I. Introduction
Benefiting from the increasing improvement of network infrastructure such as broadband and fiber optics, as well as the continuous iteration of internet technology, the penetration rate and recognition of online media have rapidly increased. In contrast, paper media is facing the dilemma of a sharp decline in advertising and subscription volume, as well as a decline in influence [1] . To cope with this trend and maintain sustained revenue growth, major media organizations are also actively seeking strategies to address it, such as replacing advertising departments with project teams to coordinate business cooperation, encouraging the provision of information and public relations services for enterprises, and carrying out integrated media reforms. At the end of 2017, Caixin Media Group, officially announced a new way of operating after 8 years of establishment, with news business as the core, and began the commercial practice of comprehensive paid reading.
In fact, Paid media is not new. In the late 20th century, mainstream American financial newspapers such as The New York Times and The Wall Street Journal had already tried this model and accumulated rich experience [2] . In China, Caixin's approach is pioneering. From three years of practice, Caixin's paid reading business has achieved a certain user accumulation. In the first half of 2020, the number of Caixin paid reading subscription users reached 510'000. In addition, the production and distribution of Caixin's paid reading business, as well as the expansion of users, have formed a relatively mature process, and market-oriented operation has shown initial results. In the past two years, several media outlets, including the People's Daily, have imitated the Caixin model and attempted to adopt fee-based measures for important content. The popularity of the paid reading model in the media industry is gradually increasing. In addition, Hu Shuli, the founder of Caixin Media, has always been the initiator and practitioner of advanced media models in the media industry. Whether the paid reading model launched this time will trigger a demonstration effect and promote its popularization throughout China is also widely anticipated by industry insiders.
Of course, the paid reading business also faces challenges from various aspects [3] . On the one hand, due to the high overlap of news organizations' same topic reports, to stand out in the competition with free content, paid media need to have better news quality and innovative ways to attract audiences, and corresponding distribution channels also need to be as comprehensive as possible [4] . In addition, the paid wall has also had a negative impact on the enthusiasm of advertisers for a period. So the widespread popularization of paid reading services still requires layers of hard work.
At present, existing research still focuses on the operational status of media that have already conducted paid media abroad, and there is a lack of in-depth and systematic analysis of the practice of paid media in Chinese media. The launch of Caixin's paid media business is a milestone event in the development history of China's media industry. Given the industry influence and future development space of Caixin's paid media business, further attention, and systematic research on it are undoubtedly of certain significance.
This article integrates marketing into the study of media economics and observing the media industry from a marketing perspective helps to gain a professional understanding of the market-oriented operation of media. This article defines paid reading business as a content product, using the Expectation Confirmation Model as the framework, and selects 4P marketing theory as the analysis tool to study the problems and corresponding improvement measures of Caixin's paid media business in production, pricing, channels, and promotions. The aim is to optimize the strategy of media engaged in paid reading and provide a reference direction for its future development.
II. Theoretical Background
1. Expectation Confirmation Theory
In 1980, Oliver proposed the Expectation Confirmation Theory (ECT) [5] , which is the fundamental theory for studying consumer satisfaction in the field of consumer behavior.
The theory of expectation confirmation holds that the comparison between consumers' pre purchase expectations and their post purchase experiences (post purchase performance) determines their desire and behavior for a second purchase. In the theory of expectation confirmation, there is a negative correlation between expectation and confirmation. When consumers' expectations are higher and their actual sense of use is not met, their confirmation is lower, and their satisfaction with the product also decreases. On the contrary, when consumers' expectations are low and their actual usage experience is highly satisfied, the degree of confirmation is high, and user satisfaction will also increase.
