Online business or e-commerce is now very popular and is a growing industry in Indonesia. Although the growth is high with a 15% rise from 2011, it is found that only 57% of the Indonesians used the Internet for online shopping (MasterCard, 2012). To increase purchases through online stores, a study on customer satisfaction is needed. This study designs a preferred service for fashion online shops based on customer preferences, which would then increase customer satisfaction. By analyzing service attributes that enhance satisfaction for online customers, a suitable service for fashion online shops could be generated. The Kano model is used to classify which service attributes are important for improving and developing the quality of fashion online shops. To understand customer preferences for fashion online shops, conjoint analysis is used to calculate the preferences statistically.