simultaneously increasing market demand and consumer surplus.

(3) Retailers need to closely monitor consumer feedback to different marketing approaches, namely, consumer sensitivity to SVM and LSM activities. The reason for this is that greater consumer sensitivity to a particular marketing method provides retailers with greater pricing power for their products under that strategy.

In summary, this paper proposes the following suggestions.

(1) For Retailers:

Retailers should dynamically adjust their marketing strategies according to market development stages. In the initial phase, adopting a hybrid strategy (Strategy H) combining SVM and LSM can help quickly penetrate the market. During the growth stage, shifting focus to influencer live streaming (Strategy L) can expand sales scale, while in the mature stage, merchant self-streaming or focusing on SVM (S strategy) helps optimize costs. Retailers must closely monitor consumer feedback to different formats, optimize content to improve conversion rates, and carefully select creators/streamers with large, engaged audiences while balancing commission and slotting fee to ensure cost effectiveness. Leveraging platform analytics tools to evaluate performance and continuously refine strategies is key to maintaining competitive advantage.

(2) For Platforms:

Platforms should develop differentiated commission policies tailored to different development stages—offering preferential rates to encourage hybrid marketing adoption initially, then optimizing live streaming commission structures during growth phases to enhance profitability. Strengthening data analytics capabilities to provide merchants with precise traffic insights and performance metrics is equally important. Additionally, platforms should establish incentive mechanisms (e.g., traffic support, commission discounts) to encourage high-quality content creation, thereby improving overall ecosystem quality while increasing user engagement and monetization potential.

(3) For Short Video Creators and Live Streamers:

Content professionals should focus on quality enhancement—crafting creative, story-driven short videos while developing strong interactive demonstration skills for live streams to better engage target audiences. Building and nurturing fan communities through regular interactions and exclusive benefits is crucial for establishing a solid foundation for commercial collaborations. Proactively providing brands with audience insights (e.g., user profiles, peak traffic periods) helps establish long-term partnerships that benefit both content monetization and brand growth. Staying attuned to platform policy changes and promptly adjusting content direction ensures continued access to traffic support opportunities.

Theoretically, this study contributes to the marketing literature by incorporating SVM, LSM, and their hybrid strategies into a unified analytical framework, addressing the current research gap where most studies focus solely on single strategies. Through modeling and comparing the pricing, demand and profit differences among these three strategies, we provide retailers with dynamic strategy selection guidelines based on different development stages.

This study has several limitations that suggest directions for future research. First, our model assumes deterministic consumer behavior; future studies could incorporate random utility models to analyze behavioral uncertainty more thoroughly. Second, while we conducted simulation analysis through theoretical modeling, empirical verification through case studies on major platforms like TikTok or RED could assess whether simulation results align with real-world enterprise behavior. Finally, although we calibrated key parameters using industry reports, the lack of longitudinal consumer behavior data limits our ability to validate the model’s dynamic predictive performance. Future research should employ panel data or field experiments to further test the theoretical model’s applicability.