Trust on WhatsApp: Designing Faith‑Aware, Human‑in‑the‑Loop Hotel Recommenders for Pakistan

In Pakistan, hotel bookings often take place through informal channels like WhatsApp, phone calls, and personal referrals rather than through formal apps or platforms. Travelers prioritize family privacy, faith-aligned amenities, and payment flexibility—needs that are rarely addressed by conventional recommendation systems. This study explores how hotel recommendations can be made more trustworthy and relevant in such a setting. Using interviews, booking simulations, and chat transcript analysis across Abbottabad, Murree, and Swat, we find that travelers respond more positively to recommendations that are delivered inside familiar channels, explained in clear and culturally sensitive language, and supported by human staff. Features like cash-on-arrival options, safety cues, and the ability to override AI suggestions increase trust and booking confidence. The study proposes a conceptual framework for low-resource recommendation ecologies, emphasizing cultural congruence, human oversight, and explainability as key drivers of adoption. These findings offer practical guidance for designing hotel recommendation systems that work with, rather than against, the realities of informal travel in Pakistan.   Keywords: Hotel Recommendation Systems; WhatsApp-Native Booking; Faith-Aware Design; Human-in-the-Loop AI; Cultural Congruence; Informal Tourism; Pakistan Hospitality; Explainable Recommendations; Low-Resource Digital Systems; Trust in Travel-Technology