According to the data, Tokopedia has faced a consistent drop in consumer buying interest over the last three years.

To explore the factors that influence consumer purchase intention, a preliminary survey was conducted among 30 Tokopedia users to identify the most influential factors driving online purchase intention on the platform. Based on the results and using the Pareto principle, the selected variables—contributing ≥80% of cumulative responses—include: e-trust, brand image, e-service quality, online customer reviews, and online customer ratings.

Based on the aforementioned background, the researcher is interested in conducting a study titled “Factors Influencing Purchase Intention on Tokopedia.” This research aligns with the 9th goal of the Sustainable Development Goals (SDGs)—Industry, Innovation, and Infrastructure—as e- commerce platforms such as Tokopedia contribute to technological and logistical innovation and support the development of digital economic infrastructure.

Consumer behavior is defined as the study of the processes involved when individuals or groups select, purchase, and use products to satisfy their needs and desires. It plays a vital role in how consumers make purchasing decisions. E-commerce

E-commerce refers to the use of online platforms for conducting buying and selling activities, including both goods and services (Jarti et al., 2023). In this era of globalization, transactions have become easier, with no limitations in terms of space or time.

Purchase intention is the consumer's willingness to buy a particular product after considering various alternatives (Selly & Purba, 2021). Consumers typically compare the strengths and weaknesses of similar products or services before making a final decision.

E-trust refers to the relationship built between sellers and potential buyers in an online environment (Lien et al., 2015). It reflects consumers’ confidence in the reliability and quality of the products offered (Sativa & Puba, 2016). Since online transactions pose higher risks compared to offline transactions—ranging from payment security to personal data protection—building e-trust is essential for fostering consumer confidence.

Brand image is the perception held by customers regarding a particular brand, formed through their experiences, trust, and emotional connections (Kotler & Armstrong, 2008). According to Tjiptono (2015), brand image is a set of associations or beliefs consumers have about a brand, influencing their buying decisions.

According to Aaker and Biel (2009), brand image is the consumer’s perception of a brand formed through personal experience or information obtained from others and the media, and it represents the collection of knowledge, beliefs, and emotions that consumers hold about the brand.

E-service quality refers to a website’s ability to deliver a smooth and efficient shopping experience (Sativa & Astuti, 2016). Saragih (2019) defines it as the website’s capacity to facilitate purchasing, order processing, and delivery in an effective and efficient manner. A high level of e-service quality helps consumers throughout the transaction process and enhances customer satisfaction.

Online customer reviews and ratings are features that allow consumers to share their opinions or feedback regarding products they have purchased (Elwada, 2016). These can be expressed in the form of written reviews or star ratings. According to Mo and Fan (2015), such reviews offer insights into product quality and help other consumers make informed decisions based on the experiences of previous buyers.

Table 3. Results of Outer Loadings Test

Table 5. Composite Reliability

Source: SmartPLS Output, 2025

A Composite Reliability value above 0.70 indicates a high level of internal consistency. Therefore, all constructs in this study meet the composite reliability criteria and are deemed highly reliable.

Convergent validity can be measured using the Average Variance Extracted (AVE) value.

Tabel 7. Fornell-Larcker

Based on the SmartPLS 4 output, the correlation between each indicator and its construct is higher than its correlation with other constructs. Therefore, it can be concluded that the model exhibits good discriminant validity.

According to Hair et al. (2022), an HTMT value of < 0.9 indicates that the model has good discriminant validity. The HTMT value in this research model is < 0.9, thus this study demonstrates good discriminant validity.

Table 9. Variance Inflation Factor (VIF)

Source: SmartPLS Output, 2025

According to Hair et al. (2022), a VIF value below 5 indicates that multicollinearity is not a concern in the structural model. In this study, all indicators meet this criterion except PI2, which has a VIF of 5.246. As a result, PI2 was removed from the model, leaving 19 indicators in total.

Structural Model Evaluation (Inner Model)

a. Coefficient of Determination (R²)

Table 9. R-Square (R²)

Source: SmartPLS Output, 2025

The R² value for E-Trust is 0.662, indicating that Brand Image, E-Service Quality, and Online Customer Review/Rating collectively explain 66.2% of the variance in E-Trust. Meanwhile, the R² for Purchase Intention is 0.542, meaning that the same independent variables explain 54.2% of the variance in Purchase Intention.

