How to Survive in the Shadow of Sharing Economy Giants: Business Model Innovation for Small and Medium-Sized Platforms

This study aims to explore how small and medium-sized platforms in the sharing economy can gain a competitive advantage in a market monopolized by giants. Taking China’s ride-hailing industry as an example, the k-means clustering method was employed to compare and analyze the characteristics and differences of 207 small and medium-sized platforms based on the framework of the business model canvas. An empirical typology comprising four representative successful business models of small and medium-sized ride-hailing platforms is extracted from the dimensions of key partners, value propositions, customer segmentation, and cost-revenue structure: high-end business platforms, Minibus-hailing platforms, Intercity carpool platforms, and Aggregation mode platforms. The results show that small and medium-sized platforms in the sharing economy can rely on their own characteristics and advantages to seek new potential development directions and gain development space by constructing value propositions and operating models that are different from industry giants. Specifically, it is necessary to further implement refined market identification and segmentation to continuously develop differentiated competitive advantages. In particular, more effective and open business strategies should be explored and practiced to greatly expand the network effects and business boundaries of small and medium-sized platforms, thereby significantly enhancing their competitiveness in the ride-hailing market. This study provides a new perspective on the reconstruction of market exchange in the sharing economy through the prism of small and medium-sized digital platforms and has certain important theoretical significance for supplementing and advancing existing research.


