Journal of Retailing and Consumer Services 72 (2023) 103240

Contents lists available at ScienceDirect

Journal of Retailing and Consumer Services

journal homepage: www.elsevier.com/locate/jretconser

What motivates users’ viewing and purchasing behavior motivations in live

streaming: A stream-streamer-viewer perspective
Shiyong Zheng a, b, c, Jiada Chen d, Junyun Liao e, *, Hsin-Li Hu f
a
School of Business, Guilin University of Electronic Technology, Guilin, Guangxi, China
b
Management School of Hainan University, Haikou, Hainan, China
c
College of Digital Economics,Nanning University, Nanning, Guangxi, China
d
School of Management, Jinan University, Guangzhou, China
e
Research Institute on Brand Innovation and Development of Guangzhou, School of Management, Jinan University, Guangzhou, China
f
School of Communication, The Hang Seng University of Hong Kong, China

A R T I C L E I N F O A B S T R A C T

Keywords: Live streaming has offered new opportunities for retailers to increase their sales, hence its reception of contin­
Live streaming uous scholarly attention. However, understanding the antecedents of viewer watching and purchasing behaviors
Flow theory in live streaming remains largely insufficient. This study thus constructs a framework from a stream-streamer-
Social presence
viewer perspective to provide a deeper understanding of viewer watching and purchasing behaviors. Using
Interactivity
flow theory, both stream dimension (social presence and interactivity) and streamer dimension (streamer
Optimal stimulation level
attractiveness and streamer expertise) as proposed as antecedents of flow, which motivates viewer behavior.
Moreover, the optimal stimulation level of viewers (i.e., viewer dimension) was found as a moderator which
alters the way stream and streamer dimensions affect flow experience and downstream behaviors. Structural
equation modeling was applied on 367 survey questionnaires to test our framework. Results indicate that both
social presence and interactivity enhance flow. Furthermore, flow has a significant influence on continuous
watching and purchase intention. Meanwhile, optimal stimulation level negatively moderates social presence on
flow, but positively moderates interactivity on it.

1. Introduction streaming market itself is also consistently growing— with 638 million
users in China alone as of 2021, which is a 47.2% increase compared to
The increasing popularity of live streaming has led its use as tool to 2020 figures (Media Entertainment Industry Research Center, 2022).
promote sales by various firms (Mao et al., 2022). The act refers to Promoting user views and purchasing is the fundamental goal of live
simultaneously streaming media which is recorded and broadcast in commerce (Chen et al., 2022; Chen and Liao, 2022; Ma, 2021; Wong­
real-time over social media platforms on the internet (Hilvert-Bruce kitrungrueng and Assarut, 2020). A series of factors have been found to
et al., 2018). Recently, new online sales model based on live streaming affect viewers’ watching and purchasing behaviors. These include
has emerged, aptly called “live commerce” (Ming et al., 2021). In live streamer characteristics (Guo et al., 2022; Liao et al., 2022) such as
commerce, streamers introduce products and instantly communicate source credibility (Lu and Chen, 2021; Park and Lin, 2020; Wongki­
with viewers through live streaming, and who are then able to purchase trungrueng and Assarut, 2020; Zhang et al., 2022) and live-streaming
these introduced items while watching the stream (Zhang et al., 2022). characteristics, such as social immersion and interactivity (Chen and
The real-time video-based product introduction and instant Liao, 2022; Ming et al., 2021; Yang et al., 2022). However, viewers’
streamer-viewer communication is an advantage compared to tradi­ watching and purchasing behaviors can be influenced by the stream,
tional electronic commerce (R. Zheng et al., 2022). Since the COVID-19 streamers and viewers themselves. Therefore, a more comprehensive
outbreak and subsequent lockdown policies, the number of live­ perspective encompassing these dimensions is required to examine
streaming viewers has greatly increased (Chen et al., 2022). The live viewers’ watching and purchasing intentions. The motivations and

* Corresponding author.
E-mail addresses: shiyongzheng123@whu.edu.cn (S. Zheng), 528053213@qq.com (J. Chen), haoyueshan@foxmail.com (J. Liao), sammyhu@hsu.edu.hk
(H.-L. Hu).

https://doi.org/10.1016/j.jretconser.2022.103240
Received 7 August 2022; Received in revised form 20 December 2022; Accepted 21 December 2022
Available online 28 December 2022
0969-6989/© 2022 Elsevier Ltd. All rights reserved.
S. Zheng et al. Journal of Retailing and Consumer Services 72 (2023) 103240

characteristics of viewers who watch live streaming always vary (Zhang elements to stimulate user behavior in a non-gaming environment, is
and Li, 2022), thus viewer characteristics should be considered when widely used in the commercial market and have provide significant re­
investigating their behaviors in live streaming participation. sults (Deterding et al., 2011). On the one hand, selling experiential
Three dimensions (stream, streamer, and viewer) are examined products through live streaming has been purported to increase sales by
herein to explore users’ watching and purchasing behaviors. Social 27.9% compared to pre-recorded ones (Chen et al., 2019). Moreover,
presence and interactivity are considered for the stream dimension merchants further enhance their credibility and reputation by answering
given extant studies which posit social presence and interactivity being consumer queries and communicating their brand philosophy in real
the most significant stream variables in live streaming (Li and Peng, time during the live streaming (Wongkitrungrueng and Assarut, 2020).
2021; S. Zheng et al., 2022). The streamer perspective considers two On the other hand, streamers receive financial support and in­
main characteristics: streamer expertise and attractiveness (Guo et al., centives from viewers, such as gifts and commissions on purchases
2022; Liao et al., 2022). For the viewer dimension, optimal stimulation (Wohn et al., 2018). Furthermore, streamers gain financial support by
level is used herein as it represents the level of anticipation and stimu­ subscribing, donating, and “cheering”, advertising, sponsorships, com­
lation pursued (Rodríguez-Ardura and Meseguer-Artola, 2019), thus petitions and targets, unpredictable rewards for viewers, and imple­
significantly shaping viewer reactions to live streaming. Because flow is menting games into streaming channels themselves, among other things
an essential viewer experience which motivates their engagement with (Johnson and Woodcock, 2019). Also, due to the monetary incentive,
live streaming and purchase behaviors (Chen and Lin, 2018; Kim and streamers further improve their communication skills for a more
Kim, 2022; Li and Peng, 2021; Wang, 2020), a theoretical framework enjoyable viewer experience, increasing viewer willingness to either
based on flow theory is proposed herein to examine how both stream purchase products or give gifts.
and streamer dimension lead to flow, and how the optimal stimulation
level of viewers alter their effect on flow. 2.2. Flow theory
Structural equation modeling (SEM) was used on 367 valid ques­
tionnaires which generated three interesting findings: first, stream Flow is the state of being entirely focused on the current activity and
dimension factors significantly affect flow in live streaming, although filtering out irrelevant information (Csikszentmihalhi, 2020; Csiks­
the streamer dimension does not have a significant effect. The impor­ zentmihalyi, 1990). It emphasizes enjoyment and satisfaction instead of,
tance of flow in enhancing viewers’ watching and purchasing behavior say, financial returns when consumers spend large amounts of energy on
was also confirmed, along with optimal stimulation levels negatively particular activities (Leung, 2020). Generally, experiencing flow in­
moderating the influence of social presence and positively moderating cludes four key dimensions: immersion, perceived control, time distor­
the influence of interactivity on flow. These results expand the literature tion, and enjoyment (Pelet et al., 2017). Immersion is the experience of
on live streaming by providing a more comprehensive framework and the user being fully immersed in the activity and the user’s reactions
offering new insights into users’ watching and purchasing behaviors. from giving full attention to said activity (Nah et al., 2011). Perceived
The rest of the study is organized as follows. First, we summarize the control is the sense that the user is in control of the activity, which is a
relevant literature to support the development of the hypotheses found prerequisite for focus and immersion (Suh et al., 2017). Time distortion
in Section 2. Second, we develop and propose further related hypotheses refers to the user’s complete immersion in the current activity, therein
in Section 3. Third, we report the methodology and results in Section 4. forgetting the passage of time (Agarwal and Karahanna, 2000). Enjoy­
Last, the study presents its theoretical contributions, implications, and ment is the user’s feelings of pleasure and utility, reflecting their
limitations in Section 5. intrinsic interests and motivations (Guo and Poole, 2009). All four di­
mensions combine to reflect the user’s flow experience when interacting
2. Literature review with information systems (Guan et al., 2022).
Flow theory provides insights in explaining the positive online ex­
2.1. Live streaming periences of users (Hyun et al., 2022). Additionally, flow theory also
provides insights in creating an attractive online environment for cus­
Live streaming allows real-time media recording and broadcasting tomers to extend their browsing time (Chen and Lin, 2018; Kim and Kim,
and enables streamer-viewer interaction (Zhang et al., 2022). It has 2022; Li and Peng, 2021; Tong et al., 2022). Some recent studies have
fundamentally changed the way information is disseminated through used flow theory to explain user experience in live streaming (Chen and
(what was once mainly done through) static media such as text and Lin, 2018; Kim and Kim, 2022; Li and Peng, 2021; Liao et al., 2022). Liao
images, altering the one-way delivery of information to consumers in et al. (2022) even found that streamers’ interaction-oriented commu­
traditional shopping contexts (Sun et al., 2019). Viewers, streamers, and nication enhances immersion and prosocial interaction. Kim and Kim
platforms have, in turn, become the key participants. Because of the (2022) found that streamer attributes (friendship) and live-streaming
actual interaction and the scrolling danmu, viewers can actively mobile application attributes (technical and functional quality, conve­
participate and provide feedback. As a result, they become content nience, and security) significantly impacted consumers’ game loyalty
producers instead of passive information recipients (Kang et al., 2021) when doing online games. Similarly, this study uses flow theory to
and are also able to reward streamers for their valuable content. investigate flow formation from the stream-streamer-viewer
Live streaming with real-time interactions enhances social presence perspective.
and makes users feel as if they are present (Chen and Liao, 2022).
Streamers, in turn, motivate the viewers by adopting various strategies 2.3. Optimal stimulation level
such as responsiveness, humor, and motivation (Guo et al., 2022; Kim
and Kim, 2022; Liao et al., 2022). During which, viewers get acquainted Individuals are internally motivated to obtain stimulation at a spe­
with the streamer and may be drawn to their humorous and empathetic cific expectation level (Rodríguez-Ardura and Meseguer-Artola, 2019),
content, further deepening both connection and attachment. Moreover, with the optimal stimulation level representing the amount of stimula­
the streamer’s vicarious expression and synchronicity cater to viewers’ tion people prefer in life (McReynolds, 1971; Richard and Chebat,
psychological needs to express themselves, especially in electronic 2016a). Optimal stimulation level is a universal personality trait which
gaming and sports live streaming (Li and Guo, 2021; Liu et al., 2022). is part of people’s tendency to be stimulated, and greatly varies from
Affinity and intimacy with live streaming may also be transferred to very low to very high per individual. Those with high optimal stimula­
platforms or products (Zhang et al., 2022). tion constantly require higher levels of arousal or stimulation to main­
Most live streaming incorporate elements from gameplay to better tain psychological pleasure compared to those with low optimal
engage viewers (). Gamification, which is incorporating game design stimulation levels (Park and Jang, 2014). They therefore crave exposure

