As the explosive development of short video constantly catapulting to become a kingpin in web business, there has been a growing attention in customer psychology and behavior in the scenario of short video. The relevant researches addressed the questions concerning publication marketing (Xie, 2014; Chen, 2019), tourism destination marketing (Li J. Y. et al., 2019; Liu and Gu, 2019; Sun and Zhang, 2019), sporting event marketing (Zhou et al., 2018), film marketing (Tian, 2019), etc. Although most of the existing studies explored short video transmission mechanism (Zhou et al., 2018; Liu and Gu, 2019; Sun and Zhang, 2019), business model (Wang and Sun, 2018; Dong and Bu, 2019; Wang and Guo, 2019), and opportunities and dilemmas brought by short video (Chen, 2019; Tian, 2019), few of them focused on consumer decision-making problems at the micro level of consumer psychology. Studies found that at the cognitive level, the perceived value in short video would positively affect purchase intention, with users' attitude and participation displaying a significant mediating effect (Wang and Huang, 2019). Comparatively, other studies in terms of experience and emotion revealed a significant positive effect of brand experience in short video apps on user loyalty and brand resonance. Also, brand resonance plays a partial mediating role in the impact of behavioral experience and a complete mediating role in intellectual experience on user loyalty (Jia and Hu, 2019). Moreover, both interactive and creative behaviors in short video are found to have a significant positive effect on perceived value and loyalty (Dai and Gu, 2017). In addition, user attitude toward payment is verified to be positively affected by the visual perception created by the content characteristics and emotional perception by the frequency and quality of interaction in short video (Li P. F. et al., 2019). Findings also showed that virtual haptics and pleasure experienced in short video have a significant positive effect on consumers' purchase intention (Guo et al., 2019). In spite of the limited literature in the short history of short video marketing, recent researches on the consumption decision-making mechanism in short video scenario pointed to one of the major characteristics of consumption behavior as having more “human” factors such as “experience,” “resonance,” and “interaction” during the purchase journey in short video than in traditional graphic web page.

A recently proposed theory of customer inspiration is the integration of traditional inspiration theory and consumer behavior theory in psychology, which contributes to the study of human factors on state transition from seeking and willing to accept new ideas (in experience, resonance, interaction, etc.) to changing consumption habits (e.g., purchase behavior). Inspiration is conceptualized in classic psychology as a unique state of intrinsic motivation characterized by evocation, transcendence, and motivation in social psychology. Evocation refers to the inspiration triggered by external factors than by the recipient. Transcendence, however, describes recipient's positive sense of self-enhancement. Then, motivation happens on the moment when inspiration induces a state that is close to motivation, where the recipient feels compelled to actualize new ideas (Thrash and Elliot, 2003). Based on the theory of “inspiration” and practice in marketing, Böttger et al. (2017) (Sheng and Zhao, 2018) first introduced the concept of “customer inspiration” that is defined as a temporary motivational state that facilitates a transition from the acceptance of marketing-induced ideas to the pursuit of consumption-related goals. Furthermore, Böttger et al. illustrated a two-states customer inspiration scale that consists of “inspired-by” and “inspired-to” states. “Inspired-by” state is related to customers' reception of marketing-induced concepts (i.e., evocation) and awareness of new possibilities (i.e., transcendence). On the other hand, “inspired-to” state is connected with consumers' intrinsic pursuit of relevant goals (i.e., motivation). In this case, the customers are eager to implement new ideas such as buying or using a new product rather than simply keep the idea in their minds. In other words, the customers will undergo a transition from “being inspired by the idea” to “being inspired to purchase the product” when they are converting acquired purchase inspiration into specific behaviors (Hoffman and Novak, 1996; Thrash and Elliot, 2003). In this study, ‘short video customer inspiration' is the specific manifestation of customer inspiration in short video context.