Introduction
In the past decade, the rapid development of digital economy has brought great changes to consumer patterns and concepts, while also continuously promoting enterprises to reshape their business models and build core competitiveness.Driven by digital technology, traditional business functions are being restructured, while new functions are being created (Peric et al., 2017;Rachinger et al., 2019).The organizational structure, profit model, and competitive strategy of enterprises are also undergoing significant changes (Haaker et al., 2021;Matarazzo et al., 2021).In the field of business and management, the business model of enterprises under the context of digitalization has been widely concerned by scholars and practitioners.However, these business model development and innovation practice studies mainly aim at large-sized enterprises, while there is a lack of attention to small and medium-sized enterprises (SMEs), which account for the majority of enterprises in the market (Andersen et al., 2022;Cosenz & Bivona, 2021;Pucihar et al., 2019).Referring to the business model canvas, M€ uller (2019) discussed the business model innovation (BMI) activities of SMEs in the context of Industry 4.0.Through a case study, Andersen  2022) analyzed the BMI process in the digital transformation of SEMs and identified four key BMI process activities.Pucihar et al. (2019) empirically examined the drivers and outcomes of BMI undertaken by European SMEs to maintain competitive in the digital economy.Considering the radicalness of digital transformation and the openness of SMEs to external partnerships, Albats et al. (2021) studied the business model transformation of small and medium-sized enterprises from a process perspective using a two-dimensional typology of BMI.Based on the inherent characteristics of SMEs, Cosenz and Bivona (2021) attempted to use dynamic business modeling methods to tailor lean strategic design tools for BMI activities in SMEs.As an important component of the economy, the BMI of SMEs driven by the digital wave still needs further exploration and research.
Digitalization has promoted the continuous emergence of various forms of market exchange reconfiguration.Among them, one of the most striking is the sharing economy which has brought revolutionary changes to the conceptualization of market exchange.With the great business success of emerging representative start-ups such as Uber and Airbnb, a large number of sharing economy platforms have sprung up around the world, which has aroused widespread concern in the academic, business, and policy fields.Scholars have conducted extensive and in-depth studies on this new economic model from multiple angles.Some scholars discussed the positive side of the sharing economy such as economic prosperity, employment and income, consumer behavior, and the sustainability of resources, environment, and society (Ciulli & Kolk, 2019;Guo et al., 2020;Martı´nez-Gonza´lez et al., 2021;Mensah et al., 2019;Revinova et al., 2020;Srovnalı´kova´et al., 2020), while others have paid attention to the negative impact brought about by the development of sharing economy such as discrimination, labor right, trust and safety, industry monopoly, opaque pricing strategies, and other governance issues (Collier et al., 2018;Edelman & Geradin, 2015;Enochsson et al., 2021;Frenken & Schor, 2017;Li & Canelles, 2021;Soltysova & Modrak, 2020).
In addition, the research on the sharing economy business model innovation is also an important and hot research topic.Mun˜oz and Cohen (2017) extracted an analytical framework for the business model of the sharing economy, including seven dimensions of Funding, Governance, Interaction, Mission, Platform, Resources, and Technology, and further developed five ideal types of sharing economy business models, namely, collaborative consumption, crowd-based tech, space-based low-tech sharing, business-to-crowd, and sharing outliers.Later, they also proposed a business model classification framework consisting of six dimensions (i.e., technology, type of transaction, business approach, shared resources, governance model, and platform type) for the sharing economy, called the Sharing Business Model Compass, which helps elucidate the various forms of BMI being adopted by the sharing economy platforms (Mun˜oz & Cohen, 2018).Netter et al. (2019) constructed a new theory-driven framework that includes three dimensions of ownership structure, related participants, and core organizational characteristics to describe and analyze the business-to-consumer and peer-to-peer sharing models.Based on a literature review of the definition and classification of business models and specific sharing activities of the sharing economy, M. Ritter and Schanz (2019) proposed a comprehensive business model classification framework and divided the sharing economy into four market segments, namely singular transaction models, subscription-based models, commission-based platforms, and unlimited platforms.Jiang et al. (2021) summarized four types of business model configurations of sharing economy that can achieve high performance, including productbased, labor-based, knowledge-based, and leasebased.In order to better distinguish value configurations of sharing-based business models, Reuschl et al. (2021) constructed a two-dimensional framework of customization versus standardization of shared goods and the centralization versus particularization of property rights over the shared goods.Based on the marketing structures and the categories of sharing economy activities (Sun et al., 2022) developed a typological configuration that characterizes the full spectrum of sharing economy business practice in the real world, which comprise seven categories of the collaborative support platform, resource supply platform, authentic C2C platform, C2C mutualized mobility platform, hybrid service platform, B2C service platforms, collaborative finance platform.
Although scholars have conducted exploratory research on the sharing economy business model innovation from different perspectives and made some meaningful and important contributions, there are still deficiencies that need that need to be further supplemented and improved.Firstly, almost all of the typological studies of sharing economy business models are based on diversified or even the full spectrum of business practices of sharing economy.But there is very little research targeting a particular industry and specific category of business cases.However, this kind of subdivision research can provide more specific guidelines and valuable references for the business practices of the sharing economy platform, and even the sustainable development of the sharing economy as a whole.In addition, existing studies on the sharing economy business model mainly focus on giants or representative platforms in a monopoly position of the sector (e.g., Uber and Airbnb), which occupy the majority of the market share and operate in a privileged manner in the large urban centers.Little is known about the huge number of small and mediumsized sharing platforms in the market so far, although they are also a very important part of the sharing economy.It is unclear how these small and mediumsized sharing platforms contribute to the reconfiguration of merchant exchange and articulate advances leading to a reconfiguration of the market exchange.In particular, there is a very important topic worthy of in-depth research, that is, how these small and medium-sized enterprises (SMEs) in the sharing economy can gain a competitive advantage in a market monopolized by giants?More specifically, this is also a business model research question, that is, what are the different characteristics of small and medium-sized sharing economy platforms compared with those industry giants?For example, who are their customers, what values can they provide to the clients, and how to make profits?Therefore, in order to fill these research gaps, this article uses business model analysis tools to conduct case studies on the Chinese ride-hailing market.Specifically, it tries to present a clear picture reflecting the characteristics and initiatives of the small and medium-sized platforms in the context of a specific sector of the sharing economy dominated by giants and also take a new look at the reconfiguration of market exchanges in the sharing economy through the prism of small and medium organizations.
The rest of this article is organized as follows.Section 2 provides the materials and methods used in this study.Next, Section 3 and 4 present the results and discussion of the empirical analysis, respectively.Finally, section 5 wraps up the paper with the conclusions, implications, limitation and future research directions.

Business Model Canvas
A business model is a conceptual tool used to illustrate the business logic of a particular entity, which includes a series of elements and relationships (Geissdoerfer et al., 2018;T. Ritter & Lettl, 2018).It describes the value an enterprise can provide to its customers, as well as the internal structure, partner network, and relationship capital that the enterprise uses to achieve value creation and delivery and generate sustainable profits (Geissdoerfer et al., 2018;Lanzolla & Markides, 2021;Rachinger et al., 2019).
The ''Business Model Canvas'' (BMC) is considered to be one of the most effective tools for analyzing a company's business strategy, organizational structure, profit model, and competitive advantage (Joyce & Paquin, 2016).As shown in Figure 1, based on the nine building blocks, the business model canvas provides a unique framework, depicting the internal and external forces related to the business and operations of the Source.Osterwalder and Pigneur (2010).
enterprise, which enables stakeholders to quickly and clearly understand the operation mode and value creation process of the enterprise (Carter & Carter, 2020;Frick & Murshid Mikael, 2013;A. Osterwalder & Euchner, 2019).Therefore, considering its readability and practicality, this article uses BMC as the theoretical framework for empirical analysis.