2
S. Zheng et al. Journal of Retailing and Consumer Services 72 (2023) 103240

to new, positive stimulation environments to satisfy the need for stim­ 3.3. Impact of streamer attractiveness
ulation, novelty, and adventure. In contrast, people with low optimal
stimulation levels prefer to feel comfortable in their familiar environ­ Attractiveness is another critical characteristic of streamers (Guo
ment (Rodríguez-Ardura and Meseguer-Artola, 2019). et al., 2022). Generally, attractive streamers are more favored due to
Optimal stimulation level is an essential individual trait that influ­ their various personality traits such as social ability and intelligence (Till
encing consumer emotion (Mahatanankoon, 2007), attitude (Lichtlé, and Busler, 2000). Many studies have shown attractive communicators
2007), and behavior (Richard and Chebat, 2016a). When deviating from carrying and transmitting information to influence consumer behavior
the optimal stimulation level, users seek ways and means to adjust (Baker and Churchill Jr, 1977). Streamer attractiveness was also shown
themselves (Raju, 1980). When below optimal stimulation level, users to increase warmth (Guo et al., 2022), which, in turn, is related to
become bored and explore novel activities to keep themselves enter­ preferences and social expectations along with trust, morality, sincerity,
tained. And when above the optimal stimulation level, users reduce or and kindness as main characteristics (Cuddy et al., 2008).
simplify their current activities. Consumers with high optimal stimula­ In live streaming, gender also shapes the streamer’s overall
tion levels are more willing to try new products and brands (McAlister temperament and image, impacting the viewers’ desired goals (Ruberg
and Pessemier, 1982; Raju, 1980). Those with lower optimal stimulation and Brewer, 2022). Therefore, choosing the streamer gender wisely
levels, however, tend to be more willing to purchase products and effectively improves the user’s perceived fit and closeness to the live
brands they are familiar with— this is because optimal stimulation streaming. Additionally, affective labor is an essential influencing factor
levels have goal-directed characteristics which moderate consumer which constitutes streamer attraction (Jodén and Strandell, 2022; Wang,
emotions and behaviors (Raju, 1980). In toto, because the optimal 2020; Woodcock and Johnson, 2019). As viewers connect more
stimulation level may determine viewers’ response to stimulation from emotional energy with the streamer, they also further trigger positive
live streaming, the current study considers it as an important moderator associations and emotions towards the streamer, thus building shared
in the development of flow. connections and emotions (Jodén and Strandell, 2022; Woodcock and
Johnson, 2019). Inclusive strategies such as common noun referencing
3. Research hypotheses are also essential in increasing streamer attractiveness (Jodén and
Strandell, 2022).
3.1. Impact of social presence In live streaming, the streamer acts as the spokesperson of the
product or brand (Guo et al., 2022). Attractive streamers are, therefore,
A recent study identified emotional support, interactivity, and sense likelier to grab the viewer’s attention, further improving the favorable
of community as critical influencing dimensions of social presence attitude of viewers toward information sources and increasing streamer
(Chen and Liao, 2022). Social presence ultimately achieves user pres­ persuasiveness (Chen and Liao, 2022; Guo et al., 2022; Liao et al., 2022).
ence by building alternative experiences and social interactions (Hu Furthermore, streamer attractiveness causes a sense of pleasure (Liu,
et al., 2017). It also enhances users’ source credibility by establishing 2020), enabling users to immerse themselves in live streaming and enjoy
real-time alternative product experiences and real-time interactive value the flow (Kim and Kim, 2022). Therefore, the following hypothesis is
transfer, thus reducing user uncertainty and purchase hesitation (Lu and also proposed:
Chen, 2021; Wongkitrungrueng and Assarut, 2020; Zhang et al., 2022).
H3. Streamer attractiveness in live streaming positively influences
Most importantly, it promotes user interaction, intimacy, warmth, en­
flow.
hances pleasure (Ma, 2021), and increases the overall sense of belonging
(Sun et al., 2019; Wongkitrungrueng and Assarut, 2020). Therefore, the
3.4. Impact of streamer expertise
social presence makes the user experience enjoyable and immersive,
likelier making users to become absorbed. Hence, the following hy­
Expertise is the streamer’s specific skill, knowledge, or ability. The
pothesis is proposed:
greater the communicator’s expertise, the greater the trust as well
H1. Social presence in live streaming positively influences flow. (Crisci and Kassinove, 1973). Sales literature has also recognized the
significance of the seller’s expertise (i.e., knowledge about the product
3.2. Impact of interactivity and its application) (Agnihotri et al., 2009). Knowledgeable salespeople
have better cognitive abilities to anticipate customer responses and
Interactivity is a form of communication which provides active modify their behavior to meet customer needs (Jones et al., 2002).
control and synchronized sharing of information with each other (Liu, Salespeople easily make the actual selling behavior when they have
2003; Mutum and Ghazali, 2011). In social media, interactivity has three sufficient technical knowledge (Stock and Hoyer, 2005).
main dimensions: human-to-human (two-way communication), In live streaming, streamer expertise significantly shapes users’ in­
human-to-medium (synchronization), and human-to-machine (active tentions and behavior (Guo et al., 2022): the higher the streamer’s
control) (Liu and Shrum, 2002; Mutum and Ghazali, 2011; Wu and Wu, expertise, the lower the information-seeking effort of the user (Hu et al.,
2006). Interactivity is an essential atmospheric cue which stimulates and 2016). This means that the more energy is invested in engaging with
influences consumer emotions through received response and perceived streamers, the likelier a viewer enters a state of flow. Furthermore, the
control (Mutum and Ghazali, 2011). Users generally interact with expertise of streamers leads to a higher degree of participation, thereby
streamers by liking, sending pop-ups, and sharing and giving virtual gifts immersing viewers in live streaming and bringing a higher sense of flow
(Yu et al., 2018); such interactivity is a sign of high user engagement and (Kim and Kim, 2022; Liao et al., 2022). Therefore, the following hy­
involvement (Li et al., 2021; Li and Peng, 2021). Furthermore, inter­ pothesis is also proposed:
active users are more connected with live streaming (Chen and Liao,
H4. Streamer expertise in live streaming positively influences flow.
2022; Kang et al., 2021) and gain more enjoyment (Foster et al., 2022;
Kim et al., 2021; Ma, 2021). Thus, users are also likelier to enter the flow
3.5. Impact of flow on continuous watching and purchase intention
when the live streaming is highly interactive (Huang and Hsu Liu, 2014).
Therefore, the following hypothesis is also proposed:
Flow has been considered a significant factor affecting consumer
H2. Interactivity in live streaming positively influences flow. attitude, willingness, and behavior (Obadă, 2013). In live streaming,
flow represents the user’s focused and enjoyed state (Yang and Lee,
2018a). Users become more focused on the live streaming and linger
when they enter a state of flow. This pleasurable feeling creates a

