플랫폼 서비스 유형별 전환 결정요인 실증분석 (An Empirical Analysis of Determinants Influencing Switching Decisions Across Platform Service Types)

플랫폼 경쟁은 이제 단순히 기업 간의 경쟁을 넘어 글로벌 정책 이슈로 부상하고 있다. 주요국 에서는 플랫폼의 시장 독점 및 소비자 보호 문제를 해결하기 위해 플랫폼에 특유한 새로운 규제를 도입하거나 기존 경쟁법을 개선하는 등 적극적인 조치를 취하고 있다. 이러한 노력은 플랫폼 간 자유로운 이동을 촉진하고, 소비자가 플랫폼 선택에서 실질적인 자유를 누릴 수 있도록 함으로써 시장에서의 공정한 경쟁과 경합성을 보장하기 위한 것이다. 그러나 실제로 소비자 들이 플랫폼 간 전환을 쉽게 할 수 없다면, 플랫폼 시장에서의 유효한 경쟁이 이루어지기 어렵다. 따라서 플랫폼 경쟁의 실효성을 확보하기 위해서는 소비자의 전환 비용을 분석하고, 각 플랫폼 서비스 유형에서 소비자 고착 요인을 규명할 필요가 있다.
이러한 관점에서 본 연구는 소셜네트워크 서비스, 검색 서비스, 음식 주문 및 배달 서비스, 숙박 예 약 서비스, 전자상거래 서비스 등 다섯 가지 유형의 플랫폼 서비스를 대상으로 소비자들이 해당 플랫폼에 고착되는 요인을 실증 분석하였다. 정보통신정책연구원의 부가통신실태조사의 이용자 설문 조사 결과를 이항로짓모형을 활용하여 분석한 결과, 멀티호밍 서비스 개수, 기존 데이터 축적 여부 는 플랫폼 서비스 전환에 미치는 영향이 유형별로 크게 다르지 않았으나, 이용자 및 이용사업자 규모왿� 같은 다른 요인들은 플랫폼 서비스 유형에 따라 상이한 영향을 미치는 것으로 나타났다.
본 연구 결과는 플랫폼 경쟁의 양상을 최소한 이용자 관점에서 플랫폼 유형별로 파악할 필요가 있음을 시사한다. 예를 들어, 멀티호밍 행태가 활발히 나타나는 플랫폼 서비스에서는 단순히 이용자 규모가 크다는 이유만으로 소비자 고착 현상이 발생하지 않을 수 있으므로, 이용자 규모가 크다는 점만으로 플랫폼 간 경쟁이 저해된다고 보기 어렵다. 반면, 이용사업자 측에서 발생하는 네트워크 효과는 소비자 고착에 유의미한 영향을 미친다. 따라서 이용사업자가 여러 플랫폼 서비스에 입점할 수 있도록 지원하는 정책 대안은 시장 경쟁 활성화에 기여할 수 있다. 또한, 플랫폼 유형에 관계없이 이용자의 데이터 이동성을 촉진하는 것도 중요한 과제이다. 다만 데이터 이동성 정책의 실효성은 이용자의 적극적인 의지왿� 행동이 뒷받침되어야 하므로, 데이터 이동성을 활성화하기 위한 세심한 메커니즘 설계가 요구된다.

Platform competition has increasingly emerged as a global policy issue that goes beyond just competition between companies. Major countries are actively enacting new regulations or revising competition laws to address market dominance and consumer protection concerns within platform markets. Such efforts seek to facilitate consumer mobility across platforms, thereby ensuring genuine freedom of choice and fostering fair competition. Analyzing consumer switching costs and identifying the factors that contribute to consumer lock-in acorss various platform service types are essential to supporting effective competition within platform ecosystems, as high switching barriers are likely to undermine competitive dynamics.
In view of this, This study empirically investigates factors contributing to consumer lock-in across five major platform service types: social network services, search services, food ordering and delivery services, accommodation booking services, and e-commerce services. Data on switching behavior of users of major platform services was obtained from the Korea Information Society Development Institute’s 2023 Survey on value-added telecommunications service providers. Considering the various switching costs that may be incurred in platform services, this study set up a model as follows.