Ride-Hailing Platform in China
The ride-hailing platform, as a link between supply and demand, enables the riders and drivers to match quickly and complete the transaction through the establishment of a series of mechanisms such as mobile LBS (i.e., Location-Based Services) application, dynamic algorithm and pricing, and mutual evaluation system of both sides (Clewlow & Mishra, 2017).The rise of the sharing economy has promoted the rapid development of ride-hailing platforms across the world.Some of them have made remarkable achievements, such as Uber.Inspired by the success of Uber, hundreds of ride-hailing platforms have emerged successively in China since 2010 (China State Information Center, 2018).With the support of venture capital, the entire ride-hailing service industry presents a fiercely competitive situation for years.The major platforms companies have invested billions of dollars in sustained large-scale ''cash-burning'' wars to seize market share.In 2016, with the merger of China Uber and DiDi, DiDi has become the largest car-sharing platform in China.At present, there are still hundreds of online ride-hailing companies in China.Among them, Didi occupied about 90% of the market (Finance, 2018).The rest of the ride-hailing platforms are all small and medium-sized enterprises.Since 2017, Didi has been continuously expanding its competitive advantage.Meanwhile, there are also many small and medium-sized platforms with the same operating model as Didi, which compete with Didi in cities at all levels.Due to fierce market competition, this type of platform performed poorly both in terms of scale and profitability, and most of them eventually withdrew from the market.However, there are still some small and medium-sized platforms that have continued to grow and expand in the ride-hailing market.Therefore, this study selected these well-operated small and medium-sized platforms as examples, and compared and analyzed the differences in business models between them and industry giant Didi with the help of BMC tools.

Data and Samples
The analysis presented in this study is based on a database containing the business practice information of 207 small and medium-sized ride-hailing platforms in China, which covers almost all types of ride-hailing enterprises in China.The specific process of sample data collection is as follows.In order to make the selected sample representative, all the RHPs are from the ride-hailing regulatory information exchange platform sponsored by the China Communications and Information Center, which is not only responsible for the qualification review, license issuance, market supervision, and assessment of the national ride-hailing platforms but also responsible for the collection, exchange, and sharing of operational data of nationwide ride-hailing platforms.The initial identification of individual platforms was completed using web search engines, media content analysis, and information gleaned from industry statistical reports, and field research.And then we summarized and analyzed the operation modes and the development process of these platforms.
All samples are licensed ride-hailing platforms in China from 2017 to 2022.These platforms have operated and developed well over the past 6 years.The operating status of all platforms is active.In addition, we differentiate the size of the platform based on the size of the fleet.RHP with a scale of less than3,000 vehicles is defined as a small platform, while RHP with a scale of between3,000 and 10,000 vehicles is defined as a medium-sized platform (Prospective Industry Research Institute, 2022).

Analysis Framework and Techniques
In order to facilitate a systematic analysis of the business models of successful small and medium-sized RHPs in China, this article attempts to develop a comprehensive typological configuration using an inductive and quantitative-oriented approach, which encompasses a set of characteristics of RHPs' business practices based on the BMC framework.Typology is the study of types or the systematic classification of the types of something according to their common characteristics (Robinson & Bennett, 1995;Wongkit & McKercher, 2013), which has been extensively developed in various scientific fields (Henry et al., 2020;Kaczam et al., 2022;Wongkit & McKercher, 2013).Clustering is a process of classifying objects (data or samples) into different groups or clusters according to a series of characteristics or dimensions, which makes objects in the same cluster have great similarities, while objects in different clusters have great differences (Blashfield & Aldenderfer, 1978;Kettenring, 2006).As a scientific and effective quantitative analysis method, cluster analysis has been widely used in market research and business practice typology (Blashfield & Aldenderfer, 1978;Ertz et al., 2019;France & Ghose, 2019;Helmus et al., 2020).
This article employs clustering analysis to develop a typological configuration that characterizes the successful business practice of small and medium-sized RHPs in China.First, we identified and selected representative RHPs and build a database of research samples to cover the successful RHP business practice as much as possible.Second, we identified and selected characteristics and dimensions for cluster analysis.Third, based on the database, we applied cluster analysis to derive a typology for successful small and medium-sized RHPs business practices.
In this study, we used five dimensions for cluster analysis, namely, key partners, value proposition, customer segments, cost structure, and revenue streams.For each selected BMC dimension, there are several corresponding sub-dimensions that are used to describe the characteristics of the sample.A total of 17 sub-dimensions are used in the cluster analysis of this study.The distribution of samples in different dimensions is presented in Table 1.
Before clustering analysis, the encoding process for sample data was carried out by two independent coders.The coding was binary for each subdimension (0 = no, 1 = yes) in Table 2 (Ertz et al., 2019;Magdalina & Bouzaima, 2021).The controversial coding items were settled through discussion by the two coders (Ertz et al., 2019;Sun et al., 2022).After completing the encoding process, the k-means clustering algorithm was used to perform clustering analysis.As a classic unsupervised machine learning algorithm, it is essentially a process of solving the classification problem through iterative computation.The specific steps are as follows.It first randomly selects k objects as the initial clustering center (i.e., seed cluster center).Then, based on the calculation results of the distance between each object and each seed cluster center, each object will assigned to the cluster center (CC) closest to it (Wilkin & Huang, 2007).After each assignment, the CC of the cluster will be recalculated based on the existing objects in the cluster (Ahmed et al., 2020;Ralambondrainy, 1995).This process will be repeated until a certain termination condition is met.The termination condition can be that no objects are reassigned to different clusters, the cluster center no longer changes, or the sum of squared errors is locally minimized (Bishop, 2006;Wu, 2012).
Considering that this clustering technique can provide optimal intra-cluster homogeneity and inter-cluster heterogeneity, we selected it for the taxonomy study (Likas et al., 2003;Ralambondrainy, 1995;Wilkin & Huang, 2007).Software SPSS 28.0 and Python programming were employed to conduct the entire clustering analysis process.Additionally, the elbow method was used to determine the ideal number of clusters (i.e., k) (Bholowalia & Arvind, 2014;Mehar et al., 2013;Pham et al., 2005).