3
S. Zheng et al. Journal of Retailing and Consumer Services 72 (2023) 103240

stronger desire to keep watching (Kim and Kim, 2020). Streaming, The research framework is summarized in Fig. 1.
moreover, is a positive state which represents enjoyment and concen­
tration. Consumer enjoyment leads to exploratory behavior and an 4. Data and methodology
increased propensity to purchase (Wu et al., 2020). Furthermore, a
focused state is more receptive to marketing cues from merchants and an 4.1. Questionnaire and data collection
increase in purchase intentions (Ming et al., 2021; Tong et al., 2022).
Therefore, the following hypotheses are also proposed: The questionnaire was marked on the professional questionnaire
platform, WenJuanXing, while the survey was conducted through a
H5. The flow in live streaming positively affects viewers’ continuous
web-based questionnaire. The main content of the questionnaire was
watching intention.
divided into three part: the first part contains the basic information of
H6. The flow in live streaming positively affects viewers’ purchase the respondents, referring to their experience in watching live stream­
intention. ing, such as watching frequency and purchase frequency. Here,
screening questions were set up to filter respondents, including ques­
tions on whether they have watched live streaming and purchased goods
3.6. Moderating effect of optimal stimulation level in the last two months. Only respondents who replied “Yes” were invited
to proceed. The second part is the formal measurement items, followed
Optimal stimulation levels represent the amount of stimulation by the third which is the basic demographic information.
people prefer in life (Richard and Chebat, 2016a). Individuals also Online Chinese consumers were used as the survey’s main target
experience pleasure when certain stimuli elicit optimal stimulation sample; hence, the questionnaire questions were in Chinese. To avoid
levels (Im et al., 2021). In live streaming, the stream and the streamer semantic inconsistencies, the measurement of all latent variables were
are the most arousing cues— in terms of stream features, social presence adapted from established literature, translated, and then proofread by
makes the user comfortable (Ogara et al., 2014), and interactivity en­ two marketing professors using the back-to-back translation method
ables the user to feel enjoyment (Zhang et al., 2022). Users with high (Ares, 2018). Using the WenJuanXing sampling services, the question­
optimal stimulation level may be aroused by seeking newer and more naire was randomly distributed to 1536 users who were active online
exciting live streaming session. Therefore, social presence and inter­ following the firm’s private knowledge and standard. Each user was
activity may also arouse users who have high optimal stimulation levels. restricted to answer the questionnaire. To encourage user participation,
Aroused users are thus likelier to enter the flow status. In contrast, a low respondents who passed the screening process and completed the
optimal stimulation level may simplify and reduce current activities questionnaire were awarded 3 RMB. 400 questionnaires were ultimately
when they feel overly satisfied (Novak and Hoffman, 1997; Richard and collected, reflecting a 26% response rate. We then removed those with
Chebat, 2016a). Hence, users with low optimal stimulation levels may completed the questionnaire in a very short time (less than 120 s),
not enter the flow as easily as their counterparts despite high social leaving 367 valid responses as part of our final sample.
presence and interactivity. Essentially, the optimal stimulation level The sample’s descriptive statistics are shown in Table 1 below. Of the
positively moderates the relationship between social presence and surveyed samples, 45.23% were male and 54.77% were female. User age
interactivity and flow. Similarly, streamer attractiveness and expertise is mainly distributed in the 18–23 range accounting for 40.08% of the
improve viewers’ pleasure and source credibility, thereby appealing to total sample, followed by the 23–35 range, accounting for 38.97%.
users of high optimal stimulation level as well (Guo et al., 2022). Given These indicate that the younger age group (18–23) is the leading group
these discussions, the following hypotheses are also proposed: actively participating in live commerce. The proportion of college and
H7a. Optimal stimulation level positively moderates the effect of so­ undergraduate students also reached 60.76%. Furthermore, respondents
cial presence on flow. clearly watched live streaming on a frequent basis, thus meeting the
requirements of the survey group. The demographic distributions of the
H7b. Optimal stimulation level positively moderates the effect of sample in this paper were consistent with previous live streaming
interactivity on flow. research conducted in China (Chen and Liao, 2022; Guo et al., 2022; Li
H7c. Optimal stimulation level positively moderates the effect of et al., 2021; Liao et al., 2022). Therefore, the sample is not subject to a
streamer attractiveness on flow. large background potential bias.

H7d. Optimal stimulation level positively moderates the effect of

streamer expertise on flow.

Fig. 1. Research model.

4
S. Zheng et al. Journal of Retailing and Consumer Services 72 (2023) 103240

Table 1 Table 2
Descriptive statistics. Results of confirmatory factor analysis.
Type Frequency Percent Conception Title item Item
(%) loading