The concept of customer inspiration is proposed to connect the reception of activated new ideas to the pursuit of consumption-related goals, and is conducive to further research on the effect of short video on online marketing. Specifically, how to stimulate customers' instant impulsive purchasing behavior while forcefully locking users' attention in the current ever-shortening customer journey. Customer inspiration is usually characterized as a mediating role to facilitate transition from extrinsic marketing stimuli to customer intention and behavior. It has been found by many scholars that customer inspiration has a mediating effect between variables such as perceived product innovation, product localization, vitality of ads, societal compact, and purchase intention (Dong, 2020; Sheng et al., 2020; Tang and Tsang, 2020). In addition, customer inspiration plays a significant impact on customer behavior; for example, it can mediate inspired content and engaged behaviors (Izogo and Mpinganjira, 2020).

Besides, many studies have found evidence that inspiration may lead to customer impulse buying. For example, a large-scale offline questionnaire survey done by Böttger et al. (2017) revealed that the product descriptions with high inspirational content often trigger consumer purchase impulse, while the analysis by Thrash and Elliot (2003) from the perspective of behavior indicated that inspiration elicits intrinsic motivation to the implementation of a new idea that will determine the ultimate behavior, namely, the “inspire by” state. Moreover, the study by Chen et al. (2009) proved that the flow experience components of inspiration will increase the number of unplanned purchases. In general, customers obtain new ideas through the product information cues offered by the marketers. In this process, customers may experience transcendence toward a new mental state that may stimulate an impulse to pursue a relevant goal which is quickly transformed into a purchasing behavior.

Applied in the field of marketing, inspiration also supplements some of its features to customer inspiration, such as the evocation of inspiration that is commonly triggered by an external information other than customers' personal experience, common sense, or self-induced need (Dong, 2020). Therefore, customer inspiration in short videos is actually incentivized by the content in videos. On the other hand, some scholars indicated that a new model that immerses consumers in situations (such as exhibition of product operation, visual search at Pinterest, etc.) can help create rich marketing content that might inspire customers, and thus is conceived as important source of customer inspiration (Böttger et al., 2017). Since immersiveness in situation is the manifestation of the concept of “presence” which describes a remote perceptual experience in virtual or real environment (Shen, 2018), presence may be an important source of inspiration for online customers.

Although no direct literature has been found about the relationship between presence and customer inspiration in virtual shopping scenario, other relative researches can be cited as credence to the implication of the study. On the one hand, some studies have confirmed that customer inspiration includes components of flow experience. For example, Schouten et al. (2007) defined flow experience as feelings acquired by total absorption in an activity, with features such as high concentration, loss of self-awareness, and intrinsic pleasure. They further pointed out that the transcendent customer experience is characterized by emotional intensity, epiphany, singularity, and newness of experience, etc. Therefore, the flow experience can be regarded as a manifestation of inspiration. On the other hand, relevant studies have found that presence in the virtual environment will generate flow experience. The stronger the physical presence and social presence generated by customers, the more likely the customer is to receive the perceptual stimulation in the virtual scene, and the more possible the flow experience is induced (Novak et al., 2000; Schouten et al., 2007; Animesh et al., 2011). In addition, Kim et al.'s (2013) work indicated that a higher degree of co-presence will result in a sense by consumers that they are psychologically closer to their shopping partners, their communication is warmer and friendlier; therefore, their social needs and hedonic value will be satisfied. A high level of hedonic value will deepen the sense of immersion and state of inspiration in terms of concentration. In a study on the relationship between inspirational flow experience and consumption purchase, Chen et al. (2009) found that the consumers' remote perception (presence) affects their flow experience in a significant way. Similarly, an empirical research conducted by Dai and Liu (2015) showed that the social presence of WeChat users has a significant positive impact on flow experience. In summary, short video can effectively inspire the customers by creating a rich marketing content. Specifically, by a higher level of presence, the consumers are more likely to be engaged in the virtual scenario and have a real life-like experience by staying and communicating with network anchors. Consequently, there is greater possibility that consumers will be immersed in short videos and acquire customer inspiration that can ultimately promote the impulse purchase. Accordingly, we propose Hypotheses H2a–H2c as follows:

H2a: Customer inspiration plays a mediating role between social presence and impulse purchase intention in short video scenario.

