Love is in the cloud: Uncovering the factors driving continuous use intention of online dating applications

Abstract Romance has never felt the same way since the inception of online dating. A match is made not in Heaven but in the Cloud. Swiping profiles could sweep someone’s feet off. This work aims to assess the factors predicting the continuous use intention of online dating applications among young Malaysians through the lens of the UTAUT2 model and privacy risk. The research model was analyzed using Partial Least Squares Structural Equation Modelling per 282 valid responses from online daters in Malaysia. The results indicate the model’s substantial predictive power with social influence, price value, and habit as significant drivers to continuous use intention of online dating applications. Habit has the strongest effect on the intention to use online dating applications continuously. This research adds to the body of knowledge on the behavioral intention of online dating applications and offers practical suggestions for online dating service providers to increase the likelihood of their continuous adoption.


Introduction
"Online Dating is Like Online Shopping Except You're Looking for People No One Wants and It's $50 a Month." -Phil Pivnick. An avid online dater might self-deprecatingly resonate with the comedian's sarcasm. Although, nothing in the history of humankind has ever paralleled the revolutionary effect of online dating on courtship and matchmaking. Plenty of fish are on the Net, romance is ABOUT THE AUTHORS Tze Wei Liew is a Senior Lecturer at the Faculty of Business, Multimedia University Malaysia. His research interests and contributions intersect cognitive psychology, learning sciences, media psychology, cyberpsychology, human-computer interaction, and human-agent interaction.
Su-Mae Tan is a Senior Lecturer at the Faculty of Information Science & Technology, Multimedia University Malaysia. Her research interests are human-agent interaction, media psychology, and elearning.
Joe Yin Sung is a research student at the Faculty of Business, Multimedia University Malaysia. Her research interests are knowledge management and information sciences.
Chin Lay Gan is a Senior Lecturer at the Faculty of Business, Multimedia University Malaysia. Her research interests are cyberpsychology, mobile learning, and technology management.
Yi Yong Lee is a Ph.D. candidate at the Faculty of Business, Multimedia University Malaysia. Her research interests are knowledge management, cyberpsychology, and consumer behavior. a two-way swipe, and a match is made in the Cloud. Lucky suitors swiping in an online dating application could sweep someone's feet off.
Reports forecast the online dating market in Malaysia to acquire RM 21.8 million by the end of 2023 and RM 24.4 million with 1.19 m users by 2027 (Muller, 2023a(Muller, , 2023b. The same reports indicate that 42% of Malaysian online dating app users are between 24 and 34. Several reasons prosper the online dating market. Contemporary busy lifestyles restrict time and flexibility for conventional dating practices, and online dating remediates this conundrum by initiating meetings, dialogues, and courtships that transcend time and physical borders. Today's hyper-connected digital society facilitates friendships ascending to romances. What was once a stigma and taboo, online dating is increasingly regarded as socially acceptable and even lauded for its practicality and efficacy in exploring potential relationships. Online dating services are powered by cloud services, AI algorithms, geolocation, communication interfaces, and network infrastructures. The technology allows user profile creation, prospecting via search and filter, communication via text, voice, or video, and AI-assisted pairing to help them specify and relate with potential counterparts based on favored socio-demographics, interests, preferences, and location traits. An array of online dating applications is available, with Tinder, Dating.com, Tantan, OKCupid, Grindr, Paktor Bumble, and Coffee Meets Bagel leading the Malaysian market (Muller, 2023a). They feature extremely intuitive actions, including swiping right, tapping a heart-shaped love icon, and conveying virtual gifts or emoticons to exhibit interest in a potential partner.
Not all online dating applications enjoy continuous user adoption success and popularity. Some platforms, such as Facebook Dating, have far fewer active users than contenders like Tinder, while other variants, including HowAboutWe and Missed Connections, are now defunct (Iovine, 2022). Amid a flurry of dating applications, some are bound to accrue traction while others dwindle in user acceptance over time.
What factors might influence the continuous adoption of dating applications? From a usability perspective, users may assume that an online dating platform fails to promote their profile visibility or maximize matchmaking success likelihood (Kolhoff, 2021). Users ultimately reject the application when it does not help accomplish their dating goals. Also, users expect online dating operations to be smooth and pleasant. Challenging navigation, complicated interface, technical faults, and frequent failures frustrate users until they exit the application and seek alternatives.
Social norms and acceptance may affect the adoption of online dating applications, especially within a conservative culture where online daters face stigmatization (Rochadiat et al., 2018;Siddiqui et al., 2023). The preceding relates to social and moral reputations affixed to certain online dating applications; for instance, Tinder users lean toward casual hook-ups, while Coffee Meets Bagel users emphasize long-term commitment. Another user-critical concern is privacy risk (Brandtzaeg et al., 2019;Stoicescu et al., 2019), where personal and sensitive information may be leaked, hacked, tracked, or shared without consent. Users will hesitate to use an online dating application deemed risky (Canta et al., 2021;Siddiqui et al., 2023). Amid increasing living costs, users may also becoming less inclined to pay for subscriptions and premiums associated with online dating applications (Criddle, 2022).