First, the dependent variable captures whether users have switched their primary service provider within the past year. Independent variables included multi-homing (number of services currently used), network effect, inertia, service quality, and data (usage history). The network effect can be viewed in terms of both the number of users using the same service and the number of platform business users that provide services on the platform. As we have seen, network effects are one of the important variables that affect consumer lock-in in platform usage, so we included the size of users and the size of service providers as independent variables in this study. The size of users is the same for each type of service, and the size of platform business users is included for food ordering and delivery services, accommodation booking services, and e-commerce services. Considering the findings of previous studies (Engels, 2016; Gao et al., 2014) that inertia in familiar platform services is a factor that hinders service switching, we included inertia as a variable of whether users value familiarity with the service. The quality of service is also included as an independent variable because it is one of the service stickiness factors, and the more data accumulated on the existing platform, the higher the cost of switching to a new platform (Nicholas, 2020), so we set whether users value accumulated usage history as a reason for not switching as an independent variable.
Additional factors specific to each platform type were incorporated. For search services, we considered the amount of information provided in search results and the accuracy of the search as separate variables. For food ordering and delivery services, accommodation reservation services, and e-commerce services, the diversity of available platform business users was included, with further consideration given to factors such as subscription memberships and delivery speed for e-commerce services.
The results of this study indicate that, across all types of digital platforms, a higher degree of multi-homing—defined as the use of multiple services within the same platform category—positively influences users' likelihood of switching platforms. This finding is consistent with previous studies, which suggest that digital platforms that allow multi-homing offer greater switching flexibility compared to traditional services.
For users who place significant value on the accumulation of usage history within existing services, switching behavior across all types of platform services was negatively impacted. This suggests that data portability could facilitate user switching between platform services, thereby fostering competition in the market.
Network effects from the user base are largely insignificant across platform types, except for accommodation reservation services, suggesting that a large user base may not necessarily equate to market dominance in a multi-homing environment (Tucker, 2018), and that consumers may even prefer smaller networks (Cennamo, 2020).
Conversely, the perceived importance of platform business user network size had a negative effect on switching behavior, except for e-commerce services. For food ordering and delivery services, users who placed high importance on the availability of a large number of listed restaurants also demonstrated reduced switching behavior. This implies that, in most platform service categories, the network effects associated with the provider side—i.e., the number of providers available on the platform—can serve as a factor contributing to user lock-in.
However, as previously discussed, in the case of horizontally differentiated e-commerce services, neither the user base size nor the provider network size—representing network effects—significantly influenced switching behavior. This suggests that multiple services can coexist by targeting distinct user segments.
Lastly, in terms of inertia—defined as user reluctance to switch due to familiarity with existing services—an unexpected positive effect on switching behavior was observed for search services. This may be attributed to the fact that the usage of search functions does not differ substantially across different services, meaning that switching to a new service does not result in increased learning costs or heightened uncertainty, both of which are typically factors that elevate switching costs.
The results of this study suggest that the factors affecting switching differ by platform type, suggesting that it is necessary to understand the competitive landscape, at least from a user perspective, by platform type. In the case of platform services where multihoming behavior is actively practiced, it is unlikely that competition is hindered simply because of the large user base, as the large user base alone may not cause consumer lock-in. On the other hand, network effects arising from service providers have a significant impact on consumer lock-in, so policy alternatives that promote multi-homing from the perspective of service providers so that service providers can enter multiple platform services may help to promote market competition. It is also necessary to promote data mobility of users regardless of the type of service, but the effectiveness of data mobility policies is predicated on the active will and behavior of users, so mechanisms should be carefully designed to ensure that data mobility is actively used.
This study offers empirical insights suggesting that reducing switching barriers by platform type is crucial to fostering effective competition. However, a limitation of this study is that the data used in this study is based on users' stated preferences rather than actual switching behavior in the market, and the factors that affect switching are based on users' opinions rather than the actual size of service users. Nevertheless, this study is meaningful in that it attempted to conduct empirical analysis in a field where it is difficult to utilize actual market data, and considering that platform services are highly influenced by technological evolution and government policies, it is necessary to continue to conduct research related to platform services and user switching in the future.