This Thai luxury hotel group operated 14 properties across Bangkok, Phuket, Chiang Mai, and Koh Samui, with a combined inventory of 3,200 rooms ranging from five-star business hotels to boutique resort properties. Revenue management was handled by a centralized team of six analysts using a combination of spreadsheet models and a legacy revenue management system that relied on historical occupancy averages and basic seasonal adjustments.
Revenue per available room (RevPAR) across the group had stagnated at THB 4,800, while competitive set properties using AI-driven dynamic pricing were achieving RevPAR of THB 5,700 or more. The group's pricing was consistently either too high (resulting in low occupancy during shoulder seasons) or too low (leaving money on the table during peak demand). The revenue team adjusted rates only twice per week — on Mondays and Thursdays — while competitors were adjusting prices multiple times per day based on real-time demand signals.
Thailand's tourism market added complexity: demand was influenced by an unusually wide range of factors including Chinese holiday calendars, European winter season bookings, Thai festival dates, political events, airline capacity changes, and weather patterns that varied dramatically between the Gulf coast and Andaman coast properties. The existing system could not integrate these diverse demand signals.
Pertama Partners' AI Readiness Audit analyzed the group's historical booking data spanning five years, including rate performance, booking lead times, cancellation patterns, channel mix, guest nationality profiles, and competitive pricing data. We identified that the group was systematically underpricing during the 60-to-90-day booking window when business travelers and group bookings were price-inelastic, while overpricing during the 7-to-14-day window when leisure travelers were highly price-sensitive.
Our AI Pilot Program deployed a dynamic pricing engine at four properties representing different market segments. The system incorporated 43 demand signals including real-time flight search data for Thai destinations, competitor pricing scraped from OTA platforms, local event calendars, weather forecasts, social media sentiment about Thai tourism, and the group's own booking pace data. The model generated optimal room rates by room type and channel, updated every four hours, with a revenue manager approval workflow for changes exceeding defined thresholds.
Executive Training helped the group's general managers understand the principles behind AI-driven pricing and trust the system's sometimes counterintuitive recommendations — such as raising rates during a period of seemingly low demand because leading indicators showed a surge was imminent. Team Training equipped the revenue management team to monitor, interpret, and override the AI system when their hospitality expertise warranted it.
"Our revenue managers have deep hospitality instincts, but they could not process 43 demand signals simultaneously. Pertama Partners built a system that thinks as fast as the market moves while respecting the art of hospitality pricing."— Pimchanok Wongsawat, Group Director of Revenue
Every transformation starts with a conversation. Let's discuss your challenges and opportunities.
Discuss Your Challenge