Related works on behavioral intentions of online dating applications
Based on the foregoing, examining the factors influencing the use intention of online Dating applications through the lens of technological acceptance frameworks is beneficial. Alam et al. (2018) tested a research model comprising word-of-mouth, attitude, perceived enjoyment, perceived playfulness, and trust as potential predictors driving intention to use online dating sites among young people. The researchers discovered that word-of-mouth, attitude, perceived enjoyment, and perceived amusement are significant predictors of Malaysian young adults' willingness to use online dating sites. Balan et al. (2021)'s study revealed that perceived usefulness and continuous use intention of dating applications in Malaysia were affected by gratification elements, including entertainment, sexual activity, friendship, romantic relationship, social inclusion, and location-based search.
Respati and Amalia (2021)'s study drew on theories concerning self-esteem, permissive sexual attitude, and motivation to understand the factors driving the intention to use online dating applications in Indonesia. Their findings indicated that self-esteem, permissive sexual attitudes, and motivation significantly influence the acceptance of online dating platforms. Grounded on the innovation resistance theory, Siddiqui et al. (2023) demonstrated that barriers related to risk, usage, and tradition negatively impact the adoption intention of online dating applications in India. Chakraborty (2019) explored whether word-of-mouth, attitude, perceived delight, perceived playfulness, and trust impact the willingness to use dating applications in India. The researcher found that playfulness most profoundly affects acceptance of dating applications. Huwae and Ng (2022) observe that less research has investigated the Unified Theory of Acceptance and Use of Technology (UTAUT2) contextual to online dating applications among young people, despite the model's touted robustness in explaining adoption intention across other technological domains. Indeed, recent studies have only begun to adopt this framework to predict the use intention of online dating platforms. Huwae and Ng (2022)'s investigation discovered that two UTAUT2 constructs, i.e., hedonic motivation and habit, significantly predict the use intention of dating applications among Gen-Zers in Taiwan. Chan et al. (2023)'s recently extended the Unified Theory of Acceptance and Use of Technology (UTAUT) model by incorporating the trust variable for the technology adoption of online dating applications. Their findings identified performance expectancy, social influence, and trust as significant antecedents of the intention to adopt online dating applications among young people in Malaysia.

Research objective and question
Findings across studies suggest that the effects of the UTAUT model's variables on the behavioral intention of online dating applications differ and, therefore, cannot be generalized between countries (Chan et al., 2023;Huwae & Ng, 2022). Within the Malaysian context, Chan et al. (2023) adopted the UTAUT's constructs of performance expectancy, effort expectancy, social influence, and facilitating conditions. They also integrated trust as an additional variable.
In this work, we extend the research to incorporate UTAUT2 as the base framework for explaining the continuous use intention of online dating applications among young people in Malaysia. Beyond the original UTAUT applied in prior work (Chan et al., 2023), this study features the UTAUT2 model, which includes the additional constructs of hedonic motivation, price value, and habit. This expanded model may offer a more comprehensive insight into the continuous use intention of online dating applications in Malaysia. In addition, we adopt the perceived risk as an additional factor, given that online dating applications solicit and manage private and sensitive data susceptible to leaks, unscrupulous tracking, or misappropriation. Noteworthy, we omit facilitating conditions construct from our research model, following Huwae and Ng (2022)'s contention that the technological and organizational framework for using technology is irrelevant to the online dating application domain where most users are digitally-savvy Gen-Zers. The research question of this work is formulated as follows:

RQ:
To what extent do Performance Expectancy, Effort Expectancy, Social Influence, Hedonic Motivation, Price Value, Habit, and Privacy Risk affect the continuous use intention of online dating applications? Venkatesh et al. (2012) proposed the Unified Theory of Acceptance and Use of Technology (UTAUT) to forecast users' behavioral intentions regarding information systems and technology. The seminal version of the model comprises four primary factors: performance expectancy, effort expectancy, social influence, and facilitating conditions. The model was then upgraded to UTAUT2, incorporating three additional constructs: hedonic motivation, price value, and habit. The UTAUT 2, considered the most comprehensive integrative theory for predicting personal technology use (Tamilmani et al., 2021), has demonstrated superior explanatory power for understanding acceptance factors of new technology compared to other models such as TAM, TAM 2, TRA, and UTAUT (Gharaibeh et al., 2018). This study excludes facilitation conditions from the research model as we are not concerned with how accessible the system's technological and administrative resources are since it is focused on online dating applications for the younger generation, especially Generation Z (Huwae & Ng, 2022). Online daters are likely already familiar with smartphones, the Internet, and social media; thus, facilitating conditions are insignificant in this work. Chan et al. (2023) 's recent finding affirms this contention; it was found that facilitation conditions were not significantly related to the acceptance of online dating applications among young people in Malaysia.

Performance expectancy
Performance expectancy encapsulates users' assumption that using a system will help them improve task performance and fulfill intended goals (Venkatesh et al., 2012). Performance expectancy promotes user satisfaction and continuous use intention of technology for health IT services, e-payment, online tax filing, mobile learning, online ticketing, and e-learning webinars (Khechine et al., 2016). Online dating services offer flexible and convenient access to a larger population of potential matches than traditional face-to-face meetings, thus increasing the likelihood of finding suitable companions (Finkel et al., 2012). Further, AI-powered algorithms can generate compatible pairings based on geolocation, interests, lifestyles, and socio-demographic traits. Therefore, online daters believing that online dating applications will facilitate and increase matchmaking possibilities are likelier to continue using the technology (Balan et al., 2021;Chan et al., 2023). Contrariwise, if users perceive that online dating applications' algorithm does not effectively optimize their visibility or partner recommendations, they will not accept the platform (Kolhoff, 2021). Thus, H1: Performance expectancy positively affects the continuous use intention of online dating applications.

Effort expectancy
Effort expectancy signifies users' belief that using a system is simple, effort-free, and requires a low learning curve (Venkatesh et al., 2012). The literature shows that digital healthcare, e-payment, and online tax filing systems that are easy to use garner more user acceptance and behavioral intention (Khechine et al., 2016). Online dating applications have intuitive interfaces allowing users to effortlessly swipe right to express interest or left to reject potential partners based on profile traits. In addition, tapping actions enable users to communicate likes and emoticons to profiles and send virtual gifts to others. Signing up for and establishing a profile is a simple guided process where users input basic information, upload pictures, and construct a bio. Online dating platforms notify users of new matches, messages, and profile activity, which eases the matchmaking process. It is thus expected that: H2: Effort expectancy positively affects the continuous use intention of online dating applications.

Social influence
Social influence represents the degree to which the opinion and judgment of significant others, such as family, friends, colleagues, celebrities, and admirably respected persons, influence users' decisions to use technology (Venkatesh et al., 2012). A review of the UTAUT model places social influence as a robust predictor of technology acceptance across different domains, including healthcare IT, e-payment, e-library, online tax filing, and mobile learning (Khechine et al., 2016). Although modern society increasingly accepts online dating, negative connotations still occur. For instance, Tinder users often project the casual hook-up stereotype, invoking social stigmatization and moral reproval, particularly in a conservative culture (Cöbek & Ergin, 2021;Rochadiat et al., 2018). However, Huwae and Ng (2022) found that social influence does not predict online dating application use among Taiwanese Gen-Zers. On the other hand, Chan et al. (2023) demonstrated that social influence significantly relates to the adoption of online dating applications among young Malaysians. Malaysia is a progressive nation, albeit interwoven with conservative and collectivist Asian values. In this regard, Malaysian online daters are susceptible to the sentiments and critiques of societal members that impact the continuous adoption of online dating applications. We, therefore, expect that: H3: Social influence positively affects the continuous use intention of online dating applications.