In the theory of Expectation Confirmation Model theory, there are four keywords, namely "expectation", "performance", "confirmation", and "satisfaction". Expectations directly affect the level of confirmation of user expectations and user satisfaction. Expectation is a prediction and assumption made by consumers in the early stages of purchasing a product or service, with a strong subjective color. The so-called expectation refers to the prediction of the product or service effectiveness by consumers or based on past shopping experience, paraphrases from surrounding relatives and friends, or information provided by marketing personnel. Fornelletal (1996) Research suggests that consumers' expectations for products or services during early purchases mainly come from three aspects: including consumers' search for external information, word-of-mouth communication within their circle of friends, and sales guidance from salespeople during the first purchase.But when making a repeat purchase, the expectation at this time is not only external information from the early stage, but also direct user experience from within the individual consumer. Therefore, the more complete the consumer's information, the higher their expectations for the product or service. [6]
Performance, as a factor compared to expectations, is used to confirm consumer satisfaction. The "expectation failure theory" proposed by Oliver and DeSarbo (1998) proves that the results formed by comparing consumers' pre purchase "expectations" with actual performance form failure. When actual performance exceeds expectations, there will be a positive lack of validation, leading to an increase in user satisfaction. When performance falls short of expectations, negative failures occur, negatively affecting user satisfaction. Confirmation is a result of consumers' actual experience, resulting from the gap between expected and actual performance.
Confirmation can be divided into three forms: objective confirmation, inference confirmation, and perceptual confirmation. Firstly, objective confirmation refers to the actual gap between consumers' objective usage feelings and expectations, which is usually considered to exist; Inferential confirmation is a reasonable inference made by researchers by observing consumers' reactions before and after purchase; Perceived confirmation is a subjective evaluation of performance and benchmark differences by consumers, including psychological factors, whose results often measure performance and acceptance of expectations.
2. Design of Acceptance Model
This study refers to ECT theory and and designs a user expectation confirmation model for paid reading (hereinafter referred to as the "acceptance model"), as shown in Figure 1. The acceptance model follows the four independent variables in the ECT model, namely social impact, performance expectations, perceived usefulness, and expected confirmation. Based on the actual development of financial paid media and relevant research results, the intermediary variable is designated as satisfaction.
Fig. 1. Research Model
III. Research Design
1. Research Hypothesis
Based on the current development status of China's Caixin paid reading media and the main factors affecting the acceptance of platform paid users, this study proposes eight hypotheses.
1.1 Social Influence
Social influence is the basis for the survival and development of the media, and is the ultimate goal of realizing news value [7] . Scholars suggest that social influence should be taken as the key index to evaluate the success of the media. If the content that the media provided to users is more accurate, timely and influencal, the readers will have a stronger user perception of the media, so as to improve user adoption behavior . Therefore, this paper puts forward the following hypothesis:
H1: Social impact has a positive impact on the adoption behavior of Caixin App readers.
1.2 Performance expectations
Performance expectations refers to the extent to which readers believe that using information systems will help them improve life or job performance, including perceived usefulness, extrinsic motivation, job suitability, relative advantages, and outcome expectations [8] .Individuals use information technology and products because they believe it will improve their work performance. The scholar found when readers feel that using paid media platforms can improve their study or work efficiency, improve their quality of life, and enjoy more resources and services provided by the platform, allowing them to get twice the result with half the effort when creating, then readers' willingness to use paid content platforms will increase. Therefore, this paper puts forward the following hypothesis:
H2: Performance expectations have a positive impact on the adoption behavior of Caixin App readers.
1.3 Adoption behavior
"Adoption" means to accept and adopt, especially after selection or recognition . Paid media adoption behavior refers to the behavior of readers accepting and adopting paid media through their choice or recognition of paid media. The adoption behavior has an impact on the reader's willingness to continue using it. Therefore, this paper puts forward the following hypothesis.
H3: Adoption behavior has a positive impact on willingness to continue using.
1.4 Perceived Usefulness
Perceived usefulness refers to the extent to which people believe that using a specific system or Application can improve their job performance [9] . The significant impact of perceived usefulness on user experience has been widely confirmed. When the reader perceives that the paid
media can enhancing cognition and acquiring knowledge , the user will be satisfied with the paid media and enhance the willingness to continue to use it. Therefore, this paper puts forward the following hypotheses:
H4: Perceived usefulness has a positive impact on reader satisfaction.
H5: Perceived usefulness has a positive impact on readers' willingness to continue using Caixin App.
1.5 Satisfaction
Satisfaction refers to the emotion generated by people after using a specific system or Application, which is expressed as a positive or negative feeling [10] .In the research context of paid media, if the user is satisfied after using the paid media, the users may continue to use the when they want to read or get some news , so as to enhance the user's willingness to continue to use it. Therefore, this paper puts forward the following hypotheses..