Cluster Analysis Results
According to the results of elbow method (see Appendix A), this study chooses k = 4 as the number of clusters.Combining the results of cluster analysis and the business practices of the small and medium-sized platforms, we summarized four types of RHPs : the high-end business ride-hailing platform (HBP) designed for the high-end business crowd in the first and second-tier cities (Cluster A), online minibus-hailing platform (MHP) which provides the commuting service for ordinary office workers who live in the suburbs but work in the city center (Cluster B), intercity carpool platform(ICP), which provides flexible and convenient inter-city travel services for people with inter-city travel needs (Cluster C), and the platform with aggregation mode (Cluster D).The specific characteristics of each cluster are shown are shown in Table 2.
The general statistics of cluster analysis are shown in Appendix B. It can be seen that the chi-square test between each variable and the four-cluster grouping was highly significant.The Cramer's V value, which measures the strength of the association between descriptive variables and cluster groups, ranges from 0.7 to 1.0, indicating a strong relationship.In addition, the distribution of the variables within each cluster is very diversified, further suggesting relative intragroup homogeneity and intergroup heterogeneity.These results demonstrate the robustness and validity of the typology of the business model of small and medium-sized RHPs in this study.Therefore, the cluster analysis results can be used for further analysis.

Comparative Analysis Between Different Type of Platforms
Four types of small and medium-sized platform business models were extracted through cluster analysis.With the help of BMC, the business models of Didi and different categories of small and medium-sized ride-hailing platforms were analyzed from five dimensions (i.e., key partners, value propositions, customer segmentation, and cost-revenue structure).The results are as follows.
Business Model Canvas of Didi.Combining Didi's growth history and actual operating conditions, we briefly summarize Didi's business model and map it to the BMC framework.The specific results are shown in Figure 2.
Didi has three important types of key partners.First, Vehicles and drivers are the most important components of the ride-hailing platform's supply side, which help deliver the value proposition to the end customers.The drivers with the private-owned car are the main components of the platform supply side.Second, technology partners facilitate supply and demand sides to complete transactions efficiently, thus helping the platform achieve its unique value proposition.These technology partners include map navigation systems, third-party payment platforms, and mobile terminal service operators, etc.Investors and venture capitalists are the most powerful supporters and the most important source of funds during the growth of the platform., especially at the start-up stage of the development.
As to the key activities, the scale effect is the key ingredient of platform businesses to achieve competitive advantage.Therefore, the key activities should be to continuously expand the scale of both supply and demand sides of the platform.At the same time, the platform should constantly reduce negative externalizes to create a favorable external environment such as government support and public acceptance.In addition, the platform should continue to improve the value proposition and seek complementary value propositions to better meet customer needs and attract more participants to join the platform.Last but not least, the platform should enhance technological lead and intellectual property to steepen barriers of entry, and also expand the market (e.g., expand to more cities) and business scope (e.g., involved in food delivery, logistics, travel services, and other businesses) to seek new growth space and profit growth point.
Key Resources.The master resource of the platform is the scale effects that need to be built and nurtured.In the age of big data, the data, algorithms, and the capability to analyze and gain insights are essential for enterprises, especially for those digital platforms.With the continuous expansion of the platform scale, a large amount of data and information about users, drivers, and vehicles will be generated and accumulated, which will be the most valuable wealth of the platform.In addition, the capabilities of the platform innovation and development are also very important, which can support the platform to continuously expand the business scope and business model innovation to achieve sustained growth.The brand effect is one of the most important intangible assets of an enterprise, so it is also a key resource in the process of platform development.
Value Proposition.A ride-hailing platform is a multisided platform and as such it has to have a value proposition to both sides, the passengers as well as the drivers.For riders, the value propositions are to meet customers' demand for car travel, such as providing personalized and diversified, low price, convenience, fast pick-up security, and safety service.For drivers, it's the opportunity to earn extra income and the freedom of choosing their work hours.
Customer relationships include establishing a good brand image, providing transparent pricing and route, maintaining a safe and secure transaction environment, building a Review, Rating & Feedback system, and enhancing interaction through social network platforms (e.g., WeChat, Weibo).
Channels of Didi for initial awareness and market cultivation are usually promotional campaigns (e.g., subsidies.free vouchers), social media and advertisement, word of mouth, and mobile apps.
Customer Segments.There are some of the more important and meaningful segmentation as follows: demographic feature (e.g., age, socio-economic status, family status); geographic feature, that is market area (e.g., a big city or small city, suburb); consumer behavior and consumption habits (e.g., preference to price or preference to service quality, frequency of usage).Considering the agglomeration effect and scale effect of consumption, the major market area of Didi is inside the first and second-tier cities, and it mainly focuses on the middleand low-end consumer groups (e.g., commuters, users of daily traveling business travel).
Cost structure.For Didi, the biggest cost element at the early stage of development is the user and driver acquisition costs, that is ''subsidies'' on both sides the drivers and the passengers.In addition, there are also some major types of operating expenses such as technological infrastructure and workspace, cost of platform operation management (e.g., human resource cost), R&D investment, marketing cost, and other business investments.
Revenue streams mainly include commissions from ride-hailing services, Ad revenue from other companies (display ads in cars and mobile apps), capital management, and other business cooperation revenue.
High-End Business Ride-Hailing Platform.As the price of the online ride-hailing service is generally lower than 10% to 25% of the price of normal urban taxis, this has greatly stimulated people's consumer demand, especially the middle and low-end consumer groups, thereby enabling the platform to quickly build a user base.Didi has grown rapidly by targeting mass consumers who expect cheap and convenient services.Unlike Didi, HBP is designed for the high-end business crowd in first and second-tier cities.It is also usually used for important business meetings, special celebrations, etc.
We use the normal urban taxi price as a baseline value and observe the impact of price changes on the scale of users.The result of the price sensitivity analysis of the users in China's ride-hailing market is shown in Figure 3.
With the gradual increase in the premium rate, the scale of users has gradually decreased.The target users of HBP are those who can accept a premium rate of over 40% and even exceed 100%.It adopts a different driver and vehicle management mode from Didi.Different from the C2C operation model adopted by Didi, these platforms adopt the B2C model, that is, these platforms have their own fleets and professional driver teams.The service standards and safety standards are much higher than those of ordinary car-sharing platforms.Compared to Didi, it has the following advantages.