Hedonic motivation
Hedonic motivation denotes how users derive fun and pleasure from a technological application (Venkatesh et al., 2012). This construct has significantly affected behavioral intention across Fintech, mobile payment, digital banking, e-learning, and mobile AR systems (Tamilmani et al., 2019). Online dating activities are often driven by seeking thrills and excitement (Bryant & Sheldon, 2017;Sumter et al., 2017). Online dating applications exhibit multiple features projecting hedonic cues, including visually-arousing profile pictures and biodata, flirty and emotive messages, and virtual gifts. In addition, searching, swiping, and tapping sparks exhilaration, eliciting hedonic motivation from anticipating romantic rewards. Studies have confirmed the positive relationship between hedonic motivation and the intention to use dating applications (Alam et al., 2018;Chakraborty, 2019;Huwae & Ng, 2022;Respati & Amalia, 2021). Hence, we predict that: H4: Hedonic motivation positively affects the continuous use intention of online dating applications.

Price value
Price value denotes users' perceived worth or utility of a technological application against its financial costs (Venkatesh et al., 2012). To maximize matchmaking success rates, online daters consider spending money on subscriptions, premium status upgrades, and virtual gifts. Upgrading to Tinder Plus features the ability to reverse reject swipes, limitless liking capacity that exceeds the daily limit imposed by Tinder's algorithms, and the Passport function, which allows users to interact with people worldwide (David & Cambre, 2016). Upgrading to Bumble Premium extends matches by 24 hours and privileges unlimited swipes, advanced filters, and incognito mode (Barros et al., 2022). Obtaining a Tantan VIP badge allows infinite swipes, modifying location filters, unlimited rewinds, and conveying five "Super Likes" daily (Pleines, 2023). These additions enhance users' chances of meeting compatible prospects and striking up conversations. Users can improve their chances of finding a good match by narrowing their search parameters using features like extending matches and sophisticated filters. Further, virtual gifting is a booming business, where online daters spend money, tokens, or credits to gift other roses, airplane tickets, movies, spa treatments, dinner, drinks, jewelry, or lingerie (Davis, 2019;Wade, 2011). Virtual gifting spotlights a generous member among the crowd, thereby maximizing their matchmaking success. Thus, it is expected that: H5: Price value positively affects the continuous use intention of online dating applications.

Habit
People are creatures of habit and routine. Applied to IS/IT adoption, habit signifies repeated and consistent technology use behaviors that become automatic, natural, mindless, and routine derived from accumulated information and experiences over time (Tamilmani et al., 2018;Venkatesh et al., 2012). Long-term utilization of a technological application, specifically throughout product development, results in users forming habits around that technology or product. Online dating applications entail habitual usage, with online daters performing swiping for about 55 minutes daily (Reporter, 2023). Gen-Zers are technologically savvy, which promotes habit formation. Intuitive actions like swiping and tapping require less conscious thoughts, increasing the odds of developing reflexive habits around using online dating applications. More critically, online dating activities prime the brain's reward systems to release dopamine, the "feel good" neurotransmitter. This generates a reward-centric and instant gratification feedback loop that fosters addictive use habits of online dating applications (Narr, 2021;Waters, 2021). Tamilmani et al. (2018) 's paper documents habit as a significant antecedent of use intention across technological services, including e-learning, e-banking, mobile payment, online shopping, and social media. In light of this, users will likely use the dating app regularly once a routine is formed. Familiarity with a certain technology, such as online dating applications, increases the effects of habit in terms of intention and usage (Tamilmani et al., 2018;Venkatesh et al., 2012). Hence, H6: Habit positively affects the continuous use intention of online dating applications.

Privacy risk
Privacy risk refers to users' concern that using a technology entails the vulnerability of having sensitive and confidential information hacked, leaked, divulged, tracked, or misused without consent (Kasilingam, 2020). This factor is especially critical to using online dating applications, given their personal and intimate information, including biodata, geolocations, photographs, messages, and videos (Bonilla-Zorita et al., 2021;Chen et al., 2021;Lutz & Ranzini, 2017;Siddiqui et al., 2023). Thus, online daters must balance their preferences and the application's privacy parameters when sharing private information (Stoicescu et al., 2019). A user's Tinder profile is linked to their social media profile, and mutual friends are revealed to other users who have also connected their Tinder and social media accounts. However, Tinder's privacy policy is vague, indicating they may gather and keep users' information without overtly stating so. This may challenge the assumption that online dating applications are fully transparent and question what other data these services have access to. As a result, online daters fearing privacy risks will be hesitant to use online dating applications. More formally, H7: Privacy risk negatively affects the continuous use intention of online dating applications.

