With the advent of social media, access to health information has increased, with more than 55% of adults using social media to seek health information and advice [1]. Vranica and Kruppa [2] found that about 23% of social media users used short video platforms, specifically TikTok. TikTok handles a high search activity, which reached 1.2 million search requests per minute in 2024 [3]. Liu et al [4] explained that the short video-based social media format is most popular for users conducting health information searches because short videos can help convey complex health information in a short time with a combination of visual, audio, and narrative that is easy to understand.

Short video platforms can present information in various forms of media, such as text (captions), images or photos, audio (podcasts), and videos, both long-duration videos and short-duration videos. This multimedia format allows users to see visualizations of the disease they are experiencing so that it is easier to recognize early symptoms and compare their condition with visual examples in videos or photos for reference [5]. The format of information media in the search for health information can be explained through media richness theory (MRT). MRT explains that “richer” media have an effective ability to explain ambiguous issues [6]. Short video platforms can combine visual, audio, text, and live demonstration elements in a short duration. This allows complex health information to be conveyed well and improves audience understanding quickly. In addition to the media aspect, several other factors affect the behavior of searching for health information on short video platforms, such as personalization and interactivity.

Han et al [7] found that TikTok’s algorithm forms an information space that facilitates the discovery of relevant content that encourages exposure and further searches on the platform. This algorithm serves relevant content based on interactions, such as likes and watch time, so that users can find health content without conducting an explicit search [7]. The behavior of seeking health information on short video platforms can also be understood from a psychological perspective, one of which is through the theory of health belief model (HBM). HBM explains that an individual’s decision to engage in the search for health information is influenced by perceived susceptibility (PSU) and perceived severity (PSE) to a health condition [8]. The higher a person’s risk perception, the greater their tendency to seek information through digital platforms, including social media [9,10]. Priyambodo et al [11] showed that users tend to seek explanations when they are worried about the potential risks they could face. This pattern is visible in the realm of skin health, which is visual and easy to observe. Concerns about acne, dark spots, eczema, or other changes in skin conditions make short video platforms the main choice because they are able to present visual demonstrations, real examples, and personal experiences in person [12,13].

As of 2025, Indonesia has a population of 284.4 million people, with internet penetration reaching 80.66% [14]. The most widely used short video platforms in the country are TikTok, Instagram Reels, and YouTube Shorts [14]. With this high level of social media penetration, Thang et al [15] found that many users accessed skin health information through short video platforms. Short video platforms are a popular source used to search for or share dermatological or skincare information, with high average views per video of 1.6 million views on TikTok alone [15]. Thus, social media plays a major role in the beauty discovery phenomenon. Beauty discovery is the process of searching for information or definitions related to beauty, such as information about products, personal care techniques, and beauty standards.

Although they have many uses, short video platforms also have the risk of presenting disinformation. According to Zhang et al [16], the dissemination of information by lay users can lead to widespread misinformation on social media. Other research has shown that the quality of health videos on these platforms is often low. For example, Ming et al [17] found that the proportion of misinformation in certain disease videos on TikTok could reach 41%‐78% of the total content. In their study, Kirkpatrick and Lawrie [18] found that almost all respondents believed that health misinformation was common on TikTok. The absence of special quality evaluation standards for health information on short video platforms is another obstacle to finding health information on short video platforms. The risk of adopting information without further information verification is further exacerbated by the dominance of beauty content in Indonesia, where the representation of bright and smooth skin is a visual standard widely reproduced by content creators [19]. Ariana et al [20] showed that using TikTok significantly affected users’ body image satisfaction. This phenomenon further emphasizes the strong influence of social media on shaping public perception of the definition of an “ideal” body and skin, especially among young audiences who are active in consuming short video content.

Several studies have been conducted on the quality of skin health information content shown on short video platforms. Thang et al [15] discussed acne management information on various short video-based social media platforms and showed that the quality of education varies and does not always correlate with popularity. Khan et al [21] analyzed 120 videos on TikTok related to eczema and found that 65.8% of the content was produced by individuals who did not have a professional background in the health field, while only 15.8% of the content was produced by individuals with a health background. Thang et al [15] found that the quality of content produced by certified dermatologists was significantly higher than that of individual content that did not have a health background. However, videos of better quality made by doctors actually received lower engagement [15]. In addition, several studies have shown misperceptions of information circulating on short video platforms, especially regarding drugs and skin health procedures. Khan et al [21] found many misperceptions about the use of tretinoin, especially exaggerated claims regarding its effectiveness and recommendations for use without medical supervision. A similar pattern can be observed in the content related to derma rolling procedures. For example, Awad et al [22] found that most videos provided recommendations that were not aligned with clinical practice and tended to promise instant results. In general, the most viral videos were those that were demonstrative, emotional, and visually appealing but were not based on strong scientific evidence.