First, it has a more stable user base, while Didi's user base is susceptible to price.This small group of users has very low-price sensitivity, pays more attention to service quality, and has high user stickiness.
Second, the potential safety hazard caused by the lax examination of drivers' qualifications has become an important factor affecting the development of China's ride-hailing market.For example, Didi has been affected by a series of negative events recently, such as beating passengers and sexual harassment (Prospective Industry Research Institute, 2022).The provision of full-time drivers can greatly improve the safety of the service, especially for those who wish to obtain high-quality and safe travel.
Third, in this market segment, some platforms are derivative brands launched by well-known auto manufacturers and the transformation of traditional car leasing companies, which have inherent advantages in terms of vehicle supply, cost, and brand effect(Analysys International, 2019;Baijiahao, 2018;China State Information Center, 2019;Shouqi, 2020).At present, the high-end business ride-hailing service is still a blank incremental market in China.With the continuous improvement of economic level and consumption capacity, such platforms will have broader development space.
Minibus-Hailing Platform.With the serious imbalance between the supply and demand of public transportation in lots of China's big cities, the market demand for online ride-hailing is still huge.Take Beijing for example, the ratio of the number of public travelers to the number of available seats on public transport is 100: 5 (BBS, 2020; BTDRI, 2020).Most commuters travel by crowded buses and subways, and the user experience is very poor (Baidu, 2019).Coupled with the automobile purchase restriction policies in many large cities, more people who need private cars cannot purchase vehicles due to policy reasons, hence they choose online ride-hailing as an inevitable way to travel.However, due to the limitation of vehicle scale and urban space, the transport capacity of the normal ride-hailing platforms (i.e., Didi) is very limited, especially during the rush hours in the morning and evening.In this case, the online minibus-hailing platform (MHP) appeared.
Its main user group is ordinary office workers who live in the suburbs but work in the city center.They use mobile apps for online booking, seat selection, payment, and other operations (Rainbow Bus, 2020).The routes and modes of MHP are similar to the public buses.But they have fewer stops, which can basically achieve one stop from home to the office.The number of seats provided by minibuses is generally about 9 to 15, while the average number of passengers per vehicle of Didi is about 2 (Didi, 2016).And the price is also cheaper than Didi.Therefore, it can better meet the travel demand commuters and relieve the pressure of traffic congestion to a large extent, especially during rush hours.In addition, it's more comfortable than a public bus and safer than an ordinary ride-hailing platform due to having more passengers in a car.In the past 2 years, many minibus-hailing platforms such as the Rainbow Bus, the Smooth Home Bus, and the Shuttle Bus, have grown up quickly.They focus on the route from the densely populated suburb to the city center, providing commuting services for commuters.
Compared to Didi, the vehicles are belonging to the platform, and the vehicles and drivers are controllable, so it can provide a stable service volume.In addition, as China is vigorously promoting new energy vehicles currently, most platforms are introducing new energy vehicles.This provides good external conditions for the development of the mini platform.On the one hand, the platform has received government policy support and capital subsidies.On the one hand, new energy vehicle manufacturers also actively participate in various forms of cooperation such as joint development, which has greatly reduced the platform's operating costs.Therefore, there will be a broader development space for the minibus-hailing platforms.
Intercity Carpool Platform.Ride-hailing service in urban areas has been relatively mature and has even become an indispensable part of citizens' life, especially in first and second-tier cities.However, the transportation between the city and rural areas, as well as between third-, fourth-, and fifth-tier cities, still mainly depends on traditional transportation modes such as intercity buses and trains.Due to the uneven development of Chinese cities, the distribution of transportation resources is also extremely uneven (CNBS, 2019).The transportation capacity of fourth-, and fifth-tier cities are difficult to meet people's increasingly diversified travel needs.In particular, due to the large population, when it comes to holidays and especially the Spring Festival, transportation will become a big problem.With the rapid development of ride-hailing in big cities, the emergence of intercity carpool platforms provides a new way to alleviate this problem.China is striving to promote urbanization and integration of urban and rural areas, and traffic between cities is becoming more and more frequent.The main features of the intercity carpool platform are as follows.
First, it provides a door-to-door shuttle service.Within a certain area, it can realize point-to-point and door-to-door pick-up and drop-off services, reducing the pain of passenger transfer and waiting.Specifically, it allows passengers to reduce the hassle of taking buses, subways, or taxis and directly transports them from home to their destination.
Secondly, the cost of the intercity carpool service is relatively low, which is basically the same as the cost of traditional intercity buses.In addition, compared with the traditional intercity buses which have fixed schedules every day, the operating time of the intercity carpool platform is more flexible.
These small platforms have strong regional characteristics.They operate very flexibly and have a good user base in certain areas.There are more than 300 third-, fourth-, and fifth-tier cities in China, with a population of about 500 million (CNBS, 2019).Although the markets are smaller and more fragmented, with the rapid progress of China's urbanization process, city dwellers' differentiated and diversified transportation needs will gradually increase, which undoubtedly brings huge development space for such platforms.
Aggregation Mode.The platform under the aggregation model is those small and medium-sized platforms with the same business model as Didi.These platforms have a limited number of users and insufficient platform network traffic.In order to gain living space in the market monopolized by Didi, they form an aggregation model by opening their own apps to cooperative platforms.In the aggregation mode, consumers do not need to switch between different Apps when submitting travel requirements.These platform apps are connected to each other.Users only need to select vehicles on different platforms Sun et al. in the same app interface according to their own needs (vehicle type, departure time, price, origin, and destination) (see Figure 4).This can greatly simplify the user's appointment and ordering procedures to provide users with a more convenient travel experience.At the same time, the network effect of the platform at both ends of supply and demand can be greatly expanded.This aggregation mode is based on a series of cooperation agreements and frameworks between platforms to operate normally, such as information and resource-sharing mechanisms, and profit-sharing modes.The sharing economy emphasizes the scale effect.Metcalfe's law indicates that the value of the Internet platform is proportional to the square of the number of network users.This way of forming an alliance can greatly expand the commercial boundaries of small and medium-sized ridehailing platforms, thereby significantly enhancing their competitiveness in the ride-hailing market.

