Event marketing is an experience-based marketing tool focused on consumer experience and driven by both emotional and rational factors to create a complete experience for consumers (Setiawan et al., 2022). Event marketing has an extraordinary ability to create memorable experiences for its target market because it can shape emotions and memories (Satriya et al., n.d.). According to Wood and Masterman in Maharani and Mustikasari (2020), the dimensions of event marketing are as follows: involvement, which is the company's ability to make the audience feel the atmosphere of the event; interaction, which is the communication process between consumers and service or product providers; impression, the impression felt by consumers during the event; intensity, referring to memorable and impactful experiences; personalization, which is a unique and memorable experience for the individual; innovation, the organizer’s ability to create unique events; and integrity, which concerns the authenticity of the event and the organizer’s ability to contribute.

Promotion is an effort made by companies to highlight product advantages while convincing consumers to make purchases (Farisi & Siregar, 2020). Promotion is one of the factors that can lead to customer satisfaction (Indriani, 2024). According to Kotler and Armstrong in Hasanawi and Imsyar (2022), the dimensions of promotion are as follows: advertising, which involves using media to promote products; personal selling, the direct interaction between sellers and prospective buyers; sales promotion, activities that encourage short- term sales; publicity and public relations, which build a positive company image through press releases, events, and other activities; and word-of-mouth information, recommendations from consumers to others about products or services.

Brand image can be defined as the perception formed in the minds of consumers based on how a brand reflects itself in consumer memory during interactions with the brand (Marlius & Anwar, 2023). According to Lassar in Mashuri (2020), the dimensions of brand image are as follows: good name, which refers to the status held by a particular brand; recognition, the consumer’s awareness of the brand; emotional attachment, the emotional bond experienced between consumers and producers; and brand loyalty, which is the consumer’s long-term use and recognition of the brand.

Customer loyalty is defined as the customer’s commitment to a brand, store, or service provider, based on a strong positive attitude and manifested through consistency in repeat purchases (Arianto & Kurniawan, 2021). According to Kotler in Mashuri (2020), the dimensions of customer loyalty are as follows: repeat purchase, where consumers buy products again due to satisfaction or loyalty; recommending to others, where consumers suggest products to others based on positive experiences; and resistance to influence, where consumers remain loyal to the brand despite the presence of competitors.

In the normality test, the hypothesis states that if the significance value (Sig) is greater than 0.05, the residual data are normally distributed; conversely, if the Sig value is less than 0.05, the residual data are not normally distributed. The test result using the One-Sample Kolmogorov-Smirnov method produced a Sig value of 0.200, indicating that the residual data are normally distributed, and thus the null hypothesis (H₀: data are normal) is accepted.

In the linearity test, the hypothesis states that if the Sig value from the deviation from linearity is greater than 0.05, the relationship between the independent and dependent variables is considered linear (H₀: linear relationship). Conversely, if the Sig value is less than 0.05, it indicates a deviation suggesting a non-linear relationship (H₁: non-linear relationship). In this study, the Event Marketing variable obtained a Sig value of 0.842, the Promotion variable 0.654, and the Brand Image variable 0.748. Since all values are greater

than 0.05, the null hypothesis (that the relationship is linear) is accepted, indicating that there is no deviation from the linear pattern between each variable and user loyalty.

The multicollinearity test is used to check whether there is a strong correlation among the independent variables in the regression model. High correlation among independent variables can lead to inaccurate or biased model estimates. Therefore, a good regression model should have independent variables that are not correlated or multicollinear. This test is typically performed by measuring the Variance Inflation Factor (VIF) and tolerance coefficient to assess the level of collinearity between variables. The hypothesis for this test is as follows: if the tolerance value > 0.10, no multicollinearity occurs; if the tolerance value < 0.10, multicollinearity occurs. For the tolerance test, the obtained values were Service Quality: 0.630 > 0.10, Product Quality: 0.620 > 0.10, Timeliness: 0.980 > 0.10

For the VIF hypothesis test, the following applies: if the VIF value < 10, no multicollinearity occurs; if the VIF value > 10, multicollinearity occurs. The VIF values obtained were Service Quality: 1.587 < 10, Product Quality: 1.612 < 10, Timeliness: 1.021 < 10. Since the obtained values meet the hypothesis requirements, it can be concluded that no multicollinearity exists.

The heteroscedasticity test is used to check whether the residual variance in the regression model is constant or varies. If the residual variance is constant, the model is said to experience homoscedasticity, while if it varies, the model experiences heteroscedasticity. Heteroscedasticity can reduce the accuracy of model estimates, so this test is important to perform.

The hypothesis for this test is as follows: if the Sig value > 0.05, heteroscedasticity does not occur; if the Sig value < 0.05, heteroscedasticity occurs. The obtained Sig values were Service Quality: 0.087 < 0.05, Product Quality: 0.087 < 0.05, Timeliness: 0.351 > 0.05, Since the points do not follow any pattern in the spread, it can be concluded that no heteroscedasticity is present. Since all classical assumption tests have met the requirements, the data in this study can proceed to the next test, which is multiple linear regression.

