This study aims to assist decision makers to improve hotel service performance by developing strategic orientations. Strategic orientations are imprecise and subjective: they vary according to domain experts’ opinions and practical experience. This study develops an integrated rough multiple attribute group decision-making (MAGDM) model that supports decision makers to acquire a systematic view of strategic methods for improving performance to the desired level. The model incorporates three methods. First, we use the rough number approach to determine group opinions from the practical experience of domain experts. We then apply the rough decision making trial and evaluation laboratory (DEMATEL)-based analytic network process to develop a rough influential network relation map (RINRM). The rough influential weights based on strategic orientations and their corresponding attributes are calculated. Finally, we use the modified complex proportional assessment of alternatives with gray relations (mCOPRAS-G) and the influential weights to calculate the relationship ratio between performance and aspiration level. Our model flexibly assesses the vagueness involved in decision-making, enabling more objective expert estimation of attributes in subjective surroundings than traditional multiple attributes decision-making models. We use the Taiwanese hotel industry as an empirical case study. Our study provides useful information on how to enhance hotel services and achieve optimal performance based upon expert judgments and the RINRM. The results show that customer orientation is the optimal strategy for hotel service promotion.