Previous studies have examined the behavior of searching for health information on social media in general [9,11,15,23,24]. However, previous studies have only discussed content quality with regard to the credibility of actual information [21,22,25]. Further research exploring the user behavior of social media platforms with short video formats is needed because skin problems are often related to aesthetic factors, social stigma, and comparison of appearance [26]. Thus, the research question of this study is, “What factors affect the use of short video platforms in searching for skin health information?” The findings of this study can be used by the government, academics, and educational institutions to develop digital literacy and public health programs.

This study was approved by the Faculty of Computer Science, University of Indonesia (reference S-20A/UN2.F11.D1.5/PPM.00.00/2025 and dated June 3, 2025). All of the respondents involved in this study agreed in writing to participate in the study, and all respondent data were anonymized. The questionnaire was filled in by the respondents voluntarily, and they could withdraw from participating in the research at any time. We did not provide incentives for the respondents. We used the survey tools of the University of Indonesia and stored the data exclusively on the university’s server. All respondents approved the written consent to be involved in this study.

This study adopted a mixed methods approach, using an online questionnaire (quantitative) and interviews (qualitative). We designed an online and open questionnaire based on the CHERRIES (Checklist for Reporting Results of Internet E-Surveys) checklist (Checklist 1). All respondents approved the written consent to be involved in this study. The respondents were selected according to the criteria of the research target—that is, Indonesian people who used short video platforms to search for skin health information (purposive sampling). The questionnaire link was shared for 28 days from September 22, 2025, until October 20, 2025, through social media platforms widely used by Indonesian people, such as Facebook, Instagram, and TikTok. In the early stages, quantitative research was used as the primary method to test the hypotheses and the relationships between the variables in the research models. Furthermore, a qualitative approach was carried out to explain and enrich the understanding of the quantitative results through interviews with 30 respondents.

The research instrument was in the form of a questionnaire, which was divided into 3 pages: respondent validity, respondent demographics, and 49 measurement items (Multimedia Appendix 1). The questionnaire completion time was 10‐15 minutes, and the respondents could review their answers by clicking the back button for each page. We did not randomize the questionnaire’s statements and did not use adaptive questions. For mandatory questions and statements in the questionnaire, we put a red star where the respondents must fill in the questions and statements before they could click the submit button. The respondent validity section consists of questions regarding the knowledge and experience of using one of the short video platforms used by the user, that is—TikTok, Instagram Reels, or YouTube Shorts. In addition, the respondent demographics section consists of questions related to the background of the study respondents. The measurement items were measured using a 5-point Likert scale, in which a value of 1 was interpreted as “strongly disagree,” and a value of 5 was interpreted as “strongly agree.”

The readability test was conducted to evaluate the ease of the questionnaire questions and to identify aspects that needed improvement from the respondents’ perspectives. The readability test was carried out in the form of interviews with 13 respondents. The respondents provided input on the right word selection and the correction of word or punctuation errors. After the readability test stage was completed, the pilot study stage was conducted with 30 respondents to test the reliability of the research indicators that had been made previously. The total Cronbach α (CA) score for the results of the pilot study was 0.928; thus, the questionnaire could be disseminated to a wider range of respondents. Multimedia Appendix 2 presents the demographics of the 603 valid respondents.

Of the 889-questionnaire data received, we conducted a manual data completeness check using Microsoft Excel to ensure that all questions had been answered completely. An examination of the results showed that 245 respondents did not complete the questionnaire until the end; thus, the data from these respondents were unable to be used for hypothesis testing. The remaining 644 questionnaires were completely filled out for hypothesis testing.

Questionnaire data were processed using the covariance-based structural equation modeling method using AMOS SPSS (version 26; IBM Corp) software, which aims to test the relationships between the variables formulated in the research model. This method was chosen because the study focused on testing hypotheses based on theories that had been tested in previous research. Qualitative data analysis was conducted using thematic analysis. Thematic analysis was chosen because it provides flexibility in identifying, analyzing, and interpreting patterns or themes that systematically emerge from qualitative data [64]. The social, cultural, and psychological contexts behind user behavior can also be found using thematic analysis [64].