Table 2. Multiple Linear Regression Results

Based on the table above, the constant value is 8.361, with coefficients for Event Marketing (X1) at -0.153, Promotion (X2) at 0.458, and Brand Image (X3) at 0.029. Thus, the multiple linear regression model can be written as:

Y = 8.361 – 0.153X1 + 0.458X2 + 0.029X3

The interpretation of the regression result is as follows: the constant value of 8.361 indicates that when all independent variables (Event Marketing, Promotion, and Brand Image) are equal to zero, user loyalty is expected to be at the level of 8.361. The coefficient of -0.153 for the Event Marketing variable suggests that for every one-unit increase in event marketing intensity, user loyalty is expected to decrease by 0.153 units, assuming other variables remain constant. In contrast, the coefficient of 0.458 for the Promotion variable indicates that every one-unit increase in promotional activities contributes to an increase in user loyalty by 0.458 units. The very small coefficient of 0.029 for the Brand Image variable suggests that its effect on user loyalty is not significant in this model.

Overall, the model shows that promotional activities are the key factor in increasing user loyalty, while event marketing tends to have a negative impact, and brand image does not contribute significantly. Additionally, the significance value (Sig) for the Brand Image variable is 0.707, indicating that there is no statistically significant effect.

For the t-test, the test hypothesis states that if the significance value (Sig) is less than 0.05, then the independent variable has a significant effect on user loyalty; conversely, if the Sig value is greater than 0.05, the effect is not significant. Based on the results, the Event Marketing variable has a Sig value of 0.035 with a coefficient of -0.153, indicating a significant negative effect on user loyalty. The Promotion variable shows a highly significant value (p < 0.001) with a coefficient of 0.458, meaning it has a significant positive effect on loyalty. Meanwhile, the Brand Image variable has a Sig value of 0.707, indicating no significant effect.

The F-test hypothesis states that if the significance value is less than 0.05, the independent variables namely Event Marketing, Promotion, and Brand Image simultaneously have a significant effect on user loyalty; conversely, if the significance value is greater than 0.05, their combined effect is not significant. Based on the analysis results, the significance value obtained is well below the 0.05 threshold, indicating that the combined effect of these three variables has a significant impact on user loyalty. This means that when these independent variables are included in the model, together they are able to explain the variation in user loyalty significantly.

The coefficient of determination (R²) indicates the proportion of variance in the dependent variable (user loyalty) that can be explained by the independent variables in the model. An R² value of 0.201 means that the model consisting of Event Marketing, Promotion, and Brand Image explains approximately 20.1% of the variation in user loyalty. In other words, considering only these variables, the model accounts for just 20.1% of the differences in loyalty among users. The remaining 79.9% of the variation suggests the presence of other factors beyond these three variables, or random variability, that also influence user loyalty. This indicates that although the variables in the model contribute significantly in combination, the model does not fully capture the entire dynamics affecting user loyalty. Other factors such as user experience, price, service quality, and external influences may also play crucial roles and should be considered in future research.

The research findings indicate that event marketing has a significant negative effect on user loyalty on Bukalapak. Although event marketing is generally regarded as a tool capable of creating memorable experiences and enhancing consumer engagement (Setiawan et al., 2022), this study’s results suggest that an increase in event marketing intensity is associated with a decline in user loyalty. This aligns with the perspective of Nurochman & Suryawardani (2020), who also noted an influence of event marketing on loyalty, implying that not all event marketing efforts translate positively in this context.

Promotion shows a significant positive contribution to user loyalty, as evidenced by a regression coefficient of 0.458. This result confirms that increased promotional activities such as advertising strategies, personal selling, sales promotions, and public relations effectively boost consumer trust and purchase interest. These findings are consistent with Farisi & Siregar (2020), who emphasized the impact of promotion on consumer loyalty.

This study found that the influence of brand image on user loyalty is not significant. This suggests that, within the Bukalapak context, the brand image may not yet be strong or consistent enough to directly drive loyalty. Therefore, other factors like promotion play a more dominant role. This outcome opens opportunities for further research to explore additional external factors that might have a greater impact on determining user loyalty in the highly competitive e-commerce market.

Based on the findings, e-commerce platforms, especially Bukalapak, are advised to optimize promotional strategies as the strongest driver of loyalty, while reevaluating event marketing approaches that may unintentionally reduce user commitment. Strengthening brand image requires consistent communication, trust-building initiatives, and improved user experiences that extend beyond price-based promotions. Future strategies should incorporate personalized marketing, gamification, and loyalty reward systems to enhance engagement. Additionally, collaboration with data-driven tools such as AI-powered recommendation engines and big data analytics can help platforms identify customer preferences more accurately, thereby improving retention. For long-term sustainability, companies should not rely solely on aggressive discounting but invest in building emotional connections with users through transparency, customer service excellence, and community-driven initiatives.