Full lengh articleAnalyzing the trend of O2O commerce by bilingual text mining on social media
Introduction
Online to offline (O2O) commerce is a business combination model that includes both online and offline channels; the advantage of the online channel is that it can drive offline sales and coupon redemption (Phang, Tan, Sutanto, Magagna, & Lu, 2014). O2O commerce completely changes price competition in the traditional channel supply chain, and it can be seen as a specific form of multichannel integration, where the focus is on conducting online promotions (Zhang, Chen, & Wu, 2015). With the popularization of information technology, O2O commerce has become localized and integrates internet or mobile technology for linking online shopping and front-line transactions to entice consumers to make purchases in physical establishments (Kang, Gao, Wang, & Wang, 2015). Compared to traditional ecommerce approaches, such as B2C and C2C, the approach of O2O focuses on linking online transactions to local services, such as restaurants, cinemas, gyms, tourism, beauty salons, spa shops and other service-oriented products. Although the abovementioned approaches provide online purchase and payment services, traditional ecommerce delivers goods or services directly to customers, while O2O provides services at brick and mortar businesses with offline logistics (Yan, 2013). Due to the growth of the O2O market, enterprises and retailers need to focus on building a seamless experience between online and offline interactions to acquire and retain customers. Participants in the O2O model include consumers, retailers, third-party service providers, and O2O operators. Meanwhile, online matchmaking, online payments, offline consumption, and message back indispensable are included in the transaction process of O2O. The four aspects for price differences in products or services, revenues from online advertising, business purchases for consumers, and membership fees are considered as the profits of using an O2O business model (Xue & Li, 2016). As a type of business combination, the O2O model focuses on local service, the mobile and internet commerce technology to build a long-term competitive advantage, in terms of both individual customers and business customers. The O2O business model is gradually dominating the online shopping market. According to Japan's Nomura Research Institute survey in 2011, the size of the Japanese O2O market was approximately 24 trillion yen (US $219 billion) and was estimated to reach approximately 51 trillion yen (US $465 billion) in 2017. The size of the Chinese O2O market was approximately RMB 665.94 billion (US $96.56 billion) in 2016 and was estimated to reach approximately RMB 834.32 billion (US $120.96 billion) in 2017.
The popularity of O2O has also resulted in extensive discussions in the academic field. However, seldom do studies explore the O2O trends analysis based on the perspective of social media, especially those that address different languages. In this study, we focus on analyzing the trends of O2O by applying a text mining approach on the Twitter messages posted in English and Chinese. Trend analysis is the widespread practice of collecting information and attempting to spot a pattern, and it could be used to estimate uncertain patterns (John, 2004). It is not only used to predict future outcomes, but also used to observe the past events, to track the variances and frequencies in the past historical data. Trend analysis by using text mining approach has been studied in different disciplines. For example, social media analytics is the practice of gathering data from social media websites and analyzing that data using social media analytics tools to make business decisions. It is considered as information tool to collect, monitor, analyze, summarize, and visualize social media data to extract useful findings (Zeng, Chen, & Li, 2010). Comparing to traditional data analytics, social media tools can improve interactivity between customers and the companies, while the traditional media can only do in one way connection (Bertot, Jaeger, & Glaisyer, 2010). Regarding the O2O business model, social media helps to build close relationships between online interactions and offline marketing. The social media, such as Instagram, LinkedIn, Twitter, Facebook etc. has experienced in outstanding expansion, and became important channels for business and marketing (Herrero, San Martín, & Garcia-De los Salmones, 2017). Social media has become a crucial role, not only in changing people's communication with their friends but also by changing the way providers communicate with customers in O2O commerce (Asur & Huberman, 2010). Through social media's function of referrals and guiding users to real-world stores, retailers can create social events and promote social word-of-mouth to determine who the valuable customers are and promote products or services to social customers (Tsai, Yang, & Wang, 2013). This study focuses on the analysis of Twitter data, because Twitter is one of the most popular microblogging service providers with popular features such as hashtag (Fahd, El Habib, & Omar El, 2017) and tweets from Twitter are a rich source of data that represents trends analysis. The Twitter data was used to make predications and examine the reflective of the climate of public trends during the 2011 Singapore General Election campaign (Skoric, Poor, Achananuparp, Lim, & Jiang, 2012). We applied bilingual text mining approach by means of supervising Twitter data and identified related keywords with building the corresponding concept linking diagram for the first and second layers. Afterwards, terms related to O2O were investigated, and the implication of relevant words were explored. We also compared the differences and similarities of O2O keywords mentioned in English and Chinese tweets to understand the recent development of O2O commerce in different language regions. The study provides promising implications for enterprises aiming to adopt O2O commerce business models by suggesting that they can apply similar text mining approaches for developing similar models based on their available resources. These enterprises can use text mining to obtain news about the industry, information regarding their competitors and comments from the public about products and services. In this way, enterprises can evaluate the market and predict future market trends (Dai, Kakkonen, & Sutinen, 2011). By observing the issues and opinions of people who post in the social media, we can further understand the industry trends and develop immediate responses for a market strategy.