To detect potential bias, one of the approaches carried out was the confirmatory factor analysis (CFA) single-factor test, which determines one common latent factor that contains all the research indicators [65]. This approach is conducted by estimating the CFA model by pointing all indicators to one common factor as a form of testing whether one common factor can explain most of the variance of the data [65]. Common method bias (CMB) testing using single-factor CFA can be performed by analyzing the feasibility of the model [66]. If the single-factor model does not have a good goodness of fit (GoF), then CMB is not identified in the research model, and the research can continue to the next stage [66]. Table 1 describes the CMB results for measurement and structural model testing.

The convergent validity test was carried out by analyzing the average extracted variance (AVE) value, in which the minimum AVE value for each variable tested must be above 0.5 (Hair et al, 2019). As shown in Table 2, all variables met the AVE value ≥0.5. Therefore, the convergent validity test stage was completed, and the study continued to the next stage. The reliability of the construct was measured using CA and composite reliability. Acceptable CA and composite reliability values are greater than 0.70 [67].

The discriminant validity test was performed to ensure that each construct measured in the research model had a significant difference from the other constructs [67]. This stage was conducted to ensure that a construct not only measures the same concept as other constructs but also has unique characteristics so that it has to be separated as a construct in the research model [67]. Table 3 presents the results of the discrimination validity test.

GoF testing was performed to evaluate the feasibility of the research model by determining whether the GoF values met the specified criteria [67]. The evaluation was carried out by evaluating several GoF values, such as the relative chi-square, GoF index, root mean square residual, normal fit index, comparative fit index, Tucker-Lewis index, and root mean square error of approximation [67]. Table 4 presents the GoF values of the research model.

Hypothesis testing was performed using a 1-tailed test to assess the significance of the evaluated hypothesis coefficient. A hypothesis was declared acceptable if the P value <.05, while the hypothesis was rejected if the P value was ≥.05 [67]. The results of the hypothesis test are summarized in Table 5, which shows that out of the 10 hypotheses submitted, 8 hypotheses were accepted, and 2 hypotheses were rejected because they had a P value ≥.05.

Based on the results of the effect size analysis, it can be seen that each construct in the model has a different influence power. The attitude variable showed an effect size of 0.569, which was in the range of 0.50‐0.79 so it was included in the category of medium effect. These findings show that PI has a strong and substantive influence in shaping users’ attitudes. Furthermore, the SST variable has an effect size of 0.392, which is in the category of small effects, so although USC has an effect on increasing skin stigma, the power of influence is relatively weaker than other constructs. The PU variable showed an effect size of 0.562, so it was also included in the medium effect category. This indicates that the combination of the influence of HCE and PHIs makes a significant contribution in increasing the perception of the usefulness of skin health content. Meanwhile, the variable SHE has an effect size of 0.579, which is also in the medium category. This shows that related variables in the research model, such as PU, attitude, PSE, and PSU, have a strong contribution in encouraging skin health information–seeking behavior on short video platforms.

The study found that HCE and PHIs affected users’ PU of short video platforms in searching for skin health information, thus affecting SHEs. SC had no effect on SHEs because the functionality of the platform and familiarity with the platform were the main points chosen by users. PI had an effect on attitudes, thus affecting SHEs. In the context of short video platforms, interactivity is one of the factors that play a significant role because it is considered to increase personalization, which was previously found to have a positive influence on users’ feelings on a platform. USC also had an effect on SST, which had no effect on SHEs. Finally, PSE and PSU affected SHEs.

All qualitative results in each hypothesis strengthen the quantitative results and provide the detailed reason of each hypothesis results according to the Indonesian context. Table 6 describes the mapping summary of the hypothesis testing results with the qualitative results. The unique findings of this study were on the results of SC and SST that has no effect on SHEs. SC only affected an individual’s desire to conduct further searches to validate the information they have just watched. Trusting a figure was also conducted one by one when the content was being watched and not before doing a search and through the dynamics of interactions in the platform ecosystem. Furthermore, in the context of users of short video platforms in Indonesia, SST is not the main factor that drives the behavior of seeking skin health information, especially on short video platforms. The next section will discuss further justification of each hypothesis.

The study found that HCE significantly affected PU. These findings are consistent with Yu et al [23], Armstrong et al [68], and Lenczowski et al [33]. Yu et al [23] found that the use of media elements, such as visual, audio, and text, could enrich the user experience and improve the understanding of complex health information. Armstrong et al [68] revealed that skin health education through online videos was more effective than written media because videos allow for live demonstrations. Lenczowski et al [33] showed that their respondents preferred skin health education through video rather than written materials, confirming that expressive media improves the perception of the usefulness of information.

