The purpose of this study was to determine whether brand ambassadors and product variations have a direct influence on buying interest, and to determine the effect of brand ambassadors and product variations on buying interest through Brand Image mediation. The research approach used is quantitative research. The type of data used in this study is primary data, namely research data obtained directly from original sources (respondents). The population in this study are all consumers who have and have not used Scarlett Whitening products. From the calculation using the formula (hair, 2020) recommends that the minimum sample size is 5 times the number of indicator items contained in the questionnaire, the indicators in this study consisting of 3 independent variables and 1 dependent variable totaled 23 indicators, so the sample size of this study was determined to be 115 respondents. The research data was processed using the smartPLS 4.0 analysis tool and the research method in this study used Structural Equation Modeling (SEM), namely an equation model with a variance-based approach or component-based structural equation modeling. The results of this study indicate that the first and third hypotheses are not supported because, Brand ambassador (X1) and Product Variety (X2) have no significant effect on Purchase Intention (Y), the second and fourth hypotheses are supported because, Brand ambassador (X1) and Product Variety (X2) have a significant effect on Brand Image (M), the fifth hypothesis is supported because, Brand Image (M) has a significant positive effect on Purchase Interest (Y), and it can be concluded that Brand ambassadors (X1) and Product Variety (X2) cannot affect Purchase Interest (Y) directly, but Brand ambassadors (X1) and Product Variety (X2) can affect Purchase Interest (Y) by going through Brand Image (M) as a mediating variable.