Research model
Based on the preceding hypotheses, Figure 1 illustrates the research model of this research.

Data collection and sampling
This study used a cross-sectional survey with questionnaires distributed online to collect primary data. According to a recent report, Malaysia's internet penetration rate was 89.6 percent of the total population at the beginning of 2022 (Kemp, 2022). According to recent reports, 93% of Malaysian Internet users actively utilize social media (MCMC, 2020). An online survey form was created, and the link was shared on social media, including Facebook, Instagram, WeChat, and WhatsApp, to facilitate data collection. Respondents gave informed consent upon understanding and agreeing to participate in the online questionnaire. The data and personal information collected from respondents were guaranteed confidentiality and used solely for research purposes, with participation in the survey being entirely voluntary.
A non-probability purposive sampling technique was employed in this study. The purposive sampling design implies that samples are chosen depending on the researcher's judgment (Chan et al., 2023). Per ethical considerations, the survey was limited to respondents 18 years old and above. Many popular online dating sites require an 18-year-old minimum age; minors under 18 are prohibited from online dating services and applications (Bee, 2021;OkCupid, 2022).
The measurement items for each research construct were adapted from previous literature (Escobar-Rodríguez et al., 2014;Schomakers et al., 2022;Xu et al., 2015) and modified according to the current research objectives, and the Five-Point Likert Scale with a scale ranging from strongly disagree (1) to strongly agree (5) was used.

Data analyses and results
To evaluate the relationships in the structural model, the current study employed structural equation modeling (SEM) using partial least squares (PLS) predictions. Because this study aims to analyze the theoretical framework for prediction objectives and explain the endogenous variables, PLS-SEM is an appropriate and preferable technique (J. J. Hair et al., 2014).

Descriptive statistics
Three hundred survey response sets were acquired for this study. Eighteen straight-lining responses were omitted at the data examination stage (J. J. Hair et al., 2014), generating 282 valid responses. This resulted in a 94% response rate, far exceeding the G*Power suggestion; a sample size of 103 is sufficient to achieve a statistical power of 80% (Faul et al., 2009). Tantan is the most popular online dating app, with around 59.9% of people having used it previously, followed by Soul (37.9%), Tinder (25.9%), Coffee Meets Bagel (11%), Bumble (5.7%), and OkCupid (4.6%). 64.5% of the respondents have used them for less than a year. Of those surveyed, 100 (23.4%) have used it for over a year. All respondents indicated that beyond free-to-use services, they had spent money on subscriptions, virtual gifts, or premium privileges.

Common method bias
This study examined the issue of Common Method Bias by assessing full collinearity (Kock, 2015). All constructs will be regressed on a common variable. The analysis revealed that all variables had a VIF less than 5, including adoption intention (3.982), effort expectancy (2.239), hedonic motivation (2.816), habit (4.240), performance expectancy (2.535), privacy risk (1.049), price value (2.884), and social influence (2.281), indicating that single-source bias is not a significant issue with the current research data (J. J. Hair et al., 2014).

Multivariate normality assumptions
Based on the values of Mardia's multivariate skewness (β = 17.120, p < 0.01) and kurtosis (β = 118.041, p < 0.01), the findings demonstrated that the dataset was not multivariate normal. The skewness and kurtosis values exceed the multivariate normality criterion of ±3 and ±20 (Kline, 2016). The multivariate assumption test revealed that the data's normal distribution assumption was violated. Therefore, the bootstrapping technique was utilized to correct the standard error using SmartPLS software, a non-parametric analysis tool (J. J. Hair et al., 2014).

Measurement model assessment
The outer loading, composite reliability (CR), and extracted average variance (AVE) were used to evaluate the measurement model. The outer loading, composite reliability (CR), and extracted average variance (AVE) were used to assess the measurement model. For a reflective measurement model, indicator loadings should be significant, with a minimum value of 0.708 (J. J. Hair et al., 2014). Three items for measuring privacy risk (PR2, PR3, PR4) and one for measuring pricing value (PV3) were removed due to lower outer loading. Table 1 shows that the outer loading (≥0.708), CR (≥ 0.70), and AVE (≥ 0.50) values of all constructs fulfill the minimum benchmarks (J. F. Hair et al., 2011). Table 2 shows that the heterotrait-monotrait correlation ratio (HTMT) was less than 0.90 (Henseler et al., 2015). These findings presented a sufficient basis for the discriminant validity of all constructs.