H2b: Customer inspiration plays a mediating role between co-presence and impulse purchase intention in short video scenario.

H2c: Customer inspiration plays a mediating role between physical presence and impulse purchase intention in short video scenario.

In accordance with the formulated hypotheses, a conceptual model is constructed in this study to show the relationship among presence, short video customer inspiration, and impulse purchase intention (Figure 1).

This study presents the results of a quantitative analysis of the data collected in a questionnaire survey. The first question is “Have you ever bought something after viewing a short video?” If the response is “No,” the respondent can stop his/her survey; but if it is “Yes,” the follow-up questions will be answered. In the introduction of the questionnaire, the respondent is prompted to “recall an impressive short video ads and answer the questions”; and at the end of the questionnaire, the respondent needs to answer whether he/she “has bought the product promoted in the short video.” Through this part of survey, experienced short video consumers can be selected as the research subject in this study. Before the formal investigation, a pilot study was conducted among the consumers with experience of short video ads in six cities in China. The questionnaire was revised in response to the feedback to ensure the content validity of the scale. Then a formal structured questionnaire was designed on WJX, a professional platform working on questionnaire design and data collection. Thanks to the assistance of marketing research association of Chinese universities, the questionnaire data were collected by the council members of the association in Beijing, Shanghai, Guangdong, Shaanxi, Sichuan, Jiangsu, and other provinces in China. The survey lasted for 1 month and collected a total of 420 questionnaires, of which 41 questionnaires with inconsistent answers or those finished in <2 min were taken as invalid and thus eliminated, and 379 questionnaires were found valid for the final analysis, with an effective recovery rate of 90%.

The demographic characteristics of the subjects are listed in Table 2. Age characteristics show that the young consumers are in the majority; those between 18 and 25 years of age are accounting for the highest proportion (63%). The gender characteristics reflect almost equal proportion between male and female, in spite of a slightly higher proportion of female by 2.12%. As to education characteristics, the majority of consumers have higher education qualifications, in which bachelor's degree makes up the highest percentage (57%). In addition, in the characteristics of online shopping times, the largest ratio goes to those whose online shopping times are <6. Finally, the characteristics of monthly income reveal that the low-income group is in large number; those with an income of <3,000 yuan are accounting for the highest proportion (56%).

To ensure the reliability and validity of measurement, this study employed the scale derived from the relevant concept of authority and 7-point Likert scale to measure the variables involved. First, to fully depict the concept of presence, we adapted phrasing in the preexisting scales from Klein, Bulu, Hassanein, and Head (Vijayasarathy, 2002; Klein, 2003; Hassanein and Head, 2007). Social presence, co-presence, and physical presence include three items, individually. Second, customer inspiration, in accordance with two-state 10-item scale of customer inspiration by Böttger et al. (2017) can be categorized into the following two dimensions: Cognitive activation and customer intention, namely, the dimension of “inspired-by” and of “inspired-to,” each of which contains five items. Third, impulsive purchase intention was measured by a rephrased scale derived from the research of Vijayasarathy (2002); three items involved in the measurement. Furthermore, all translated scales in the questionnaire were back-translated to validate the accuracy of phrasing in each item.

In this study, the reliability of the variables was tested by Cronbach's α coefficient with a threshold of >0.7. The corrected item-to-total correlation coefficients (CITC) were calculated between each item and global item score with statistical significance set at >0.5, as suggested by Churchill. The reliability values of each variable are listed in Table 3.

It can be seen from the reliability analysis that the Cronbach′s α coefficients of all variables in the model are >0.8, and the CITC are >0.7, fully satisfying the criteria that Cronbach′s α is >0.7 and CITC is >0.5, suggesting that the reliability of the questionnaire and the internal structure consistency of variables are high.