H6: Satisfaction has a positive impact on the willingness of Caixin App readers to continue using it.
1.6 Expected Confirmation
Readers’ expected confirmation is composed of five dimensions: information, social, emotional, entertainment and social identity expectation, and all five kinds of confirmation can positively and significantly affect the user's satisfaction and perceived usefulness. Therefore, this paper puts forward the following hypotheses.
H7: Expected confirmation has a positive impact on the perceived usefulness of Caixin App readers.
H8: Expected confirmation has a positive impact on reader satisfaction of Caixin App.
2. Research Subjects
To ensure the scientific and effective nature of the research, this study selected Caixin, a local paid reading media in China, as the research object. Throughout the development of the media industry, the significance of Caixin's paid reading business cannot be underestimated, but due to its relatively short start-up time. Zhao Xiaoxin believes that the news value and influence of Caixin are driven by the founder of Caixin Media, Hu Shuli, and a group of celebrity editors. The improvement of influence can help the development of paid reading business.
For users who already choosed Caixin and the potential users: 200 readers were selected for pretesting in the first round. In the second round, 178 readers were selected as the formal research subjects, and six readers who performed positively in the survey were randomly selected for semi-structured interviews.
3. Questionnaire Design
To gain a deeper understanding of the factors that affect the acceptance of paid reading and provide a better explanation of the acceptance model, this study followed the voluntary principle and randomly selected 6 readers who performed positively in the survey for interviews.
Before the interview, the researcher first introduced the concept of paid reading to the respondents to ensure the accuracy of their answers. During the interview, the researchers mainly discussed with the respondents about the popularity of paid reading and its application status, readers' feelings about using paid reading, and pursued relevant questions to explore the reasons behind the phenomenon.
After the interview, the researcher sorted and analyzed the interview records and found that: Readers have a high enthusiasm for paid reading and are more willing to try paid reading. Readers hope that paid media can further improve the quality of paid reading news and promote the popularization of paid reading. In the development process of paid reading, efforts should be made to simplify operations and pay attention to the convenience of use.
4. Data Analysis
4.1 Sample and Usage Statistics
In this study the questionnaire survey in this study was conducted online and was conducted among passengers in Beijing. the online questionnaire was distributed by snowballing through the Questionnaire-Star platform. Through the investigation, 378 valid samples were ultimately collected. The basic information and usage of the samples are shown in Tables 1 and 2:
Table 1. Sample Statistics
Table 2. Use Caixin App under what circumstances
4.2 Research Variable Description statistics
From Table 3, the minimum, maximum, mean, and standard deviation of each study variable. The mean expected confirmation is 3.707, with a standard deviation of 0.933; The mean of perceived usefulness=3.858, standard deviation=0.841; The mean of social influence=3.530, standard deviation=0.986; The mean of performance expectations=3.651, standard deviation=0.999; The mean of adoption behavior is 3.523, with a standard deviation of 0.972;
Table 3. Statistical Description of Research Variables
The mean of satisfaction is 3.881, with a standard deviation of 0.863; The mean of willingness to continue using is 3.968, with a standard deviation of 0.826.
4.3 Reliability Analysis
Reliability refers to the consistency or stability of measurement results obtained from measurement tools. We usually use α the coefficient (i.e., Cronbach's Alpha) is used to measure the reliability of the questionnaire, α The larger the coefficient, the higher the reliability of the questionnaire, that is, the higher the credibility and stability of the questionnaire. Devellis (1991) proposed the following viewpoint: α If the coefficient is between 0.60 and 0.65, it is best not to; α The coefficient value between 0.65 and 0.70 is the minimum acceptable value; α The coefficient value is quite good between 0.70 and 0.80; α The coefficient value is very good between 0.80 and 0.90.
From Table 4, the reliability of expected confirmation is 0.832, the reliability of perceived usefulness is 0.839, the reliability of social influence is 0.836, the reliability of performance expectations is 0.826, the reliability of adoption behavior is 0.845, and the reliability of satisfaction is 0.869.
Table 4. Reliability Analysis
The reliability of willingness to continue using is 0.848. The reliability of each research variable is above 0.7, indicating good reliability of the scale.