Table 10. Q-Square (Q²)

Source: SmartPLS Output, 2025

According to Hair et al. (2022), a Q² value greater than 0 indicates that the model has predictive relevance. Based on the results, both E-Trust and Purchase Intention possess satisfactory levels of predictive relevance.

Table 11. Effect Size (f²)

Source: SmartPLS Output, 2025

Effect size (f²) is interpreted based on thresholds from Hair et al. (2022), where:

In this study:

Figure 2. PLS Analysis Results

Table 12. Direct Effect Significance Test

Source: SmartPLS 4 Output, 2025

Table 13. Indirect Effect Significance Test

Source: SmartPLS 4 Output, 2025

The hypothesis testing results show that Brand Image does significantly affect Purchase Intention. These findings suggest that brand image plays a critical role in enhancing consumers’ purchase intention on Tokopedia. Accordingly, it is essential for Tokopedia to concentrate on strategies and elements that can positively shape and strengthen its brand image.

The results show that E-Service Quality has a positive and significant effect on Purchase Intention. This implies that Tokopedia must focus on continuously improving its service quality to encourage consumers to make purchases.

Consumers are more likely to buy from a platform where they feel comfortable due to good service. Future research could further explore or narrow the scope of E-Service Quality in the context of Tokopedia.

The hypothesis test found that Online Customer Review/Rating does not significantly affect Purchase Intention on Tokopedia.

This may be due to different standards among consumers in assessing products. A product rated as “good” by one user may receive a different numerical rating from another user, even with similar sentiments. Additionally, repeat customers may no longer read reviews or ratings if they consistently purchase the same product, making reviews and ratings less influential in their buying decisions.

The analysis shows that Brand Image does not significantly influence E- Trust. This could be because Tokopedia users are already familiar with and trust the platform’s brand. Therefore, the brand image does not play a substantial role in building or increasing trust for these users.

The results indicate that E-Service Quality has a positive and significant effect on E-Trust. Accordingly, e-service quality serves as a key determinant in fostering consumer trust toward Tokopedia. Hence, it is essential for Tokopedia to prioritize the enhancement of its service quality in order to build and sustain consumer trust.

The hypothesis is supported, showing that Online Customer Review/Rating has a positive and significant impact on E-Trust. The findings of this study demonstrate that online customer reviews and ratings serve as a key factor in building consumer e-trust.

The hypothesis testing results show that e-trust does significantly affect Purchase Intention. The findings of this study reveal that e-trust is a significant determinant of consumers’ purchase intention. Accordingly, Tokopedia should prioritize efforts aimed at strengthening consumer trust in its platform.

The Effect of Brand Image on Purchase Intention through E-Trust

The findings reveal that Brand Image does not significantly affect Purchase Intention via E-Trust.

Respondents, who are already familiar with Tokopedia's strong brand and trust it, are not influenced in their purchase intentions by these factors. Their purchase intentions are not mediated by E-Trust.

The hypothesis test results indicate that E-Service Quality does not significantly affect Purchase Intention through E-Trust. Respondents already trust Tokopedia and are familiar with the platform's service quality. Hence, this relationship is perceived as routine, and not a driving factor behind their purchase intention.

The results confirm that Online Customer Review/Rating has a positive and significant impact on Purchase Intention through E-Trust. This suggests that consumers find Tokopedia's reviews and ratings to be credible and useful in product evaluation. These reviews build trust, which in turn enhances the consumer’s purchase intention. Therefore, E-Trust serves as a mediating variable between Online Customer Review/Rating and Purchase Intention.

Theoretical recommendations for future research on purchase intention via Tokopedia suggest that subsequent studies should consider exploring additional independent variables that may significantly influence purchase intention within the Tokopedia platform. Given that the present study focuses broadly on the Tokopedia application, it is recommended that future researchers refine the scope by concentrating on a specific online store hosted on Tokopedia, thereby enabling a more in-depth and contextualized analysis.

Managerial Recommendations