Structural model assessment
Per the multivariate normality results, it was specified that the data was not multivariate normal; thus, this study follows J. Hair et al. (2019) 's recommendation for reporting the path coefficients, standard errors, t-values, and p-values , and the effect sizes and confidence intervals (Hahn & Ang, 2017) for the structural model using a 5,000-sample re-sample bootstrapping procedure. Table 3, three of the seven relationships were determined to have a t-value of ≥1.645, therefore being significant at the 0.05 significance level. The result indicates that social influence (β = 0.159; t = 2.760; p = 0.003), price value (β = 0.113; t = 1.658; p = 0.049), and habit (β = 0.544; t = 8.221; p < 0.001), were positively associated with adoption intention to use online dating apps, thus supporting Hypothesis H3, H5, and H6 respectively.

As summarised in
Performance expectancy (β = 0.072; t = 1.096; p = 0.137), effort expectancy (β = 0.060; t = 1.134; p = 0.128), hedonic motivation (β = 0.047; t = 0.692; p = 0.244), and privacy risk (β = 0.029; t = 0.667; p = 0.253) were not significantly associated with adoption intention to use online dating apps, thus rejecting hypothesis H1, H2, H4, and H7 respectively. The model explained 74.9% of the variance in adoption intention to use online dating apps. The R 2 value (0.749) exceeded the 0.26 threshold that Cohen (1988) suggested, endorsing the model's substantial predictive power. Figure 2 depicts the findings of the structural model analysis.  There is no multicollinearity concern because all VIFs range from 1.046 to 3.060 (J. Hair et al., 2017). Except for habit, all determinants had a small effect on the intention to utilize online dating applications (Cohen, 1988). Habit (f 2 = 0.385) has a larger influence on predicting intention to use online dating applications. Shmueli et al. (2019) further suggested employing PLSpredict. This holdout sample-based technique generates case-level predictions on an item or construct level utilizing the PLS-Predict with a 10-fold procedure to check for predictive relevance. According to Table 4, all of the errors (RMSE) of the PLS model were smaller than those of the LM model (PLS-LM), implying that the current study model has a high predictive relevance (Shmueli et al., 2019).

Discussion
Through the lens of UTAUT2, this study found that the continuous use intention of online dating applications among young people in Malaysia is significantly impacted by social influence, price value, and habit. Concerning social influence, our finding coheres with Chan et al. (2023)'s study, indicating that the continuous adoption of online dating applications among young people in Malaysia is robustly driven by opinions and judgments of significant others, including family members, friends, colleagues, and respected society members. Malaysia's collectivist and conservative culture prioritizes social cohesion, mutual support, and norms. Online dating activities may channel positive images like long-term and honest relationships or negative stereotypes linked to one-night stands, infidelity, or casual hook-ups (Castro & Barrada, 2020;Rochadiat et al., 2018;Siddiqui et al., 2023;Sumter & Vandenbosch, 2019). Collectivist society members strive for social unity and harmonious well-being; thus, they are highly aware of whether their actions may provoke moral concerns threatening the group's well-being. Framed within this context, Malaysian online daters' continuous acceptance of dating platforms is strongly shaped by social approval and expectations regarding online dating activities.
Our data demonstrate that price value is a significant antecedent of continuous intention to use online dating applications among young people in Malaysia. Therefore, online daters emphasize whether they believe their financial investment in subscriptions, premium services, and virtual gifts is justified based on the derived benefits. Indeed, a wide array of premium services exist across online dating applications (Barros et al., 2022;David & Cambre, 2016;Davis, 2019;Pleines, 2023;Wade, 2011), inspiring online daters to stand out from the crowd and maximize compatible matches. This study's respondents perceived that the monetary costs for their online dating