4.4 Confirmatory Factor Analysis
This article will use AMOS24.0 software to conduct confirmatory factor analysis on the samples. Confirmatory factor analysis is a statistical analysis of social survey data that explores whether the factor structure model of the scale fits the actual collected data, and whether the indicator variable can effectively serve as a measurement variable for potential variables. This study used the maximum likelihood method to estimate the model and verified its fit through the following indicators: (1) chi square x2) Inspection. x the 2 index is the most basic test indicator for model fitting, and is generally used x The smaller the 2/df value, the higher the simulation fit. usually x When 2/df0.9, TLI>0.9, and IFI>0.9 indicate good model fit (Hu&Bentler, 1999).
From Table 5, x the value of 2/df is 1.614, the value of RMSEA is 0.040, the value of SRMR is 0.035, the value of CFI is 0.971, the value of TLI is 0.965, and the value of IFI is 0.972. The model fitting indicators have all reached ideal values, indicating that the confirmatory factor analysis model fits well. The factor loadings of expected confirmation, perceived usefulness, social influence, performance expectations, adoption behavior, satisfaction, and willingness to continue using are all above 0.5, CR values are all above 0.7, and AVE values are all above 0.5. According to Hair's suggestion in validity evaluation, the absolute value of factor load estimation should be at least above 0.5, the mean variance extraction (AVE) indicator value should be above 0.5, and the construct reliability indicator value should be above 0.7. The factor loadings for expected confirmation range from 0.867 to 0.693, with all factor loadings above 0.5, CR values above 0.7, AVE values above 0.5, and p-values below 0.001, indicating significant factor loadings. Therefore, all questionnaires have good convergent validity. This article uses the method proposed by Fornell Rocker to determine whether there is differential validity by determining whether the square root of AVE is higher than the correlation coefficient between two variables.
Table 5. Confirmatory Factor Analysis Parameter Estimation
From the correlation coefficient table 6, the AVE square roots of expected confirmation, perceived usefulness, social influence, performance expectation, adoption behavior, satisfaction, and willingness to continue using are 0.797, 0.762, 0.794, 0.787, 0.806, 0.829, and 0.809, respectively, which are greater than their corresponding correlation numbers, indicating that the questionnaire has good discriminative validity.
4.5 Correlation Analysis
The Pearson correlation coefficient is used to measure the linear relationship between two fixed distance variables. The value of the correlation coefficient is between -1 and 1, and the larger its absolute value, the stronger the correlation between the two. The closer the correlation coefficient is to 1 or -1, the stronger the correlation, and vice versa, the weaker the correlation. In addition, judging the correlation requires a comprehensive consideration of the correlation coefficient and significance level. Only when the correlation coefficient is greater than 0 and the significance level is p<0.05 can it be inferred that variables are correlated. Therefore, this article uses Pearson correlation coefficient to verify whether there is a correlation between variables.
From Table 6, the correlation coefficients between various research variables. There is a significant positive correlation between social influence and adoption behavior (r=0.430, p<0.01).
Table 6. Correlation Coefficients
There is a significant positive correlation between performance expectations and adoption behavior (r=0.384, p<0.01). There is a significant positive correlation between adoption behavior and willingness to continue using (r=0.336, p<0.01). There is a significant positive correlation between perceived usefulness and satisfaction (r=0.382, p<0.01).
There is a significant positive correlation between perceived usefulness and willingness to continue using (r=0.400, p<0.01). There is a significant positive correlation between satisfaction and willingness to continue using (r=0.481, p<0.01). There is a significant positive correlation between expected confirmation and perceived usefulness (r=0.294, p<0.01). There is a significant positive correlation between expected confirmation and satisfaction (r=0.339, p<0.01). Therefore, the hypothesis receives preliminary support.
4.6 Structural Equation Model Analysis
This article uses a structural equation model for hypothesis testing. This article uses x2/df, RMSEA, SRMR, NFI, CFI, TLI, IFI validation of structural model fitting.