Section snippets
Literature review
There is a rich body of literature investigating the O2O business model and its approach from a different perspective. The following literature review discusses how past studies investigated the O2O commerce from four different perspectives: business models, customer management, channel management and marketing issues. This discussion can help us to understand the necessity of using social media mining approach to detect the pattern of O2O development.
Traditionally, studies discuss the O2O
Introduction of text mining
The real-time and important characteristics of social media, such as Twitter, can provide a great opportunity for ecommerce companies to learn more about their customers and potential ones as well (Shen & Kuo, 2015). In this research, text mining approach was adopted to understand the trend of O2O development. Text mining is an extension of data mining techniques, and text mining relies on feature-based representations of documents that can include words, concepts, characters and terms (
Mining results from English tweets
After applying text mining to the English tweets, the most strongly associated terms with the keyword “O2O” in the first layer of concept linking include “China”, “Alibaba”, “Baidu”, “DianPing”, “Meituan”, “Asia”, “ecommerce” and “service”. These 8 terms can be classified as four types of keywords with similar meanings: countries or areas, companies, O2O platforms and service models. There are 43 terms in the second layer, which are considered as the most strongly associated terms with the
Discussion and conclusions
In the big data era, social media data like Twitter can provide a great resource for the humanity to build wisdom based on collective intelligence mechanisms and advanced analysis capabilities on value contexts (Miltiadis, Vijay, & Ernesto, 2017). Trends analysis based on social media allows a business to get possible business advantages by analyzing the publicly available social media data of a business and its competitors (Fan & Gordon, 2014). The social media data from a business and their
Acknowledgements
This research was supported by the Ministry of Science and Technology, Taiwan, under contract number MOST 106-2410-H-008-031.
References (36)
- et al.
Turning like to buy social media emerges as a commerce channel
(2011) - et al.
Predicting the future with social media. Paper presented at the web intelligence and intelligent agent technology (WI-IAT)
- et al.
Data mining in social media
Social Network Data Analytics
(2011) - et al.
Social media technology and government transparency
IEEE Computer Society
(2010) - et al.
Improved Precision Recommendation Scheme by BPNN Alogrithm in O2O commerce. Paper presented at the e-Business Engineering (ICEBE)
- et al.
A study on consumers' discontinuance intentions towards O2O commerce: Focused on the mediating effects of perceived risk and user resistance
(2015) - et al.
MinEDec: A decision-support model that combines text-mining technologies with two competitive intelligence analysis methods
International Journal of Computer Information Systems and Industrial Management Applications
(2011) - et al.
Study on the Development of O2O E-commerce Platform of China from the perspective of offline service quality
International Journal of Business and Social Science
(2014) - et al.
Harnessing semantic features for large-scale content-based hashtag recommendations on microblogging platforms
International Journal on Semantic Web and Information Systems
(2017) - et al.
The Power of Social Media
Communications of the ACM
(2014)
Selection criteria for text mining approaches
Computers in Human Behavior
Integrating bricks with clicks: Retailer-level and channel-Level outcomes of online–offline channel integration
Journal of Retailing
Explaining the adoption of social networks sites for sharing user-generated content: A revision of the UTAUT2
Computers in Human Behavior
Social media competitive analysis and text mining: A case study in the pizza industry
International Journal of Information Management
Quanzhou agricultural e-commerce platform based on O2O modes
More Precise: Stores Recommendation under O2O Commerce
International Journal of Computing and Digital Systems
Introduction to text miner
Public Agenda. Public Attitudes on Higher Education – A Trend Analysis, 1993 to 2003
Cited by (157)
The effects of non-fungible token platform affordances on customer loyalty: A Buyer–Creator duality perspective
2024, Computers in Human BehaviorAssessment of the cultural-natural importance of the scenic spots from a network perspective
2023, Landscape and Urban PlanningExploring the nexus between food and veg*n lifestyle via text mining-based online community analytics
2023, Food Quality and PreferencePreparing an exhibition in the post-pandemic era: Evidence from an O2O-based exhibition of B2B firms
2022, Technological Forecasting and Social ChangeThe Intelligent Advertising Image Generation Using Generative Adversarial Networks and Vision Transformer: A Novel Approach in Digital Marketing
2024, Journal of Organizational and End User Computing