Indicator
Outer Loading CR AVE PR6 -I am concerned that the online dating apps will sell my personal information to others without my permission. services are reasonable in generating added value, utility, and benefits, which encouraged them to adopt online dating applications continuously.
Based on this study's analyses, habit was found to have a large effect on the continuous use intention of online dating applications in Malaysia. The emergence of online dating applications in Malaysia is hardly new; therefore, the technology adoption generally has passed the early adoption phase. Malaysian online daters are likely to have accrued substantial experience and savviness to form routines in using online dating applications regularly. Online dating activities can be highly addictive, given their ability to trigger the brain's reward system to release dopamine in anticipation of achieving relationship goals (Narr, 2021;Waters, 2021). In light of these, habits are formed, prompting automated and subconscious behaviors that promote the continuous usage of online dating applications among young people in Malaysia. Our finding accords with those of Huwae and Ng (2022)'s recent study, which has documented habit as the most significant predictor of the use intention of online dating applications among Taiwanese Gen-Zers.
Interestingly, the online dating applications' functional aspects comprising performance expectancy and effort expectancy have no impact on the continuous intention to use online dating applications among young Malaysians. This parallels Huwae and Ng (2022)'s results contextual to young Taiwanese online daters. Pending further exploration, we offer some conjectures as follows. Malaysian young people may not always consider online dating applications as the sole means to search for prospective dating partners. They may explore dating activities via physical avenues or prototypical social media such as TikTok, Instagram, or Facebook. It is also possible that the use of online dating applications was not exclusively motivated by the intent to seek romantic partners. Rather, users may tinker with online dating applications to satiate their curiosity, boredom, loneliness, or fear of missing out. As for effort expectancy, Malaysian online dating application young users are likely to be tech-savvy; thus, the extremely low learning curve for operating platforms' intuitive interfaces is too inconsequential to impact their continuous use intention directly (Chan et al., 2023).
Despite studies suggesting that hedonic motivation can positively affect the use intention of online dating applications (Alam et al., 2018;Chakraborty, 2019;Huwae & Ng, 2022;Respati & Amalia, 2021), we discovered no significant relationship between hedonic motivation and the technology's continuous use intention among young Malaysians. This result is surprising given that online dating applications convey aesthetically pleasing visual (e.g., attractive pictures) and emotive verbal (e.g., flirtatious messages) cues that evoke fun and enjoyment, potentially stimulating continuous adoption. Plausibly, this is attributed to different motivations for online dating activities: entertainment versus meaningful connection (Kallis, 2020). Our result retrospectively suggests that young Malaysians' primary basis for continuously using online dating applications is not purely to pursue pleasure and entertainment. Rather, beyond initial excitement and novelty derived from the initial technology engagement, utilitarian objectives associated with seeking long-term, serious, compatible, and committed romantic connections may become the dominant motivation for utilizing online dating platforms among young Malaysians. This notion fits with Malaysian sociocultural traits; the respective literature has shown that members of a collectivist compared to an individualistic society, are less likely to seek hedonic values and self-gratification when using digital technology (Evanschitzky et al., 2014;Jadil et al., 2022;Ozen & Kodaz, 2012). In addition, the collectivist and traditional cultural norms in Malaysia may socially favor seeking meaningful relationships while casting moral aspersions on purely pursuing thrills, pleasure, and fun associated with casual relationships or hook-ups in online dating applications.
This study revealed that privacy risk does not impact young Malaysian's continuous intention to use online dating applications. This finding discords with studies documenting perceived risk as a significant barrier to the adoption intention of online dating applications (Chen et al., 2021;Siddiqui et al., 2023). Although this research did not measure users' perceived trust, we deduce that this study's respondents generally placed their faith in online dating platforms to safeguard private and sensitive data. Chan et al. (2023)'s recent work discovered that Malaysian young people trust online dating applications, subsequently elevating the adoption intention of the technology. Similarly, our study's respondents might have been firmly confident about their ability to monitor and control their privacy parameters in online dating platforms. Collectively, trusting beliefs might have limited the effects of perceived privacy risk on the continuous use intention of online dating applications among young people in Malaysia.