The findings of the qualitative research also support the results of the quantitative analysis (hypothesis 1). In total, 9 interviewees (9/30, 30%) stated that the expressiveness of content on short video platforms helped them to see first-hand evidence when searching for skin health information. Interviewee 15 said:

In addition, 13 interviewees (13/30, 43.33%) felt that the expressiveness of content on short video platforms made it easier for them to capture information. Interviewee 9 stated:

A total of 7 interviewees (7/30, 23.33%) also found this helpful because the expressiveness of short video content on short video platforms can show how the product is used. As Interviewee 22 explained:

The expressiveness of the content on the short video platform provides visual evidence, makes the content easier to understand, clarifies how to use the product, and presents actual pictures of skin conditions, thus making skin health information more useful and relevant for users.

This study found that PHIs had a significant effect on PU. The results of this study are in accordance with Yu et al [23], Zhang et al [34], and Nittas et al [35]. Yu et al [23] found that PHIs are one of the features of media richness that significantly affect PU on short video platforms. Information that is appropriate to the users’ preferences, needs, and context provides the content with more practical value and can effectively address individual health challenges [23]. Zhang et al [34] showed that PHIs in the context of mobile health had a positive and significant influence on PU because they tailored services to the unique conditions of users in China. Nittas et al [35] found that personalization of skin health education, such as the use of sunscreen, was well received by participants in Switzerland and improved their perception of the platform’s usability.

The results of the interviews support the findings from the questionnaire results (hypothesis 2). In total, 9 interviewees (n=30, 30%) stated that personalization helped them obtain skin health information according to their needs. As Interviewee 8 mentioned, “Personalization makes it easier because the content that appears is already relevant to the question or skin condition that is being experienced.” Personalization in the context of skin health on short video platforms is able to display content that is aligned with users’ individual dermatological needs based on search history, duration of viewing, interactions, and user preferences for certain skin types or skin concerns. In addition, 7 interviewees (n=30, 23.33%) felt that personalization helped them obtain more information related to skin health. Interviewee 3 said:

Personalization allows users to receive content relevant to their personal needs, expand insights through additional recommendations, improve search efficiency, and simplify the process of finding skin health information. While some concerns have been raised regarding privacy and information overload, personalization is still perceived as a feature that can improve the usefulness of skin health information on short video platforms.

PU had a significant effect on SHEs. The results of the quantitative analysis are in accordance with Yu et al [23], Liu et al [4], Xia et al [38], and Boyce et al [39]. Yu et al [23] showed that PU is one of the main factors driving the search for health information on short video platforms, as users assess the content provided as useful and can improve their ability to understand health information. Xia et al [38] found that PU had a significant positive influence on users’ desire to search for health information online, even greater than the effect of perceived ease of use. In addition, Boyce et al [39] emphasized that the PU of a platform drives the behavior of seeking health information in online communities because users feel that the platform helps them find appropriate and relevant information.

In addition, we found that the interview results support the questionnaire result (hypothesis 3). Based on the interviews, 3 interviewees (n=30, 10%) considered short video platforms beneficial because they supported discussion and interactivity, which facilitated the process of finding and understanding skin health information. Interviewee 12 stated, “TikTok remains the main platform for me because its features are interactive and make it easier for me to understand information quickly.” Aside from interactivity, the perception of usability was also reflected in the ease of the search process. A total of 9 interviewees (n=30, 30%) stated that short video platforms offer a fast, practical, and easy-to-understand search process. As explained by Interviewee 13, “Because the content has been adjusted to my needs, the information that appears becomes relevant and quickly connects with what I’m looking for.” The benefits of the platform were also seen through the platform’s ability to provide additional information beyond the users’ initial needs. In total, 6 interviewees (n=30, 20%) revealed that short video platforms helped broaden their horizons. As stated by Interviewee 1, “Sometimes I get additional information that I didn’t think of before.” Aside from adding insights, other interviewees emphasized that the platform also provided a variety of information and the latest content that has been curated. Therefore, PU is an important factor that encourages users to search for skin health information on short video platforms through ease of access, information relevance, interactivity, and the platform’s ability to provide additional insights.

SC did not have a significant effect on SHEs. Moreover, the results of this study are not in accordance with those of Fan et al [43] and Kirkpatrick and Lawrie [18]. Fan et al [43] and Kirkpatrick and Lawrie [18] found that assessments of whistleblower credibility can affect the tendency and frequency of user information searches. However, several studies have found that the credibility of information sources did not have a significant effect on the search for health information through online platforms [69,70].