Discussion
Through the analysis of the characteristics and operating modes of the five types of platforms, we summarized some significantly different parts of the business model canvas, as shown in Table 3.
Due to the huge population size and high economic level, the online ride-hailing platform is more likely to achieve scale effects in large cities.This is the market area where ride-hailing platform companies such as Uber and Didi were originally born and grown, and it is also the most competitive area in the online ride-hailing market.Didi's main market area is first-and second-tier cities and its value proposition is to provide low-cost and convenient car services to mass consumers.After years of fierce market competition, Didi has 90% of the market share and become the superpower in China's ride-hailing market.Currently, Didi remains focused on first-and second-tier cities and is gradually expanding into other big cities around the world to improve its scale advantage.Due to constant market competition, mergers, and acquisition, a large number of ride-hailing platforms of the same type as Didi has to withdraw from the Chinese market because they cannot achieve the scale effects.However, some small and medium-sized platforms can still find market opportunities and development space through detailed market investigation and analysis.Through market segmentation, they target specific user groups and put forward corresponding value propositions to build business models with their own characteristics.There are four main types 4 categories of relatively successful small and medium-sized platforms of platforms: High-end business platforms, Minibus-hailing platforms, Intercity carpool platforms, and Aggregation mode platforms.As can be seen in Table 1, they have many different characteristics in the value proposition, custom segment, and key partner compared to Didi.
As to the value proposition, Didi aims to provide convenient service with lower prices than a normal taxi, High-end business platforms provide high-value luxury vehicle services, the minibus-hailing platforms provide comfortable and convenient commuting service, and the Intercity carpool platform provides flexible and convenient inter-city travel services.
As to the customer segments, Didi's major market area is inside the first and second-tier cities.It focuses on the middle-and low-end consumer groups.This kind of user group is sensitive to price, and the low-price strategy can help the platform realize the scale effect quickly.High-end business platforms target the high-end business crowd in first-and second-tier cities.This kind of user group has high consumption power and pays more attention to high-quality services.This low-price-sensitive character makes the user scale of HBP more stable than Didi.The major market area of Minibus-hailing platforms is the suburbs of the first-tier big city.Its user group is the commuters who live in the suburbs but work in the city center.Compared with Didi, it has a stronger and more stable transportation capacity for commuters in the morning and evening rush hours.The target market areas of the intercity carpool platform are those third-, fourth-, and fifth-tier cities.They focus on the users with intercity travel needs.
As to the key partners, Didi and Intercity carpool platforms mainly rely on social private-owned vehicles, while high-end business platforms and minibus-hailing platforms are mainly supported by auto manufacturers.Due to the different sources of vehicles and drivers on the supply side of each type of platform, the corresponding cost structure and revenue streams are also different.
It can be seen that all these small and medium-sized platforms rely on their own characteristics and advantages to develop in a specific field.They seek a new potential development direction by constructing a differentiated value proposition and operating model from Didi, thus gaining development space in an almost monopoly market.
In addition, a distinctive feature of the ride-hailing services market is the serious homogeneity of products.This feature determines that competitors in the industry will often achieve differentiated competition through price subsidy wars.Therefore, the business model of the ride-hailing platform is highly dependent on the network aggregation effect driven by subsidies, and the stickiness of drivers and passengers is low so that subsequent new entrants have opportunities as long as they make capital investments and burn money subsidies.This means that market expansion under competitive environment conditions will not only increase the operating costs of enterprises but also increase the difficulty for enterprises to achieve profitability.
In any industry, the relationship between participants is not a zero-sum game, but a relationship where competition and cooperation coexist.The online car-hailing market is no exception.When the platform develops to a certain level, the closed platform model will show problems.For a single online car-hailing company, no matter the size of the platform, its own services or business scope has boundaries.When it develops to a certain extent, the marginal cost of expansion will inevitably exceed the marginal benefit.Therefore, the development of online car-hailing platforms from closed self-operated platforms to aggregated platforms is a general trend.After