Fig. 2. Structural Equation Model
From Table 7, it can be seen that, x The value of 2/df is 1.701, the value of RMSEA is 0.043, the value of SRMR is 0.063, the value of CFI is 0.965, the value of TLI is 0.960, and the value of IFI is 0.966. The model fitting indicators have all reached ideal values, indicating that the structural equation model fits well.
Table 7. Path Analysis
Expected confirmation has a significant positive impact on perceived usefulness（ β= 0.384, p<0.001); Social influence has a significant positive impact on adoption behavior（ β= 0.444, p<0.001); Performance expectations have a significant positive impact on adoption behavior（ β= 0.381, p<0.001); Perceived usefulness has a significant positive impact on satisfaction（ β= 0.335, p<0.001); Expected confirmation has a significant positive impact on satisfaction（ β= 0.279, p<0.001); The adoption behavior has a significant positive impact on the willingness to continue using（ β= 0.253, p<0.001); Perceived usefulness has a significant positive impact on willingness to continue using（ β= 0.269, p<0.001); Satisfaction has a significant positive impact on willingness to continue using（ β= 0.378, p<0.001). Therefore, all research hypotheses are supported.
IV. Conclusions
1. Theoretical and Practical Enlightment
The research results show The research results show that paid media's social influence and performance expectations have a positive impact on Caixin App readers’ adoption behavior,perceived usefulness and expection confirmation has a positive impact on Caixin App readers’ satisfaction . Adoption behavior and satisfaction has a positive effect on continue using intention, what’s more,the perceived usefulness also has a positive effect on continue using intention.
Finally, this article analyzes the framework of 4P, which includes products, prices, channels, and promotions, and proposes targeted strategies to address the problems in paid media business. In terms of products, it aims to improve the quality of reports, increase the proportion of exclusive reports, expand the influence of thematic reports, segment user needs, and develop personalized products with the help of big data to effectively protect intellectual property rights; In terms of pricing, multiple payment walls are set up, tiered pricing is implemented, and hierarchical membership management is implemented; In terms of channels, increase e-commerce sales channels, increase terminal coverage and subscription scale for major customers; In terms of promotion, achieve horizontal integration and collaborate with partners to promote sales.
2. Limitation and Prospects
Although this study has important reference value for the development of paid media, it still has deficiencies. Firstly, compared with the huge number of users of paid media, this study collects data on the premise that the data is effective, and the research problems can be solved.
The effective sample size of is relatively small; Secondly, due to the limitation of sample size, this study does not consider the impact of sample demographic characteristics (such as the readers belong to which region ) on users' willingness to continue to use. Further research on the above deficiencies will be carried out in the follow-up to provide more perfect suggestions.
Although the research in this paper has achieved certain results, there is still significant room for improvement.
References
- W. Reinartz, M. Haenlein and J. Henseler, "An empirical comparison of the efficacy of covariance-based and variance-based SEM," International Journal of Research in Marketing, Vol. 26, No. 4, pp. 332--344, 2009. DOI: 10.2139/ssrn.1462666
- H. Mohammadi, "Investigating users' perspectives one learning : An integration of TAM and is success model," Computers in Human Behavior, Vol. 45, pp. 359--374, 2015. DOI: 10.1016/j.chb.2014.07.044
- Zhao and S. Wang, "Research on the influencing factors of WeChat Applet users' continuous use intention," Modern Intelligence, Vol. 39, No. 6, pp. 70-80,90, 2019. DOI: 10.3969/j.issn.1008-0821.2019.06.008
- Q. Dong, X. Zhou and F. Mao, "Research on users' sustainable use intention in online health community: Based on perceived value theory," Modern Intelligence, No. 3, pp. 3-14, 2019.
- H. I. Chyi, "Willingness to pay for online news: An empirical study on the viability of the subscription model," Journal of Media Economics, Vol. 18, No. 2, pp. 131-142, 2005. DOI: 10.1207/s15327736me1802_4
- P. Farhi, "Build That Pay Wall high," American Journalism Review, Vol. 31, No. 4, pp. 55-57, 2009.