Implications for theory and practice
Theoretically, this study endorses the robustness of the UTAUT2 base model for predicting the continuous use intention of online dating applications. Overall, this study's research model, which adapts UTAUT2 and integrates perceived privacy risks, explains about 75% of the variance in young Malaysian's continuous adoption intention of online dating applications, thereby affirming the model's significant predictive power. The findings of this work extend the budding research intersecting the UTAUT framework and online dating application acceptance (Chan et al., 2023;Huwae & Ng, 2022) by presenting the following theoretical insights. In this work, employing the UTAUT2 beyond the original UTAUT model (Chan et al., 2023) leads to discovering the UTAUT2"s habit and price value constructs as significant antecedents of continuous use intention of online dating applications among young Malaysians. The findings of this study, when compared with those of Huwae and Ng (2022)'s work that used the UTAUT2 model in predicting Taiwanese Gen-Zer's online dating application behavioral intention, firmly imply that habit is the most robust factor driving the use intention of online dating platforms.
We offer practical recommendations following the findings of this study. To leverage the impact of social influence, online dating service providers can bridge the online dating pool with users' social circle in their social media accounts (e.g., Facebook Dating). This enables social recommendations for prospective dating partners based on mutual relationships and interests, thereby accentuating the approval, validation, and encouragement from significant others, including friends and family. Spotlighting endorsements, testimonials, positive ratings, and success stories by peers, celebrities, and influencers can boost the positive social influence supporting the acceptance of online dating applications. Referral programs that offer incentives for recommending online dating applications to friends, family, or colleagues can boost social validation for potential users to adopt the technology.
Given the importance of price value in predicting continuous intention to use online dating applications, online dating service providers must pay careful attention to pricing strategies. Online dating service providers can design different pricing packages and tiers based on unique dating needs and financial affordability to enhance perceived price value for a wider range of diverse users. Promotional offers, free trials, and discounts can compel user upgrades to premium features. Additionally, users must be presented with transparent pricing schemes that inform the privileges and benefits derived from the added features like unlimited swipes, profile visibility spotlight, advanced search algorithms, and exclusive gifts to maximize matchmaking success. Advertising testimonials and success stories of premium users can reinforce the price value impact on users' intention to use online dating applications continuously.
How can habit be leveraged to encourage continuous use intention of online dating platforms? Habit is crystalized through a series of consistent behaviors repeated long-term that become automatic and mindless. Interface cues must be designed to prime repetitive user activities in online dating applications. Conveying interface prompts, reminders, and notifications nudges users to constantly log on and engage in online dating platforms. Intuitive and seamless actions like swiping and tapping can foster users' reflexive routines in online dating platforms. Finally, gamification can catalyze habit formation by offering recurring points, rewards, badges, gifts, challenges, and ranks; these elements motivate regular and consistent engagement in online dating platforms. Thus, habit stemming from these dopaminergic cues leads to continuous use intention of online dating applications.

Conclusion, limitations, and future outlook
This research fills a gap where scant online dating application adoption research has been done through the lens of the UTAUT2 framework. The key finding is that habit represents a large effect in predicting continuous use intention of online dating applications among young Malaysians, which is consistent with Huwae and Ng (2022)'s study with Taiwanese young online daters. Social influence and price value are significant drivers of Malaysian online daters' continuous intention to adopt online dating applications. Thus, online dating service providers must leverage design and managerial strategies to accentuate habit formation, social influence, and positive perception of price value to foster long-term usage of online dating applications.
We acknowledge this study's limitations and offer recommendations for future research. The respondents in this study were mostly between 18 and 27, leading to the findings being representative of young Malaysian adults. Thus, researchers should refrain from generalizing the results to people of different ages. Further, this study's findings are derived from Malaysian respondents, thereby, can not be generalized to other countries. We also note that this research's dataset comprised mostly Chinese-ethnic respondents. Hence, our findings may not be robust for other Malaysian cultures and ethnicities, including Malays and Indians. Significant cultural differences and values exist across these major race groups, partly shaped by religions and social-moral codes. This also relates to the data demonstrating Tantan as the dominant online dating application among the respondents. Tantan is a China-based application that is popular among Asian ethnic users. It boasts a stronger verification and security system, has lesser bots and fake profiles, and offers cheaper premiums than its rivals like Tinder (Pleines, 2023a(Pleines, , 2023b. Thus, the unique traits of specific dating applications may influence the results of this study. To counter these limitations, future works should broaden the sample groups to cover more diverse age groups, cultural traits, and specific online dating applications (Chan et al., 2023;Huwae & Ng, 2022).
Given that many UTAUT2 constructs are insignificant in predicting online dating applications' continuous adoption intention, future research may integrate contemporary factors to particularize the technology's unique characteristics. Gender as a potential moderating factor may differently influence the relationship between the variables and adoption intention of online dating applications (Huwae & Ng, 2022), on the account that males and females may hold divergent expectations, preferences, and priorities associated with online dating behaviors (Fink et al., 2023;Guadagno et al., 2012). Similarly, other individual traits such as personality, self-esteem, sexual attitude, and sexual orientation may result in different factors driving continuous use intention of online dating applications (Chan et al., 2023;Respati & Amalia, 2021); hence, extending the research stream to incorporate these potential moderating constructs can offer a more comprehensive view.