Validity analysis is to test the rationality and validity of the questionnaire. The high validity level indicates that the measurement results are in good agreement with the measurement targets. In this study, the rationality and validity of the questionnaire were tested through construct validity (including convergent validity and discriminant validity) that refers to the extent of consistency between the theory and the data. Convergent validity is to assess whether the data of items in one dimension point to the same variable and correlations between different items are high. The discriminant validity refers to the assumption that the items having no correlation with the variables do have no correlation with the variables. Furthermore, the factor analysis was conducted in this study to evaluate the construct validity of the questionnaire.

The Kaiser–Meyer–Olkin (KMO) measure of variables and the Barlett test of sphericity are required before the factor analysis. The KMO statistic is used to test the partial correlation between the variables. Ranging between 0 and 1, a KMO value >0.9 is considered adequate for the factor analysis. On the other hand, Bartlett test of sphericity is used to assess the independence between the variables. The significance of probability (p) is used to judge the adequacy for factor analysis. The results of KMO measure and Bartlett test of sphericity of each variable are listed in Table 4.

It can be seen that the KMO values of all variables are >0.7, and the probability p-values of χ² in Bartlett test of sphericity achieve the criteria of significance, indicating that the factor analysis can be performed.

In this study, confirmatory factor analysis (CFA) was used to test the convergent validity of the model. The results of the measurement model are listed in Table 5. The standard deviation (STD) scores of all latent variables are >0.7; the composite reliability (CR) scores are higher than 0.8; and the scores of average variance extracted (AVE) are >0.5. Therefore, the reported results fulfill the criteria of convergent validity (STD > 0.5; CR > 0.6; AVE > 0.5); thus, suggesting a good convergence validity of the tested model.

In this study, the discriminant validity was tested using the method of AVE, in which when the square root of AVE (AVE) obtained from the variables is greater than the correlation coefficient between the variables, the scale has good discriminant validity. It can be seen from Table 6 that the correlation coefficients between the variables are <0.5 while AVE of each variable is greater than the correlation coefficient of other variables, which meets the requirements of the discriminant validity test, indicating a good discriminant validity of the scale.

We first tested whether videos A1 and A2 successfully stimulate social presence of the subjects. It is shown by a one-way analysis of variance (ANOVA) that social presence triggered by video A1 (M_{highsocialpresence} = 4.65, SD = 1.31) is significantly higher than that by video A2 [M_{lowsocialpresence} = 3.45, SD = 1.57; F_{(1, 148)} = 26.410, p = 0.000]. The significant difference indicates a success of manipulation on social presence.

The ANOVA results show a significant higher impulse purchase intention in the group with high social presence (M_{highsocialpresence} = 4.72, SD = 1.57) than that with low social presence [M_{lowsocialpresence} = 3.77, SD = 1.64; F_{(1, 148)} = 13.264, p = 0.000] as shown in Figure 4, indicating a significant main effect of social presence on impulse purchase intention, thus Hypothesis H1a being verified.

A one-way ANOVA with customer inspiration as dependent variable indicates that customer inspiration of the high social presence group (M_{highsocialpresence} = 4.83, SD = 1.29) is significantly higher than that of the low social presence group [M_{lowsocialpresence} = 4.02, SD = 1.45; F_{(1, 148)} = 13.057, p = 0.000]. Then the mediating effect of customer inspiration was tested by Bootstrapping. In specific, Bootstrap procedure was performed with 5,000 samples on model 4 in which social presence was assessed as independent variable, impulse purchase intention as dependent variable, and the customer inspiration as mediating variable. The results show that the mediating effect of customer inspiration is significant (0.6428) when the 95% confidence interval does not contain 0 (LLCI = 0.4925, ULCI = 0.8020), which verifies Hypothesis H2a that the customer inspiration plays a mediating role in the impact of social presence on impulse purchase.

We firstly tested whether videos B1 and B2 successfully inspire the subjects' co-presence. A one-way ANOVA shows that the level of co-presence stimulated by video B1 (M_{highco−presence} = 5.23, SD = 1.01) is significantly higher than that by video B2 [M_{lowco−presence} = 3.40, SD = 1.44; F_{(1, 148)} = 49.85, p = 0.000], indicating that the manipulation on co-presence was successful.

