Why do people lose themselves in “hanging out” on social platforms? Why do people spend more time on social media than planned? The reason may be that when people “hang out” on social platforms, encounters that do not involve deliberate choice are enjoyable because the way they happen generates feelings of serendipity. Serendipity is a complicated and interesting concept commonly used in reference to ‘the happy accident’, the finding of things without seeking them. Although people yearn for the power of free choice, serendipity brings more surprises and influence personal behavior significantly. Since automated recommendations have become a pervasive part of the daily social media user experience, individuals perceptions of serendipity designed and recommended by the algorithm may be very different from those existing in the objective world; thus, comprehensively understanding the serendipity recommended by algorithms and its influence mechanism on personal behavior has important theoretical and practical value.

To this end, this study aims to explore individuals’ perception of algorithmic recommendation serendipity and its influencing mechanism on personal behavior. This study will be carried out in five steps. Firstly, introduce and define Perceived Algorithmic Recommendation Serendipity (PARS), derived from an individual’s perception of the serendipity of algorithmic recommendations of social platforms in a digital environment. Secondly, propose the research model and hypotheses. Since PARS is closely related to encounters on social media platforms, the study draws on the perspective of information encountering, a theory referring to that serendipitous information discovery or bumping into useful or interesting information while seeking or browsing for some other information. Then, serendipity is often considered a by-product of browsing, which is defined as a search, hopefully serendipitous; Moreover, browsing is often task-oriented, so we examine the effect of PARS in different task contexts (different browsing strategies). In terms of mediating and moderating variables, we take inspiration as a mediator and information overload as a moderator based on previous studies. As for outcomes, we take social engagement, media stickness as positive effects and problematic social media use as a negative effect. If possible, try to explore the possibility of its positive effects turning into negative ones. Thirdly, based on interviews and scales validated by previous studies on serendipity, form an initial scale of PARS. In the meantime, adapt other constructs’ measurements involved in this study to fit the research context, conduct an informal survey to test the validity and reliability of the questionnaire. Fourthly, conduct formal data collection by experiments, questionnaires through the “Wenjuanxing” platform, and use SEM to test hypotheses. Finally, prepare the manuscript according to the results and discuss its contributions and limitations.

This study has three main theoretical contributions: firstly, PARS is a useful concept for research on algorithm recommendation, digital labor, and other issues; Secondly, PARS is a concept with positive emotional tendencies. However, its significant negative effects should not be neglected. Thirdly, drawing on the role of inspiration and information overload, this study confirms the influencing mechanism of PARS on personal behavior, providing insights for future research. As for the practical implications, this study provides a guide for optimizing the design and application of serendipity for computer science and other disciplines (e.g., business). Most importantly, it also provides some solutions to social problems such as problematic social media use.