Gender Male 166 45.23 Social presence (SP) There is a sense of personal connection in 0.859
Female 201 54.77 Cronbach’s α = 0.785 watching live streaming.
Age Under 18 13 3.54 CR = 0.886 There is a sense of humanity in watching 0.760
Age 18-23 147 40.05 AVE = 0.660 live streaming.
Age 23-35 143 38.97 There was a kind of human warmth in 0.820
Age 35-50 63 17.17 watching live streaming.
Over 50 1 0.27 There is a human sensitivity to watching 0.807
Income (monthly) Students 156 42.51 live streaming.
3000 RMB and below 4 1.09 Interactivity (I) I am willing to send pop-ups and feedback. 0.819
3000–5000 RMB 92 25.06 Cronbach’s α = 0.852 I am willing to respond to the streamer’s 0.833
5000–8000 RMB 83 22.62 CR = 0.896 request and give feedback.
8000–12,000 RMB 17 4.63 AVE = 0.633 I want to give thumbs up, give gifts, and 0.815
12,000 RMB or more 15 4.09 share.
Education High School and 51 13.89 Streamers are happy to connect with the 0.730
below audience and answer questions.
Undergraduate 223 60.76 Streamers are willing to make personal 0.777
Master 90 24.52 connections with their viewers.
Ph.D. 3 0.08 Streamer Attractiveness Streamers give me a good feeling. 0.831
Watching frequency (per Less than 1 day 52 14.17 (SA) Streamers are charming. 0.826
week) 1–3 days 57 15.53 Cronbach’s α = 0.781 I am profoundly attracted by streamers. 0.834
4–6 days 133 36.24 CR = 0.874
Every day 125 34.06 AVE = 0.697
Streamer Expertise (SE) Streamers are qualified. 0.815
Cronbach’s α = 0.820 Streamers are skilled. 0.798
4.2. Measurement CR = 0.885 Streamers are experienced. 0.828
AVE = 0.658 Streamers are professional. 0.803
Optimal Stimulation Instead of trying something new or 0.653
The formal measurement part was divided into “social presence,” Level (OSL) different, I prefer to do something familiar.
“interactivity,” “streamer attractiveness,” “streamer expertise,” " Cronbach’s α = 0.909 *
optimal stimulation level,” “continuous watching intention,” and “pur­ CR = 0.927 I like to experience novelty and change in 0.840
chasing intention” where a five-point Likert scale was used (unless AVE = 0.647 my daily life.
I like a job that brings variety, even with 0.803
specified otherwise, 1 means “strongly disagree” and 5 means “strongly
some dangers.
agree”). To measure construct items herein, items used in the survey I prefer a fixed lifestyle to one that is 0.745
were adapted from previous literature to fit the live streaming context. unpredictable.*
For example, seven items for optimal stimulation level were adapted I’m always looking for new ideas and 0.864
from measures found in an extant study (Yang and Lee, 2018a), and the experiences.
I like changing activities. 0.857
measurement of social presence and purchasing intention comes from I like to look for new and unfamiliar 0.847
the study by Sun et al. (2019). Moreover, flow in live streaming was experiences when things get boring.
measured using four items, and interactivity with five (Chen and Lin, Flow (F) I forgot the passage of time while watching 0.825
2018). Streamer attractiveness was also measured using three items, and Cronbach’s α = 0.790 live streaming.
CR = 0.905 I was able to concentrate without being 0.853
streamer expertise with four items as previously measured by Guo et al.
AVE = 0.705 distracted while watching live streaming.
(2022). Finally, the continuous watching intention was measured in I often forget what I need to do when I 0.838
three items (Hu et al., 2017) (see Table 2). watch live streaming.
Watching live streaming took me away 0.843
4.3. Common method bias from the real world for a while.
Purchase Intention (PI) I hope I can purchase the products 0.868
Cronbach’s α = 0.837 recommended by streamers in the future.
Common method bias is an important and inevitable issue. Thus, CR = 0.895 I may purchase products recommended by 0.776
several measures were done herein to discourage common method bias AVE = 0.681 streamers in the future.
(Malhotra et al., 2006)). First, the responses made by the respondents I plan to purchase products recommended 0.838
by streamers through live streaming in the
were anonymized. Second, the questionnaire was paginated so that the
future.
independent variable did not precede the dependent variable. Following I will purchase the products recommended 0.816
Podsakoff et al. (2003), we also performed an unrotated principal by streamers in the future.
component factor analysis on all items. Here, the highest variance Continuous Watching I plan to continue watching live streaming 0.833
explained by a single factor was 43.087% (<50%), which could not Intention (CWI) in the future.
Cronbach’s α = 0.790 I’m going to keep watching live streaming. 0.846
explain most of the variance in the study (Malhotra et al., 2006). CR = 0.881 I will always try to watch live streaming. 0.851
Therefore, common variance bias is not a severe problem herein. AVE = 0.711

Note: OSL1 and OSL4 are reverse items.

4.4. Reliability and validity testing

Structural equation modeling (SEM) was constructed and subse­ The discriminant validity test requires that the correlations between
quently ran through AMOS. Before hypothesis testing of the research items in any two constructs be lower than the square root of the average
model, both the reliability and validity of items in the scale were eval­ variance shared by items within a construct (Hu et al., 2017) (see
uated. Three test criteria were adopted to evaluate reliability and val­ Table 3). The exploratory factor analysis of the samples found that the
idity: first, for reliability, the C.R. and Cronbach’s α of each variable factor load value of the item “watching live streaming temporarily
should be above 0.7. Second, the average variance extracted (AVE) from separated me from the real world” was far below 0.7, which did not
a construct should exceed the recommended level of 0.5. Third, reach the suggested threshold, and was therefore deleted. Cronbach’s α
discriminant validity was assessed to ensure differing constructs. scores of each variable fell in the range of 0.781–0.852. Therefore, all

5
S. Zheng et al. Journal of Retailing and Consumer Services 72 (2023) 103240

Table 3
Correlation matrix of variables.
Variables SP I SA SE OSL F CWI PI

SP 1
I 0.46 * * * 1
SA 0.41 * * * 0.45 * * * 1
SE 0.35 * * * 0.36 * * * 0.40 * * * 1
OSL 0.12 * * 0.15 * * * 0.14 * * * 0.12 * * * 1
F 0.43 * * * 0.45 * * * 0.39 * * * 0.30 * * * 0.15 * * * 1
CWI 0.44 * * * 0.48 * * * 0.42 * * * 0.36 * * * 0.14 * * * 0.43 * * * 1
PI 0.46 * * * 0.49 * * * 0.43 * * * 0.36 * * * 0.11 * * 0.42 * * * 0.46 * * * 1
Mean 3.97 3.99 4.07 4.25 3.55 3.86 4.01 4.01
SD 0.77 0.79 0.79 0.69 0.96 0.84 0.82 0.76

Note: *: P < 0.05; **: P < 0.01; ***: P < 0.001.