Denniss et al [70] revealed that the credibility of sources was considered less relevant because users judged the content and social interactions more. In the context of the use of short video platforms in the search for skin health information, SC did not affect the search because the characteristics of the social media ecosystem were more content-oriented and algorithmic. In the interviews, the interviewees who were laymen in the context of skin health watched the videos shown by short video platforms but did not specifically search for a figure. Therefore, trusting a figure was carried out one by one when the content was being watched, and this was not done before conducting a search. This is consistent with the demographic distribution of the study, in which 405 respondents (n=644, 67.1%) stated that skincare products were the type of skin health information they were looking for on short video platforms. This type of information is more general; thus, it can be reviewed by anyone without a formal background in the health field. The type of information that anyone can review emphasizes the role of platforms in building algorithmic ecosystems as a key factor compared with the role of individual content creators in the process of searching for skin health information. Trusting a figure was also conducted one by one when the content was being watched and not before doing a search. To date, users also face an increasing number of hoax content, algorithmic contents, and content generated by artificial intelligence [71,72]; thus, making social media users more careful in looking for informational or educational contents especially for “algorithmic truth” hindering. Figures should align their core values by emphasizing quality and environmental sustainability in their messaging that can strengthen users’ trust and emotional connection in the short video platform [73]. Moreover, quantity and quality social proof (ie, likes and comments) could also influence users’ behavior in the short video platform [74].

Moreover, interview results also support the quantitative result (hypothesis 4). We found that 20 interviewees (n=30, 67.1%) revealed that skincare products were the type of skin health information they were looking for on short video platforms. This type of information is more general; thus, it can be reviewed by anyone without a formal background in the health field. In addition, based on the results of the interviews, the validation of skin health information on short video platforms is more often done through social interactivity mechanisms, such as comments, the number of likes, and follow-up responses in the form of replies or stitch content. These interactions serve as a form of collective evaluation that helps users assess whether information is trustworthy. In this context, credibility is not built individually through a specific figure but is done every time one watches a video communally through the dynamics of interactions in the platform ecosystem.

Trusting a figure was also conducted one by one when the content was being watched and not before doing a search. Therefore, several exchange rates on the platform could make users continue to use short video platforms to search for skin health information. A total of 12 interviewees (n=30, 40%) stated that their trust in creators’ content on short video platforms was quite low, but they still conducted searches on short video platforms. According to Interviewee 27:

These findings suggest that the credibility of informants remains important, but perceptions of the level of credibility do not affect an individual’s desire not to conduct searches on short video platforms. Rather, it affects an individual’s desire to conduct further searches to validate the information they have just watched.

PI had a significant influence on attitude. This result is in accordance with Liu et al [4], who found that interactivity is one of the factors that contribute to the flow of application use and increases the desire of users to continue using the application. The finding is also consistent with Haykal [37], who found that personalization is an added value in the context of skin health, as everyone’s skin condition is different. The qualitative results also support the findings of the questionnaire result (hypothesis 5). A total of 6 interviewees (n=30, 20%) revealed that interactions on short video platforms gave them positive feelings because they helped providing personalized information. Interviewee 2 stated:

In addition, 7 interviewees (n=30, 23.33%) had positive feelings about the interactions because they could help retain information. As Interviewee 7 said, “With the save feature, I now have a kind of personal archive that I can open again at any time without having to search from scratch.” In total, 3 interviewees (n=30, 10%) also had positive feelings about the interactions because the comments showed communal support. According to Interviewee 16, “If my comment is replied to or someone responds, I feel more confident and don’t feel alone.” For 3 interviewees (n=30, 10%), the interactions provided positive feelings because they increased their sense of involvement. Interviewee 15 stated, “When I like or save skin health content, I feel more involved with the content.” Thus, interactions on short video platforms through the like, comment, and save features make users feel that they have shown appreciation to content creators who provide useful skin health information. This interaction also positively affects user attitudes because users feel engaged and have control over the videos displayed on the short video platform.

Attitude had a weak effect on SHEs. This result is in accordance with Li et al [48], Neely et al [49], and Link et al [46]. Li et al [48] showed that users’ perceptions and positive attitudes in China were associated with increased health information search behavior in digital services. Neely et al [49] found that user engagement on social media platforms and user attitudes were factors driving users’ intentions and behaviors to search for health information on social media platforms in the United States. Furthermore, Link et al [46] explained that users’ positive attitudes toward online health information searches were associated with increased online health information search behaviors in Germany.