- B. Grueskin, A. Seave and L. Graves, "The Story So Far: What We Know about Business of Digital Journalism," Digital Journalism, Vol. 1, No. 1, pp. 173-174, 2011. https://doi.org/10.1080/21670811.2012.740278
- L. Chiou and C. Tucker, "Paywalls and the demand for news," Information Economics and Policy, Vol. 25, No. 2, pp. 61-69, 2013. DOI: 10.1016/j.infoecopol.2013.03.001
- J. Macnamara, "Remodelling media: The urgent search for new media business models," Media International Australia, Vol. 137, No. 1, pp. 20-35, 2010. DOI: 10.1177/1329878X1013700104
- M. Myllylahti, "Newspaper paywalls-the hype and the reality: A study of how paid news content impacts on media corporation revenues," Digital Journalism, Vol. 2, No. 2, pp. 179-194, 2014. DOI: 10.1080/21670811.2013.813214
- I. A. Zolkepli, Y. Kamarulzaman, "Social media adoption: The role of media needs and innovation characteristics," Computers in Human Behavior, Vol. 43, pp. 189-209, Feb. 2015. DOI: 10.1016/j.chb.2014.10.050
- Y. Kim, T. Y. Choi, T. Yan, K. Dooley, "Structural investigation of supply networks; A social network analysis approach," Journal of Operation Management, Vol. 29, No. 3, pp. 194-211, Mar. 2011. DOI: 10.1016/j.jom.2010.11.001
- E. Choi, "User motivation and expectation for asking a question in online Q&A services," PhD dissertation, Rutgers The State University of New Jersey, School of Graduate Studies, 2014.
- M. R. Morris, J. Teevan, K. Panovich, "What do people ask their social networks, and why? A survey study of status message q&a behavior," Proceedings of the SIGCHI Conference on Human Factors in Computing Systems, Atlanta, Georgia, USA, 2015, pp. 1739-1748.
- N. Shi, M. K. O. Lee, C. M. K. Cheung, H. Chen, "The continuance of online social networks: How to keep people using Facebook?," Proceedings of the 2010 43rd Hawaii International Conference on System Sciences, Washington, DC, USA, 2010, pp. 1-10.
- M. Goyanes, "An empirical study of factors that influence the willingness to pay for online news," Journalism Practice, Vol. 8, No. 6, pp. 742-757, 2014. DOI: 10.1080/17512786.2014.882056
- J. Macnamara, "Remodelling media: The urgent search for new media business models," Media International Australia, Vol. 137, No. 1, pp. 20-35, 2010. DOI: 10.1177/1329878X1013700104
- S. Patil and K. Lee, "Detecting experts on Quora: by their activity, quality of answers, linguistic characteristics and temporal behaviors," Social Network Analysis and Mining, Vol. 6, No. 1, pp. 1-11, 2016. DOI: 10.1007/s13278-015-0313-x
- A. Sindik and G. Graybeal, "Newspaper micropayments and millennial generation acceptance: A brand loyalty perspective," Journal of Media Business Studies, Vol. 8, No. 1, pp. 69-85, 2011. DOI: 10.1080/16522354.2011.11073519
- E. Bonson and M. Bednarova, "The use of YouTube in western European municipalities," Government Information Quarterly, Vol. 35, No. 2, pp. 223-232, 2018. DOI: 10.1016/j.giq.2018.04.001
- F. D. Davis, R. P. Bagozzi and P. R. Warshaw, "User acceptance of computer technology: A comparison of two theoretical models," Management Science, Vol. 35, No. 8, pp. 982-1003, 1989. DOI: 10.1287/mnsc.35.8.982
- C. Gibbs, "Short-form may be long-tail for mobile video," RCR Wireless News, Vol. 26, pp. 12-13, 2007.
- H. Khan, "7 creative ways brands are using vine," PR Daily, Vol. 36, No. 11, pp. 73-77, 2019.
- R. E. Petty and J. T. Cacioppo, "The Elaboration Likelihood Model of Persuasion," Advances in Experimental Social Psychology, Vol. 19, pp. 123-205, 1986. DOI: 10.4135/9781446249215.n12
- Q. Wang, Z. Zhang and R. Gao, "Construction of Counter Effect Evaluation Index System of Internet Public Opinion Based on EMR," in LISS 2014: Proceedings of 4th International Conference on Logistics, Informatics and Service Science, Springer Berlin Heidelberg, pp. 565-570, 2015.