constructs passed the test of construct reliability, indicating adequate or low-level group, applying the covariance matrix separately. Next,
internal reliability. multiple structural equation analysis were then employed (Helm and
The results also show that the composite reliability value of the Landschulze, 2009). Consequently, the score of the optimal stimulation
variable was between 0.881 and 0.927, indicating that the combined level in the high-level group is above the average level (N = 190), ac­
reliability was high. The AVE value of each item was greater than 0.5, counting for 51.77%. Meanwhile, the score of the optimal stimulation
and the loading value of each factor in exploratory factor analysis was level in the low-level group is below the average level (N = 177), ac­
greater than 0.7, thus suggesting a reasonable degree of discriminant counting for 48.23%.
validity. According to the leading fit indices yielded regarding the pro­ To further analyze the differences between the two optimal stimu­
posed model, the structural model seemed to fit the observed data, as all lation levels group in the structural model, the Chi-square difference test
fit indices noticeably reached the acceptable level (χ 2 = 1315.99, df = was conducted by constructing the basic model (without any limitation)
510, χ 2/df = 2.58; CFI = 0.92; GFI = 0.91; NFI = 0.886; and RMSEA = and the constraint model (the path coefficients in the structural equation
0.053). This model therefore fitted the data reasonably well. In sum, all models with different optimal stimulation levels were set to be equal).
factors exhibited sufficient convergent and discriminant validity. The moderation effect is significant when the difference (△χ2) with one
degree of freedom is more significant than the critical value of 3.84. As
4.5. Hypothesis testing shown in Table 5, there were significant differences between the users of
low optimal stimulation level (β = 0.71) and high optimal stimulation
The structural model was used to test the hypotheses. The results of level (β = 0.37) in terms of the relationship between social presence and
path coefficients and corresponding levels of significance are shown in flow (△χ2 = 5.136, p < 0.05), thus suggesting that the optimal stim­
Table 4 below (see Fig. 2). According to the results of hypothesis testing, ulation level negatively moderates the effect of social presence on flow.
the path coefficient from social presence to flow in stream factors is 0.63. Therefore, H7a was not supported.
The result is significant, thus H1 is supported. The path coefficient from Likewise, there were significant differences between the users of low
interactivity to flow is 0.32, indicating user interactivity to positively optimal stimulation level (β = 0.13) and high optimal stimulation level
affect flow. Thus, H2 is also supported. (β = 0.55) in terms of the relationship between interactivity and flow
However, the standardized path coefficient of social presence and (△χ2 = 4.434, p < 0.05), suggesting the optimal stimulation level
interactivity is significantly different: the standardized path coefficient positively moderates the effect of interactivity on flow. Therefore, H7b
of social presence to flow (0.63) is significantly larger than interactivity is, in contrast, supported. However, streamer attractiveness and flow
(0.32), positing that social presence has a more significant impact on were not significantly different (△χ2 = 1.119, p > 0.05), neither were
flow compared to interactivity. Streamer attractiveness and streamer streamer expertise nor flow (△χ2 = 1.564, p > 0.05). Therefore, H7c
expertise did not pass the hypothesis test. Therefore, H3 and H4 are not and H7d were both not supported.
supported.
Following the results, the standardized path coefficients of flow on 5. Discussion
the intention to continue watching and purchasing are 0.97 (p < 0.001),
indicating user flow to positively impact continued watching and pur­ The study presents several findings. First, social presence and
chasing intention. Therefore, H5 and H6 are both supported. interactivity in live streaming positively impact the generation of flow,
paralleling extant studies regarding consumption decisions in live
streaming (Li and Peng, 2021; Ming et al., 2021). Moreover, comparing
4.6. Analysis of moderating effect the path coefficient, social presence perceived by consumers tends to be
more influential compared to interactivity on the formation of flow.
To reveal the difference among various optimal stimulation levels of Streamer attractiveness and streamer expertise, however, do not
consumers experiencing flow, this study divided the sample data ac­ significantly influence flow.
cording to the optimal stimulation level into either best high-level group Second, this study confirms the importance of flow and continuous
watching intention for consumer behavior, thus resonating with previ­
Table 4 ous studies on flow experience which consistently indicated its posi­
Hypothesis testing results. tively impact on users’ continuous watching and purchasing intention
Path Standardized Path Coefficient S.E. p Result (Chen and Lin, 2018; Li and Peng, 2021; Yang and Lee, 2018b).
SP → F 0.63 0.12 *** Supported Third, the optimal stimulation level negatively moderates social
I→F 0.32 0.12 * Supported presence on flow, but positively moderates interactivity on flow. For
SA → F 0.12 0.17 0.51 Not supported users with a high optimal stimulation level, the more interactive the live
SE → F 0.06 0.11 0.61 Not supported streaming is, the more they can concentrate on the live streaming and
F → CWI 0.97 0.07 *** Supported
enter a state of flow. However, more interactivity may not be optimal for
F → PI 0.97 0.07 *** Supported
consumers of low optimal stimulation level. Previous studies have also
Note: *: p < 0.05; **: p < 0.01; ***: p < 0.001.

6
S. Zheng et al. Journal of Retailing and Consumer Services 72 (2023) 103240

Fig. 2. Test results of path coefficients.

on the influencing factors of flow and users’ continuous use of live

Table 5
streaming platforms, thereby enriching the current research on live
Testing the moderation of optimal stimulation level.
commerce.
Paths Δχ2 Path coefficient Third, although there is a general scholarly consensus on stream and
The low optimal The high optimal streamer roles (Chen and Lin, 2018; Guan et al., 2022; Li and Peng,
stimulation level stimulation level 2021; Liao et al., 2022), explorations into users’ psychological states and
Social Presence → 5.13 0.71 *** 0.37 *** motivations produce different results. Accordingly, we identified a
Flow * moderating role of optimal stimulation level in flow formation. Mean­
Interactivity → Flow 4.43 0.13 * 0.55 *** while, although most previous studies on live streaming mainly take
*
demographic characteristics, such as gender and age, as personal traits
Streamer 1.11 0.07 0.04
Attractiveness → (Chen et al., 2020; Hu et al., 2017), our research shows that the influ­
Flow ence of flow factors is a dynamic process— We found that optimal
Streamer Expertise → 1.56 0.14 0.06 stimulus levels play different moderating roles in different pathways.
Flow
Specifically, optimal stimulation level negatively moderates social
Note: *: P < 0.05; **: P < 0.01; ***: P < 0.001. presence on flow, but positively moderates interactivity on flow. This
provides fresh insights in further understanding the role of flow, thus
found that individuals with low optimal stimulation levels likelier feel filling a relevant research gap.
anxious when interacting with a computer-mediated environment
(Helm and Landschulze, 2009). 5.2. Practical implications
Results indicating that the optimal stimulation level negatively
moderates the effect of social presence and flow may be due to users First, streamers, platforms, and viewers should each try to create a
concentrating on interaction with streamers during live streaming, better live streaming experience: interactivity clearly facilitates viewers’
rather than the presence of other users. In some situations, social pres­ entry into a state of flow and, in turn, increases their willingness to
ence may even disrupt users from experiencing flow because they do not engage in positive behaviors. Improving interactivity in live streaming
want their behaviors in live streaming to be caught by other viewers. requires the efforts of multiple parties. From the platform perspective,
platforms should focus on developing interactive functions and tools
such as real-time visual interactions to allow streamers and viewers to
5.1. Theoretical contributions interact better.
From the streamer’s perspective, streamers should increase the op­
Our research contributes to the literature by offering three crucial portunity to interact with viewers and answer viewers’ questions
insights into live commerce: First, we proposed a framework in promptly to enable viewers to produce a more substantial flow experi­
exploring users’ viewing and purchasing intentions from the stream- ence. From the viewer’s perspective, viewers should also positively
streamer-viewer perspective. Scholars have studied users’ viewing and respond to the streamer and other viewers, heling promote a range of
purchasing intentions from different perspectives (Li and Peng, 2021; psychological aspects such as self-expression, pleasure and satisfaction.
Ming et al., 2021). However, these studies always either focused on the An excellent live interactive environment therefore requires the efforts
stream (Li and Peng, 2021; Ming et al., 2021) or the streamer dimensions of the stream-streamer-viewer.
(Kim and Kim, 2022). Only limited studies have therefore examined Second, platforms should provide customized services and different
their effects in a comprehensive framework, or have examined their types of live streaming content to satisfy viewers with varied optimal
interaction effects. Furthermore, viewer dimension has received few stimulation level. Everyone has their optimal stimulation level, which
scholarly attention. The current study explores viewing and purchasing positively moderates the relationship between interactivity and flow. At
intention from the stream-streamer-viewer perspective and generates a high optimal stimulation level, the more interactive users are, the more
several new insights. they can concentrate on the live streaming and enter a state of flow
Second, our results confirm that, generally, continuous watching without thinking about it. Thus, we should provide new stimuli to users
intention and purchase intention are affected by both interactivity and with new optimal stimulus levels. Furthermore, people who watch live
social presence, which is a consistent conclusion in the traditional streaming users mostly do so to beat boredom, seek experience, among
internet environment (Hsu et al., 2012). Although, in the development others. Therefore, new stimuli and feelings further deepen the overall
of internet technologies, the usability and practicability of the live experience.
platform are no longer the main barriers which hinder consumers from
continued use, user experience in information systems nonetheless
significantly influences consumer behavior (Chen et al., 2020). There­
fore, this study provides a novel perspective and direction for research