Moreover, interview results support the findings of the quantitative results (hypothesis 6). We found that 10 interviewees (n=30, 33.33%) felt that using short video platforms helped them search for skin health information; thus, they were encouraged to use short video platforms when they wanted to find skin health information. According to Interviewee 9:

Moreover, 8 interviewees (n=30, 26.66%) were satisfied with the experience of searching for skin health information on short video platforms; thus, they were encouraged to use the platform again. As Interviewee 10 stated:

In addition, 3 interviewees (n=30, 10%) felt comfortable using the short video platform when searching for skin health information; thus, they were encouraged to use the platform again. As Interviewee 13 revealed:

Thus, in the context of searching for skin health information on short video platforms, attitudes can be judged through users’ feelings toward short video platforms when searching for skin health information. With users’ positive feelings toward the process of searching for skin health information on short video platforms, they have the desire to search for health information on short video platforms.

USC affected SST. This result is in accordance with Tian et al [50], who found that USCs, particularly on social media, increased individuals’ anxiety about their appearance in the context of general appearance, not specific to the skin. In the context of searching for skin health information on short video platforms, USC is defined as a situation in which users compare their skin conditions with those on content that displays better skin conditions. Our interview results also support the quantitative finding (hypothesis 7). This study found that 3 interviewees (n=30, 10%) admitted that they had compared their skin conditions with those on the short video platform and that this comparison process gave rise to feelings of insecurity. According to Interviewee 3:

In addition, 14 interviewees (n=30, 46.66%) were motivated to have a good skin condition similar to that shown on the short video platform, were more diligent in taking care of their skin, and strived to solve the problems they experienced. Interviewee 27 said:

Users can also understand the content displayed using filters and makeup and promotional content in the form of marketing, which is not actually used by the content creator.

SST had no effect on SHEs. This finding is not consistent with those of Berger et al [53], Aslan Kayıran et al [54], and Rowlands et al [59]. Berger et al [53] showed that stigmatization due to skin health conditions had a significant effect on the behavior of seeking health information online, as individuals who feel stigmatized tend to look for alternative information through the internet as a form of coping with negative social judgments. Conversely, Aslan Kayıran et al [54] found that the perception of stigma in patients with psoriasis had a significant effect on the search for health information through social media and the internet. Rowlands et al [59] also discovered that SST can encourage individuals to search for health information digitally, as they tend to avoid formal help and opt for private channels, such as online platforms.

In accordance with the quantitative results (hypothesis 8), 6 interviewees (n=30, 20%) did not feel insecure about their skin conditions, as explained by Interviewee 12:

A total of 4 interviewees (n=30, 13.33%) were used to their skin conditions and thus no longer felt insecure. They conducted information searches more as a preventive measure or practical need. Interviewee 21 stated:

According to 2 interviewees (n=30, 6.66%), when they felt insecure about their skin conditions, they would conduct an information search through the help of medical professionals, such as doctors or clinics, rather than on short video platforms. Interviewee 18 said, “When I feel uncomfortable or have a breakout, I will immediately stop using the product and look for a solution directly at a physical place or clinic.” A total of 5 interviewees (n=30, 16.66%) immediately sought and bought skin health products when they felt insecure about their skin conditions. As Interviewee 5 explained, “If I’m insecure because of severe acne, I usually immediately look for products that can help.” Most of the interviewees did not feel stigmatized or were used to their skin conditions; thus, the stigma did not trigger the search for health information. Even when users’ confidence in their skin health increases, some prefer to seek information from other sources (doctors, clinics, or buying products directly) rather than using short video platforms. Thus, in the context of users of short video platforms in Indonesia, SST is not the main factor that drives the behavior of seeking skin health information, especially on short video platforms. Hussain et al [75] found that many women relied on filters for social validation despite recognizing their artificiality; thus, to date, beauty standards are increasingly coproduced by platform design and artificial intelligence–driven filters. In addition, beauty standards also influenced by culture, norm, and value [75].

PSE had a significant effect on skin health information seeking. This result is in accordance with Malik et al [9], Zhao et al [10], and Neely et al [49]. Zhao et al [10] showed that PSE had a positive effect on individuals’ willingness to search for health information through social media in individuals with chronic diseases in China. Increasing PSE was found to increase health information search behavior, as users want to avoid greater health threats [10]. This study is consistent with the study by Neely et al [49], which found that users with high PSE tended to search for more health information to reduce uncertainty in health decisions. Malik et al [9] also asserted that this perception is used to help avoid or reduce the risks that can arise due to certain skin health problems in Pakistan. In the context of searching for skin health information on short video platforms, PSE can be visualized in the content shown on short video platforms.