General Conclusions
This research aims to explore how SMEs in the sharing economy can gain a competitive advantage in a market monopolized by giants.Taking China's ride-hailing industry as an example, it investigated the 207 small and medium-sized platforms and.identified four successful business models based on the BMC framework and kmeans clustering method.Compared to the industry giant (Didi), these small and medium-sized platforms have significant differences in the value proposition, custom segment, and key partner.Didi's main market area is first-and second-tier cities and its value proposition is to provide low-cost and convenient ride-hailing services to mass consumers.High-end business platforms target the high-end business crowd in first and second-tier cities.They build a business model with the support of well-known auto manufacturers and provide professional, high-value, and luxury ride-hailing services to this small group of consumers.Minibus-hailing platforms rely on policy support and cooperate with new energy vehicle manufacturers to provide comfortable and convenient commuting services to commuters who live in the suburbs and work in the city center.The intercity carpooling platform has obvious regional characteristics, providing flexible and convenient intercity travel services for people with intercity travel needs.In addition, the aggregation mode brings together some small and medium-sized platforms with the same business model as Didi.They open their own applications to other collaborative platforms, allowing users to choose from a variety of services provided by multiple platforms on a single application interface.This alliance has greatly expanded the network effects and business boundaries of small and medium-sized rail-hailing platforms, thereby significantly enhancing their competitiveness in the rail-hailing market.In sum, these types of platforms rely on their own characteristics and advantages to develop in a specific field.They seek new potential development directions by building differentiated value propositions and operating models from industry giants, thereby gaining survival and growth opportunities and space in the market.