7
S. Zheng et al. Journal of Retailing and Consumer Services 72 (2023) 103240

5.3. Limitations and future research directions Crisci, R., Kassinove, H., 1973. Effect of perceived expertise, strength of advice, and
environmental setting on parental compliance. J. Soc. Psychol. 89, 245–250. https://
doi.org/10.1080/00224545.1973.9922597.
This study has several limitations, which may offer directions for Csikszentmihalhi, M., 2020. Finding Flow: the Psychology of Engagement with Everyday
future research. First, laboratory experiments or field experiments are Life (Hachette UK).
recommended to provide deeper insights for live streaming. Second, Csikszentmihalyi, M., 1990. Flow: the psychology of optimal experience. In: Flow: the
Psychology of Optimal Experience. Harper & Row, New York.
although the current study explores live streaming users’ watching and Cuddy, A.J.C., Fiske, S.T., Glick, P., 2008. Warmth and competence as universal
purchasing intentions from the stream-streamer-viewer perspective, dimensions of social perception: the stereotype content model and the BIAS map. In:
only several prominent variables were selected for each dimension. Advances in Experimental Social Psychology. Elsevier, pp. 61–149. https://doi.org/
10.1016/S0065-2601(07)00002-0.
Other dimensional variables, such as the immediacy and entertainment Deterding, S., Dixon, D., Khaled, R., Nacke, L., 2011. From game design elements to
of live streaming, may be explored by future research. Third, although gamefulness: defining “gamification. In: Proceedings of the 15th International
this study pioneeringly introduces the optimal stimulation level in live Academic MindTrek Conference: Envisioning Future Media Environments, pp. 9–15.
https://doi.org/10.1145/2181037.2181040.
streaming research, the optimal stimulation level is a general personality Foster, J.K., McLelland, M.A., Wallace, L.K., 2022. Brand avatars: impact of social
trait and is goal-oriented, which necessitates further exploration of its interaction on consumer–brand relationships. JRIM 16, 237–258. https://doi.org/
role in live streaming. Finally, we found that streamer attractiveness and 10.1108/JRIM-01-2020-0007.
Guan, Z., Hou, F., Li, B., Phang, C.W., Chong, A.Y.-L., 2022. What influences the
expertise cannot directly affect the flow. However, there may be more purchase of virtual gifts in live streaming in China? A cultural context-sensitive
interesting variables playing a mediating role, which calls for deeper model. Inf. Syst. J. 32, 653–689. https://doi.org/10.1111/isj.12367.
explorations. Guo, Y., Zhang, K., Wang, C., 2022. Way to success: understanding top streamer’s
popularity and influence from the perspective of source characteristics. J. Retailing
Consum. Serv. 64, 102786 https://doi.org/10.1016/j.jretconser.2021.102786.
Data availability Guo, Y.M., Poole, M.S., 2009. Antecedents of flow in online shopping: a test of alternative
models. Inf. Syst. J. 19, 369–390. https://doi.org/10.1111/j.1365-
The data that has been used is confidential. 2575.2007.00292.x.
Helm, R., Landschulze, S., 2009. Optimal stimulation level theory, exploratory consumer
behaviour and product adoption: an analysis of underlying structures across product
Acknowledgments categories. Rev. Manag. Sci. 3, 41–73. https://doi.org/10.1007/s11846-009-0024-7.
Hilvert-Bruce, Z., Neill, J.T., Sjöblom, M., Hamari, J., 2018. Social motivations of live-
streaming viewer engagement on twitch. Comput. Hum. Behav. 84, 58–67. https://
This project is supported by National Natural Science Foundation of doi.org/10.1016/j.chb.2018.02.013.
China (NSFC) (72272061&71802097); The Ministry of education of Hsu, C.-L., Chang, K.-C., Chen, M.-C., 2012. Flow experience and internet shopping
Humanities and Social Science project(22YJC630077); Philosophy and behavior: investigating the moderating effect of consumer characteristics: flow
experience and internet shopping behavior. Syst. Res. 29, 317–332. https://doi.org/
Social Sciences Planning Program of Guangzhou (2021GZYB05 and 10.1002/sres.1101.
2022JDGJ06); Research Institute on Brand Innovation and Development Hu, M., Zhang, M., Wang, Y., 2017. Why do audiences choose to keep watching on live
of Guangzhou (2021CS05); Jinan University Management School video streaming platforms? An explanation of dual identification framework.
Comput. Hum. Behav. 75, 594–606. https://doi.org/10.1016/j.chb.2017.06.006.
Funding Program (GY21013); Institute for Enterprise Development, Hu, X., Huang, Q., Zhong, X., Davison, R.M., Zhao, D., 2016. The influence of peer
Jinan University, Guangdong Province (2021MYZD04 and 2020CP03), characteristics and technical features of a social shopping website on a consumer’s
China Postdoctoral Science Foundation (No, 2022M710038); Guangxi purchase intention. Int. J. Inf. Manag. 36, 1218–1230. https://doi.org/10.1016/j.
ijinfomgt.2016.08.005.
Science and Technology Base and Talent Special Project (No,
Huang, T.-L., Hsu Liu, F., 2014. Formation of augmented-reality interactive technology’s
2020AC19034), Guangxi 14th Five-Year Education Science Planning persuasive effects from the perspective of experiential value. Internet Res. 24,
Key Special Project (No, 2021A033 && 82–109. https://doi.org/10.1108/IntR-07-2012-0133.
2021ZJY1607&&2022ZJY1699), Innovation Project of Guangxi Grad­ Hyun, H., Thavisay, T., Lee, S.H., 2022. Enhancing the role of flow experience in social
media usage and its impact on shopping. J. Retailing Consum. Serv. 65, 102492
uate Education(No, JGY2022122), Guangxi undergraduate teaching https://doi.org/10.1016/j.jretconser.2021.102492.
reform project (No, 2022JGB180) and Special Funds for the Cultivation iiMedia Entertainment Industry Research Center, 2022. China live streaming Industry
of Guangdong College Students’ Scientific and Technological Innovation development status and market research analysis. Report 2022 [WWW Document].
URL. https://www.iimedia.cn/c400/84858.html.
(“Climb- ing Program” special funds pdjh2023a0059). Im, H., Ju, H.W., Johnson, K.K.P., 2021. Beyond visual clutter: the interplay among
products, advertisements, and the overall webpage. JRIM 15, 804–821. https://doi.
References org/10.1108/JRIM-10-2020-0213.
Jodén, H., Strandell, J., 2022. Building viewer engagement through interaction rituals on
Twitch.tv. Inf. Commun. Soc. 25, 1969–1986. https://doi.org/10.1080/
Agarwal, R., Karahanna, E., 2000. Time flies when you’re having fun: cognitive
1369118X.2021.1913211.
absorption and beliefs about information technology usage. MIS Q. 24, 665–694.
Johnson, M.R., Woodcock, J., 2019. “And today’s top donator is”: how live streamers on
https://doi.org/10.2307/3250951.
Twitch. tv monetize and gamify their broadcasts. Soc. Med. Soc. 5,
Agnihotri, R., Rapp, A., Trainor, K., 2009. Understanding the role of information
2056305119881694 https://doi.org/10.1177/2056305119881694.
communication in the buyer-seller exchange process: antecedents and outcomes.
Jones, E., Sundaram, S., Chin, W., 2002. Factors leading to sales force automation use: a
J. Bus. Ind. Market. 24, 474–489. https://doi.org/10.1108/08858620910986712.
longitudinal analysis. J. Personal Sell. Sales Manag. 22, 145–156. https://doi.org/
Ares, G., 2018. Methodological issues in cross-cultural sensory and consumer research.
10.2307/40471879.
Food Qual. Prefer. 64, 253–263. https://doi.org/10.1016/j.foodqual.2016.10.007.
Kang, K., Lu, J., Guo, L., Li, W., 2021. The dynamic effect of interactivity on customer
Baker, M.J., Churchill Jr., G.A., 1977. The impact of physically attractive models on
engagement behavior through tie strength: evidence from live streaming commerce
advertising evaluations. J. Market. Res. 14, 538–555. https://doi.org/10.1177/
platforms. Int. J. Inf. Manag. 56, 102251 https://doi.org/10.1016/j.
002224377701400411.
ijinfomgt.2020.102251.
Chen, C.-C., Lin, Y.-C., 2018. What drives live-stream usage intention? The perspectives
Kim, H.S., Kim, M., 2020. Viewing sports online together? Psychological consequences
of flow, entertainment, social interaction, and endorsement. Telematics Inf. 35,
on social live streaming service usage. Sport Manag. Rev. 23, 869–882. https://doi.
293–303. https://doi.org/10.1016/j.tele.2017.12.003.
org/10.1016/j.smr.2019.12.007.
Chen, Hao, Chen, Haitao, Tian, X., 2022. The dual-process model of product information
Kim, J.-H., Kim, M., Park, M., Yoo, J., 2021. How interactivity and vividness influence
and habit in influencing consumers’ purchase intention: the role of live streaming
consumer virtual reality shopping experience: the mediating role of telepresence.
features. Electron. Commer. Res. Appl. 53, 101150 https://doi.org/10.1016/j.
J. Res. Indian Med. 15, 502–525. https://doi.org/10.1108/JRIM-07-2020-0148.
elerap.2022.101150.
Kim, M., Kim, H.-M., 2022. What online game spectators want from their twitch
Chen, J., Liao, J., 2022. Antecedents of viewers’ live streaming watching: a perspective
streamers: flow and well-being perspectives. J. Retailing Consum. Serv. 66, 102951
of social presence theory. Front. Psychol. 13, 839629 https://doi.org/10.3389/
https://doi.org/10.1016/j.jretconser.2022.102951.
fpsyg.2022.839629.
Leung, L., 2020. Exploring the relationship between smartphone activities, flow
Chen, Lu, Y., Hu, Y., Hong, Y., 2019. Everyone can Be a star: quantifying grassroots
experience, and boredom in free time. Comput. Hum. Behav. 103, 130–139. https://
online sellers’ live streaming effects on product sales. In: Presented at the
doi.org/10.1016/j.chb.2019.09.030.
Proceedings of the 52nd Hawaii International Conference on System Sciences, 2019.
Li, Y., Guo, Y., 2021. Virtual gifting and danmaku: what motivates people to interact in
https://hdl.handle.net/10125/59887.
game live streaming? Telematics Inf. 62, 101624 https://doi.org/10.1016/j.
Chen, Y.-H., Chen, M.-C., Keng, C.-J., 2020. Measuring online live streaming of perceived
tele.2021.101624.
servicescape: scale development and validation on behavior outcome. Internet Res.
30, 737–762. https://doi.org/10.1108/INTR-11-2018-0487.