In addition, the interview result supports the questionnaire result (hypothesis 9). We found that 3 interviewees (n=30, 10%) stated that concerns about the content and long-term effects of using skincare products prompted them to seek further education on short video platforms. Interviewee 1 said:

In addition, 9 interviewees (n=30, 30%) were encouraged to actively seek information through short video platforms because of their concerns over the effects of skin problems, such as acne scars, stretch marks, or irritation. As Interviewee 3 stated:

Furthermore, 6 interviewees (n=30, 20%) conducted an information search to minimize the side effects when the initial symptoms began to appear. As Interviewee 23 explained, “I often look for information to reduce the contrast of acne scars when I have a breakout so that acne does not leave marks.” The higher the PSE, the greater the individual’s tendency to have skin health information–seeking behavior on short video platforms. Concerns about potential long-term effects and the desire to avoid or minimize skin problems are the main factors driving users’ search for information, both as preventive and early mitigation measures before deciding to seek professional help.

PSU significantly affected SHEs. This result is consistent with Malik et al [9], Zhao et al [10], and Link et al [46]. This indicates that the higher an individual’s perception of the likelihood of experiencing skin health problems, the greater the tendency to seek related information through short video platforms. Accordingly, Malik et al [9] showed that an individual’s perception of vulnerability to health conditions significantly affected preventive behavior, as individuals tend to take anticipatory steps when feeling vulnerable. Zhao et al [10] confirmed that PSU had a significant effect on health information–seeking behavior driven by the need to understand and control risk. Link et al [46] revealed that the perception of vulnerability to health problems significantly drives the search for information through online platforms.

The qualitative result also supports the findings of hypothesis 10 result. A total of 10 interviewees (n=30, 33.33%) were encouraged to be more active in seeking information as a form of preventive action because of their concerns about the potential appearance of skin problems. As Interviewee 3 stated:

In addition, because of their concerns about skin conditions, 3 interviewees (n=30, 10%) not only looked for solutions but also verified the accuracy of the information they obtained. As Interviewee 2 explained:

Thus, the higher the individuals’ concern about the possibility of having skin problems, the greater their tendency to be actively involved in the search, verification, and application of skin health information through short video platforms.

This study enriched previous studies that examined the behavior of searching for health information on social media in general [9,11,15,23,24] and discussed content quality with regard to the credibility of actual information [21,22,25]. We found that the role of algorithmic ecosystems in shaping users’ skin health behavior, the shift from SC to short video platform credibility, and the importance of visual media richness in aesthetic health contexts could influence users in searching skin health information on short video platforms. The findings of this study also provide theoretical implications for SOR, MRT, and HBM theories by showing that the effectiveness of media in searching for skin health information is highly determined by media wealth, which serve as a stimulus. In the context of MRT, short videos have several media that can be used by their users, such as a combination of video, photos, audio, and caption text, making them more effective in explaining visual skin health topics. This is because the combination of media used can reduce the ambiguity and complexity of information. In the context of HBM, the visualization of skin conditions increases PSE and PSU, which encourage information-seeking behavior represent organisms and responses, respectively. This implication confirms that the effectiveness of media in health communication is not only determined by the form of the media but also by the media’s ability to trigger users’ beliefs and perceptions of skin health, as described in HBM.

This study found that HCE and PHIs serve as stimuli and had an effect on PU. This finding can expand the MRT, in which the existence of content expressiveness and personalization can make media delivery effective in the context of short videos. In addition, findings related to HCE underscore that expressiveness depends not only on the media format but also on the ways of delivery, demonstration, and visualization of skin health content. These findings are consistent with those of Priyambodo et al [11], who showed that visual appeal affects PU. This study also found that PU is also formed from a combination of content expressiveness and personalization; thus, the research model of health information–seeking behavior needs to consider visual content–based factors as the main variables. This finding is in accordance with the study by Priyambodo et al [11], which revealed that PU is a factor that drives the intention to seek health information. These findings can strengthen the relevance of the technology acceptance model and confirm the research of Malik et al [9] and Jin et al [41] regarding the relationship between the usefulness of short video platforms and information retrieval behavior.

SC in this study has no effect on SHEs so that it can expand the research of Malik et al [9] by showing that the ecosystem of short video platforms is more influenced by the relevance of content and ease of access compared to the credibility of the source. This is supported by the behavior of users who consume other content before deciding to trust the information received on a short video platform. Based on Joseph [71], most of the short video platforms governed by algorithmic processes, where this platform acts as an intermediary that could shape perceptions of identity and self-representation and amplify pressures on users to conform to idealized norms. Therefore, short video platform users need to engage in active self-construction for developing digital habits that prioritize reflective awareness, diverse media consumption, and the scrutiny of artificial intelligence recommendations to offset the passive exposure to algorithmic feedback [71,72].