ANOVA indicates that the impulse purchase intention (M_{highco−presence} = 4.52, SD = 1.48) in high co-presence group is significantly higher than that in the low co-presence group (M_{lowco−presence} = 3.79, SD = 1.54; F_{(1, 148)} = 8.71, p = 0.004). As shown in Figure 7, thus, the main effect of co-presence on impulse buying intention is significant, and Hypothesis H1b verified.

First, one-way ANOVA was performed with customer inspiration as the dependent variable. The result shows that the customer inspiration (M_{highco−presence} = 4.69, SD = 1.12) in the high co-presence group is significantly higher than that in the low co-presence group [M_{lowco−presence} = 4.02, SD = 1.38; F_{(1, 148)} = 10.72, p = 0.001]. The mediating effect of customer inspiration was then tested by Bootstrapping with 5,000 samples on model 4 considering co-presence as independent variable, impulse purchase intention as dependent variable, and customer inspiration as mediating variable. The result indicates that the effect of mediation was significant (0.6126) if the 95% confidence interval does not contain 0 (LLCI = 0.4446, ULCI = 0.7884), which again verifies Hypothesis H2b, namely, customer inspiration acts a mediating role in the impact of co-presence on impulse purchase.

Likewise, stimulation test on physical presence was conducted on the subjects after they had watched videos C1 and C2. A one-way ANOVA shows that the physical presence inspired by video C1 (M_{highphysicalpresence} = 4.71, SD = 1.24) was significantly higher than that by video C2 [M_{lowphysicalpresence} = 4.12, SD = 1.27; F_{(1, 158)} = 8.78, p = 0.004]. Such a significant difference indicates a success of manipulation on physical presence.

The ANOVA test shows a higher intention of impulse purchase in the high physical presence group (M_{highphysicalpresence} = 4.29, SD = 1.66) than in the low physical presence group [M_{lowphysicalpresence} = 3.60, SD = 1.67; F_{(1, 158)} = 6.82, p = 0.010]. As shown in Figure 10, therefore, the main effect of physical presence is significant and Hypothesis H1c verified.

First, one-way ANOVA was conducted with customer inspiration as the dependent variable. It is shown that customer inspiration of the high physical presence group (M_{highphysicalpresence} = 4.53, SD = 1.40) is significantly higher than that of the low physical presence group [M_{lowphysicalpresence} = 3.91, SD = 1.52; F_{(1, 158)} = 7.04, p = 0.009]. Second, the mediating effect of customer inspiration was analyzed by Bootstrapping with 5,000 samples on model 4, taking physical presence as independent variable and impulse purchase intention as dependent variable. The result confirms a significant mediating effect (0.8653) when the 95% confidence interval does not contain 0 (LLCI = 0.7348, ULCI = 1.005), which verifies Hypothesis H2c that customer inspiration plays a mediating role in the impact of physical presence on impulse buying.

Since social presence, co-presence and physical presence produce significant positive impacts on impulsive purchase intention, as indicated by the results, consumers who are watching short videos are suggested to raise their purchase rationality and make reasonable decision in line with their actual needs and purchasing power, which is especially true when there is such high interactivity between short video anchor and audiences as conversational or responsive communication, explosion of likes or comments, and other similar situations.

It is suggested to increase presence of consumers by enhancing their viewing experience with smooth pictures and lifelike features, so that the time for decision-making can be effectively reduced and the purchase conversion rate thus improved. More precisely, the enhancement of physical, social, and co-presence in short video will elicit customer inspiration that can be converted into buying behavior over a short period of time. Therefore, technologies such as virtual and augmented reality (Rauschnabel et al., 2019), digital signage, and online tools (Böttger et al., 2017) could be used in the production of short videos to create quasi-reality experience for consumers and thus increase physical presence. Similarly, social elements could be inserted to short videos through immersive visual content and interactive expressions to increase social and co-presence that can inform customers with better knowledge about the product, enrich their imagination on its practical uses, stimulate customer inspiration, and facilitate actual conversion from an idea to purchasing behavior.