Theoretical Implications
This study contributes to the existing literature in several ways.Firstly, existing typological studies on sharing economy business model innovation are mainly based on the broadest or even the full spectrum of business practices of sharing economy (Jiang et al., 2021;Mun˜oz & Cohen, 2017, 2018;Netter et al., 2019;Reuschl et al., 2021;Sun et al., 2022), lacking detailed research on specific industries and specific categories of business cases.Focusing on China's ride-hailing industry, this paper conducted a typological study on the business practice of the sharing economy platform, which enriches the existing research on sharing economy business model.In addition, existing studies on the sharing economy business model mainly focus on giants or representative platforms (Garud et al., 2022;Mun˜oz & Cohen, 2018;M. Ritter & Schanz, 2019;Rojanakit et al., 2022;Teece, 2018), which occupy the majority of the market share and are in a dominant position in the market competition.Little is known about the huge number of small and medium-sized sharing platforms in the market so far, although they are also a very important part of the sharing economy.

Managerial Implications
This research presents a picture that reflects the development status and business practices of small and mediumsized platforms in China's ride-hailing market.It can provide a valuable reference for the business model innovation activities and sustainable growth of small and medium-sized sharing economy platforms in other regions and industries.Managers seeking to develop and maintain small and medium-sized sharing economy platforms could benefit from this study.The proposed classification not only can help managers better position their platforms in the competitive landscape but also evaluate whether their platforms have a substantial competitive advantage in the industry market.At present, industry giants have absolute dominance in almost all areas of the sharing economy.Homogeneous competition in the mainstream market area is unrealistic for small and medium-sized platforms.They should continuously accumulate strength in the tough competition and rely on their own characteristics and advantages to deeply cultivate in specific areas.To be specific, refined market identification and segmentation should be further implemented to continuously develop differentiated competitive advantages.In summary, the research framework and results of this article can help small and medium-sized sharing economy platform enterprises better identify potential market opportunities and build competitive business models and operational strategies to gain a foothold in the industry market dominated by giants.

Limitations and Future Research Directions
This study also has some limitations that need to be further supplemented and improved in the future.Firstly, due to the lack of data availability, the detailed differences in management, financial, and marketing performance among different types of platforms still require further in-depth analysis in the future.In addition, this study only focuses on the special case of ride-hailing to investigate how small and medium-sized platforms in the sharing economy can gain a competitive advantage in a market monopolized by giants.However, there are various types of sharing economy business practices in the real world, such as goods sharing (e.g., tools, clothes, equipment), mobility services (e.g., car sharing and bikesharing), space sharing (e.g., workspace, home), etc.A systematic and comprehensive comparative study on the business practices of small and medium-sized sharing economy platforms in different industries will provide more theoretical and practical value.Moreover, the development of sharing platform business practices is a dynamic process, and corresponding classifications of business models may also change over time.Therefore, it is necessary to conduct typological research within a longer time frame to better observe and analyze the temporal variability of business models of small and medium-sized platform enterprises.

1
University of Science and Technology Beijing, China 2 Hebei University of Economics and Business, Shijiazhuang City, China 3 Beijing Union University, China et al. (

Figure 1 .
Figure 1.The framework of business model canvas.

Figure 3 .
Figure 3. Price sensitivity analysis of the users in China ridehailing market.
Results of elbow method.Note.The cost function value is the residual sum of squares within clusters.Appendix B. General Statistics of Cluster Analysis.

Table 1 .
The Distribution of Samples in Different Dimensions.

Table 2 .
The Results of Cluster Analysis.

Table 3 .
Significantly Different Parts of the BMC Between Each Type of Platform.
It is not yet clear how these small and medium-sized enterprises in the sharing economy can survive in the market monopolized by the giants and achieve sustainable growth.More specifically, what are their main different characteristics compared to those industry giants in terms of business model?This study explores how small and medium-sized platforms in the sharing economy can gain a competitive advantage in a market monopolized by giants, which makes up for the shortcomings of the existing research to some extent.