The results of this study also show that through interaction features, such as likes, comments, and saves, can have a satisfaction effect on the users’ behavior of searching for skin health information. Consistent with Liu et al [4], these findings may provide further support for theories related to interactions in discussing short video platforms. In contrast to Priyambodo et al [11], who showed that user interactivity does not have a significant effect on PU, this study examined the relationship between PI and attitude.

Attitude had a weak effect on SHEs in this study. Although attitude has only a weak effect, these findings suggest that attitude’s influence on behavior is not as strong as initial assumptions and therefore requires new theoretical adaptations in the domain of short video platforms. These findings can provide insight that there are other factors that can give users a positive feeling to the users of the short video platform.

Exposure to content with certain beauty standards, such as smooth skin without acne, can trigger a process of upward comparison, which then triggers feelings of dissatisfaction with an individual’s skin condition. These findings can strengthen the research of Adkins et al [26] by confirming that USC also drives SST behavior on short video platforms. The insignificance of the effect of SST on SHEs indicates that the relationship between stigma and search behavior may be mediated by other psychological factors that have not been studied in this study. These findings may provide implications that research models related to information-seeking behavior need to integrate psychological and cognitive factors to explain the relationship between stigma and information-seeking behavior.

In the end, the strong influence of PSE and PSU on information-seeking behavior provides validation for HBM. Users who feel that their skin conditions are serious, such as inflamed acne, dark spots, and severe eczema, are likely to be encouraged to seek out skincare information, especially when they feel that the condition they are experiencing has the potential to get worse. In addition, users who have a perception of susceptibility to a disease are also encouraged to seek information to prevent getting the disease. Priyambodo et al [11] have similar results to this study; PSE and PSU have an influence on the intention to seek health information, which then affects the actual online health information–seeking behavior. These findings confirm that PSE and PSU remain the driving factors for information-seeking behavior, especially in health domains that are highly visual, such as skin health. Overall, the findings in this study show that the behavior of seeking skin health information is the result of an interaction between cognitive, psychological, social, and technological factors.

From the perspective of short video platforms, platform providers need to strengthen their role as responsible information facilitators. This can be done by providing interaction features that allow educational discussions to take place well, such as fact-check prompts. Recommendation algorithms also need to be directed to provide greater space for evidence-based educational content and corrective interaction from professionals, not just viral content. In addition, official information panels and warning labels need to be integrated into content that is vulnerable to misinterpretation so that the interaction network on the platform does not reinforce misinformation. Features that directly connect users with various trusted sources are also needed. Examples of this are displaying verified educational video recommendations, incorporating educational panels on specific dermatology keywords, and providing links to official health institutions. In this way, users can cross-check without having to leave the platform.

As content creators of health care professionals, it is important to stabilize the quality of the information circulating through their active involvement in conversations on the platform. In addition to producing educational content that is clear, visual, and easy to understand, health care workers also need to be present in digital interaction spaces, such as comment columns or video replies, to correct mistakes, answer questions, and strengthen user understanding. The presence of health care workers in the flow of public conversations on short video platforms can increase users’ confidence in accurate information. Nonprofessional creators also need to be more careful when presenting skin health information. These creators need to maintain transparency, especially in sponsored content, avoid medical claims that cannot be validated, and open up space for collaboration or clarification with health workers. By leveraging interaction features, such as duets or stitches, creators can support the spread of informative content without misleading users.

Finally, the government needs to play a role in providing official references that can make it easier for users to verify information. Governments can work with platforms to link trusted sources, such as product regulations, dangerous product lists, or clinical dermatology guidelines. In addition, health literacy programs and public education about the risks of illegal products need to be strengthened so that the public is better able to distinguish valid information from false commercial content.

A limitation of this study is the demographics of the respondents, as most of the respondents were aged 20‐29 years and were taking or already had an education level equivalent to a bachelor degree. Thus, this study could not conduct a comparative study of ages. Moreover, the study did not evaluate other cognitive or psychological factors, such as prior knowledge or self-compassion, that could influence users’ behavior in searching for health information. This study did not conduct an analysis based on the type of skin health information—that is, primary information on medical and tertiary conditions and aesthetics oriented toward beauty care.

Further studies could explore the effects of SST and SC on more general research objects, such as general health, on short video platforms. Specific content differentiation is also necessary, given that skin health information has diverse types of information, such as product reviews, skin health education, and information related to skin diseases, which are likely to drive different behaviors.