To test our hypothesis, a chi square test of proportions is conducted as customers and subscribers are two unpaired categorical variables, and there are only two groups. The ANCOVA test could be an option for testing categorical independent variables like weekdays and weekends, but it was not adopted as the dependent variables (customers and subscribers) are also categorical variables and not continuous. Therefore, the chi square test is most appropriate. 
To conduct the chi square test, we got the proportions of the weekday and weekend trips per user type and created the contingency table of values.