Resumen
Products can be divided into those that meet daily needs and those that meet unknown needs. Determining patterns of purchase of products for daily needs is relatively easy, but identifying the purchase behavior of products for unknown needs is more difficult. This study was conducted to examine a community of users of an exclusive product: the Vespa motorcycle. Validity and reliability were established, and descriptive analysis showed that all the variables met the goodness-of-fit test. The obtained results showed that the assumption of a normal line and autocorrelation cannot be determined; there is heteroscedasticity and some models fit to collinearity. The F test results showed that the independent and dependent variables are interrelated. The t-test revealed several variables that proved the null hypothesis, and several hypotheses are accordingly proposed. This means that there are several variables that are not related to the dependent variables (decision variables). Direct and indirect tests using a structural equation model showed that most do not meet the goodness-of-fit test. In conclusion, most relationships had little value, whereas other relationships had greater values. It is recommended to only measure relationships that have a value above 0.50 because the purchase behavior does not reflect the demand for scarce goods as with normal goods.Keywords: product price, product promotion, product quality, brand image, equity, brand loyalty, purchase intent, rare goods, Vespa, IndonesiaJEL Classification: M31