8
S. Zheng et al. Journal of Retailing and Consumer Services 72 (2023) 103240

Li, Y., Li, X., Cai, J., 2021. How attachment affects user stickiness on live streaming Stock, R.M., Hoyer, W.D., 2005. An attitude-behavior model of salespeople’s customer
platforms: a socio-technical approach perspective. J. Retailing Consum. Serv. 60, orientation. J. Acad. Market. Sci. 33, 536–552. https://doi.org/10.1177/
102478 https://doi.org/10.1016/j.jretconser.2021.102478. 0092070305276368.
Li, Y., Peng, Y., 2021. What drives gift-giving intention in live streaming? The Suh, A., Cheung, C.M., Ahuja, M., Wagner, C., 2017. Gamification in the workplace: the
perspectives of emotional attachment and flow experience. Int. J. Hum. Comput. central role of the aesthetic experience. J. Manag. Inf. Syst. 34, 268–305. https://doi.
Interact. 37, 1317–1329. https://doi.org/10.1080/10447318.2021.1885224. org/10.1080/07421222.2017.1297642.
Liao, J., Chen, K., Qi, J., Li, J., Irina, Y.Y., 2022. Creating immersive and parasocial live Sun, Y., Shao, X., Li, X., Guo, Y., Nie, K., 2019. How live streaming influences purchase
shopping experience for viewers: the role of streamers’ interactional communication intentions in social commerce: an IT affordance perspective. Electron. Commer. Res.
style. J. Res. Indian Med. https://doi.org/10.1108/JRIM-04-2021-0114. Appl. 37, 100886 https://doi.org/10.1016/j.elerap.2019.100886.
Lichtlé, M.-C., 2007. The effect of an advertisement’s colour on emotions evoked by Till, B.D., Busler, M., 2000. The match-up hypothesis: physical attractiveness, expertise,
attitude towards the ad: the moderating role of the optimal stimulation level. Int. J. and the role of fit on brand attitude, purchase intent and brand beliefs. J. Advert. 29,
Advert. 26, 37–62. https://doi.org/10.1080/02650487.2007.11072995. 1–13. https://doi.org/10.1080/00913367.2000.10673613.
Liu, H., Tan, K.H., Pawar, K., 2022. Predicting viewer gifting behavior in sports live Tong, X., Chen, Y., Zhou, S., Yang, S., 2022. How background visual complexity
streaming platforms: the impact of viewer perception and satisfaction. J. Bus. Res. influences purchase intention in live streaming: the mediating role of emotion and
144, 599–613. https://doi.org/10.1016/j.jbusres.2022.02.045. the moderating role of gender. J. Retailing Consum. Serv. 67, 103031 https://doi.
Liu, L.-L., 2020. Linear model predictive control for physical attractiveness and risk: org/10.1016/j.jretconser.2022.103031.
application of cosmetic medicine service. Mathematics 8, 975. https://doi.org/ Wang, S., 2020. Chinese affective platform economies: dating, live streaming, and
10.3390/math8060975. performative labor on Blued. Media. Cult. Soc. 42, 502–520. https://doi.org/
Liu, Y., 2003. Developing a scale to measure the interactivity of websites. J. Advert. Res. 10.1177/0163443719867283.
43, 207–216. https://doi.org/10.2501/JAR-43-2-207-216. Wohn, D.Y., Freeman, G., McLaughlin, C., 2018. Explaining viewers’ emotional,
Liu, Y., Shrum, L.J., 2002. What is interactivity and is it always such a good thing? instrumental, and financial support provision for live streamers. In: Proceedings of
Implications of definition, person, and situation for the influence of interactivity on the 2018 CHI Conference on Human Factors in Computing Systems, pp. 1–13.
advertising effectiveness. J. Advert. 31, 53–64. https://doi.org/10.1080/ https://doi.org/10.1145/3173574.3174048.
00913367.2002.10673685. Wongkitrungrueng, A., Assarut, N., 2020. The role of live streaming in building
Lu, B., Chen, Z., 2021. Live streaming commerce and consumers’ purchase intention: an consumer trust and engagement with social commerce sellers. J. Bus. Res. 117,
uncertainty reduction perspective. Inf. Manag. 58, 103509 https://doi.org/10.1016/ 543–556. https://doi.org/10.1016/j.jbusres.2018.08.032.
j.im.2021.103509. Woodcock, J., Johnson, M.R., 2019. The Affective Labor and Performance of Live
Ma, Y., 2021. To shop or not: understanding Chinese consumers’ live-stream shopping Streaming on Twitch.Tv, vol. 20. Television & New Media, pp. 813–823. https://doi.
intentions from the perspectives of uses and gratifications, perceived network size, org/10.1177/1527476419851077.
perceptions of digital celebrities, and shopping orientations. Telematics Inf. 59, Wu, G., Wu, G., 2006. Conceptualizing and measuring the perceived interactivity of
101562 https://doi.org/10.1016/j.tele.2021.101562. websites. J. Curr. Issues Res. Advert. 28, 87–104. https://doi.org/10.1080/
Mahatanankoon, P., 2007. The effects of personality traits and optimum stimulation level 10641734.2006.10505193.
on text-messaging activities and M-commerce intention. Int. J. Electron. Commer. Wu, I.-L., Chiu, M.-L., Chen, K.-W., 2020. Defining the determinants of online impulse
12, 7–30. https://doi.org/10.2753/JEC1086-4415120101. buying through a shopping process of integrating perceived risk, expectation-
Malhotra, N.K., Kim, S.S., Patil, A., 2006. Common method variance in IS research: a confirmation model, and flow theory issues. Int. J. Inf. Manag. 52, 102099 https://
comparison of alternative approaches and a reanalysis of past research. Manag. Sci. doi.org/10.1016/j.ijinfomgt.2020.102099.
52, 1865–1883. https://doi.org/10.1287/mnsc.1060.0597. Yang, H., Lee, H., 2018a. Exploring user acceptance of streaming media devices: an
Mao, Z., Du, Z., Yuan, R., Miao, Q., 2022. Short-term or long-term cooperation between extended perspective of flow theory. Inf. Syst. E-Bus Manage 16, 1–27. https://doi.
retailer and MCN? New launched products sales strategies in live streaming e- org/10.1007/s10257-017-0339-x.
commerce. J. Retailing Consum. Serv. 67, 102996 https://doi.org/10.1016/j. Yang, H., Lee, H., 2018b. Exploring user acceptance of streaming media devices: an
jretconser.2022.102996. extended perspective of flow theory. Inf. Syst. E Bus. Manag. 16, 1–27. https://doi.
McAlister, L., Pessemier, E., 1982. Variety seeking behavior: an interdisciplinary review. org/10.1007/s10257-017-0339-x.
J. Consum. Res. 9, 311–322. 10.1086/208926. Yang, J., Zeng, Y., Liu, X., Li, Z., 2022. Nudging interactive cocreation behaviors in live-
McReynolds, P., 1971. The nature and assessment of intrinsic motivation. Adv. Psychol. streaming travel commerce: the visualization of real-time danmaku. J. Hospit.
Assess. 2, 157–177. Tourism Manag. 52, 184–197. https://doi.org/10.1016/j.jhtm.2022.06.015.
Ming, J., Jianqiu, Z., Bilal, M., Akram, U., Fan, M., 2021. How social presence influences Yu, E., Jung, C., Kim, H., Jung, J., 2018. Impact of viewer engagement on gift-giving in
impulse buying behavior in live streaming commerce? The role of SOR theory. Int. J. live video streaming. Telematics Inf. 35, 1450–1460. https://doi.org/10.1016/j.
Web Inf. Syst. 17, 300–320. https://doi.org/10.1108/IJWIS-02-2021-0012. tele.2018.03.014.
Mutum, D., Ghazali, E., 2011. Perceived online interactivity of blogs. Eur. Adv. Consum. Zhang, M., Liu, Y., Wang, Y., Zhao, L., 2022. How to retain customers: understanding the
Res. 9, 354–359. https://www.acrwebsite.org/volumes/1007071. role of trust in live streaming commerce with a socio-technical perspective. Comput.
Nah, F.F.-H., Eschenbrenner, B., DeWester, D., 2011. Enhancing brand equity through Hum. Behav. 127, 107052 https://doi.org/10.1016/j.chb.2021.107052.
flow and telepresence: a comparison of 2D and 3D virtual worlds. MIS Q. 35, Zhang, N., Li, J., 2022. Effect and mechanisms of state boredom on consumers’
731–747. https://doi.org/10.2307/23042806. livestreaming addiction. Front. Psychol. 13, 826121 https://doi.org/10.3389/
Novak, T.P., Hoffman, D.L., 1997. New metrics for new media: toward the development fpsyg.2022.826121.
of web measurement standards. World Wide Web J. 2, 213–246. Zheng, R., Li, Z., Na, S., 2022. How customer engagement in the live-streaming affects
Obadă, D.R., 2013. Flow theory and online marketing outcomes: a critical literature purchase intention and customer acquisition, E-tailer’s perspective. J. Retailing
review. Procedia Econ. Finance 6, 550–561. https://doi.org/10.1016/S2212-5671 Consum. Serv. 68, 103015 https://doi.org/10.1016/j.jretconser.2022.103015.
(13)00173-1. Zheng, S., Wu, M., Liao, J., 2022. The impact of destination live streaming on viewers’
Ogara, S.O., Koh, C.E., Prybutok, V.R., 2014. Investigating factors affecting social travel intention. Curr. Issues Tourism 1–15. https://doi.org/10.1080/
presence and user satisfaction with mobile instant messaging. Comput. Hum. Behav. 13683500.2022.2117594.
36, 453–459. https://doi.org/10.1016/j.chb.2014.03.064.
Park, H.J., Lin, L.M., 2020. The effects of match-ups on the consumer attitudes toward
internet celebrities and their live streaming contents in the context of product
Further reading
endorsement. J. Retailing Consum. Serv. 52, 101934 https://doi.org/10.1016/j.
jretconser.2019.101934. Avornyo, P., Fang, J., Antwi, C.O., Aboagye, M.O., Boadi, E.A., 2019. Are customers still
Park, J.-Y., Jang, S.S., 2014. Revisit and satiation patterns: are your restaurant customers with us? The influence of optimum stimulation level and IT-specific traits on mobile
satiated? Int. J. Hospit. Manag. 38, 20–29. https://doi.org/10.1016/j. banking discontinuous usage intentions. J. Retailing Consum. Serv. 47, 348–360.
ijhm.2013.12.006. https://doi.org/10.1016/j.jretconser.2019.01.001.
Pelet, J.-É., Ettis, S., Cowart, K., 2017. Optimal experience of flow enhanced by Chaiken, S., 1979. Communicator physical attractiveness and persuasion. J. Pers. Soc.
telepresence: evidence from social media use. Inf. Manag. 54, 115–128. https://doi. Psychol. 37, 1387–1397.
org/10.1016/j.im.2016.05.001. Finneran, C.M., Zhang, P., 2003. A person–artefact–task (PAT) model of flow antecedents
Podsakoff, P.M., MacKenzie, S.B., Lee, J.-Y., Podsakoff, N.P., 2003. Common method in computer-mediated environments. Int. J. Hum. Comput. Stud. 59, 475–496.
biases in behavioral research: a critical review of the literature and recommended https://doi.org/10.1016/S1071-5819(03)00112-5.
remedies. J. Appl. Psychol. 88, 879. https://psycnet.apa.org/buy/2003-08045-010. Heo, J., Kim, Y., Yan, J., 2020. Sustainability of live video streamer’s strategies: live
Raju, P.S., 1980. Optimum stimulation level: its relationship to personality, streaming video platform and audience’s social capital in South Korea. Sustainability
demographics, and exploratory behavior. J. Consum. Res. 7, 272–282. 10.1086/20 12, 1969. https://doi.org/10.3390/su12051969.
8815. Li, B., Hou, F., Guan, Z., Chong, A.Y.-L., 2018. What Drives People to Purchase Virtual
Richard, M.-O., Chebat, J.-C., 2016a. Modeling online consumer behavior: preeminence Gifts in Live Streaming? the Mediating Role of Flow, p. 14.
of emotions and moderating influences of need for cognition and optimal stimulation Michael, J.B., Gilbert, A.C., 1977. The impact of physically attractive models on
level. J. Bus. Res. 69, 541–553. https://doi.org/10.1016/j.jbusres.2015.05.010. advertising evaiuations. J. Market. Res. XIV, 537–555.
Rodríguez-Ardura, I., Meseguer-Artola, A., 2019. Imagine, feel “there”, and flow! Missouri University of Science and Technology, Sheng, H., Joginapelly, T., Missouri
Immersive experiences on M-facebook, and their affective and behavioural effects. University of Science and Technology, 2012. Effects of web atmospheric cues on
ITP 32, 921–947. https://doi.org/10.1108/ITP-10-2017-0358. users’ emotional responses in E-commerce. AIS Trans. Hum.-Comput. Interact. 4,
Ruberg, B., Brewer, J., 2022. Digital Intimacy in Real Time: Live Streaming Gender and 1–24. https://doi.org/10.17705/1thci.00036.
Sexuality, vol. 23. Television & New Media, pp. 443–450. https://doi.org/10.1177/
15274764221084071.

9
S. Zheng et al. Journal of Retailing and Consumer Services 72 (2023) 103240

Nakamura, J., Csikszentmihalyi, M., 2014. The concept of flow. In: Flow and the Skadberg, Y.X., Kimmel, J.R., 2004. Visitors’ flow experience while browsing a web site:
Foundations of Positive Psychology. Springer Netherlands, Dordrecht, pp. 239–263. its measurement, contributing factors and consequences. Comput. Hum. Behav. 20,
https://doi.org/10.1007/978-94-017-9088-8_16. 403–422. https://doi.org/10.1016/S0747-5632(03)00050-5.
Richard, M.-O., Chebat, J.-C., 2016b. Modeling online consumer behavior: preeminence Xu, X., Wu, J.-H., Chang, Y.-T., Li, Q., 2019. The investigation of hedonic consumption,
of emotions and moderating influences of need for cognition and optimal stimulation impulsive consumption and social sharing in E-commerce live-streaming videos. In:
level. J. Bus. Res. 69, 541–553. https://doi.org/10.1016/j.jbusres.2015.05.010. Pacific Asia Conference on Information Systems. https://aisel.aisnet.org/pacis2019/
43.

10