The following is a guest post from Kevin Cochrane, chief marketing officer at cloud infrastructure company Vultr. Opinions are the author’s own. 

At the start of what should be a busy travel season, hotels are contending with rising operating costs, compounded by increasingly unstable demand as geopolitical tensions impact airfare and fuel prices; food and beverage supply chains; and traveler safety. 

Occupancy alone is not enough to ensure profitability. The hospitality sector is turning to technology, especially artificial intelligence, to capture guest ROI and create a cushion against uncertainty. However, these systems require high-performing IT infrastructure to operate, and many hotels lack the necessary support to run them. 

A joint report from McKinsey & Co. and Skift Research discovered that hotels are behind the curve in AI adoption and maturity compared to other industries. Even so, leaders remain enthusiastic about its potential: 90% of travel executives are already using generative AI in their businesses, and 80% plan to scale automation with agentic AI in the next three to five years. 

Facing unprecedented operational challenges, hotels can’t wait five years for advanced solutions — they need them now. By upgrading to stronger, more resource-efficient infrastructure, hotels can power the AI initiatives that will protect their profitability today.

Core capabilities

Volume is not the answer. Instead, hotels gain the most from AI when they focus on maximizing value at the margins. As supply continues to outpace demand, hotels must extract as much as possible from every booking while lowering the labor and operations cost per occupancy. 

The following use cases are already emerging at major chains across the globe: 

• Improving demand forecasting: AI tools allow hotels to update forecasting and pricing models in line with current demand influencers and guest patterns. AI can automatically trigger actions to improve data depth and quality, such as reminding guests to update their loyalty card information. 
• Labor optimization: AI systems can incorporate more accurate demand forecasts into weekly and monthly labor plans, preventing costly overstaffing while ensuring enough employees per shift to deliver the premium of service. 
• Boosting profit per room: AI-enabled insights, such as scenario modeling, can lead to more precise, strategic room pricing. Compounded across properties, these gains add up to significant growth. 
• Streamlined guest services: AI-powered interfaces for routine guest queries give staff more time to focus on deep customer engagement. 
• Expanded personalization: Advanced AI unlocks more personalized stays. For example, a routine business traveler staying at multiple properties might have their favorite room service order queued up when they arrive, or their thermostat already set to their preferred temperature.

Start with the data

Hotel guest data is highly fragmented. Many guests use third-party travel services to get lower rates and bundles, which can create discrepancies among booking records. Meanwhile, hotel loyalty programs containing guest data may be outdated if not updated manually. More consumers are also booking trips using personal AI agents, reports CNBC, which could make it harder for hotels to capture traveler data. 

Without proper visibility into their data, hotels may miss out on patterns that could lead to a better guest experience and increased revenue. This also creates blind spots in predicting demand and potential cancellations. 

Hotels need a stronger IT framework to bring siloed and inconsistent data together. However, legacy infrastructure cannot support the data demands of AI at the organizational scale. Single-cloud models without a proper software stack are often plagued by insufficient storage; a lack of visibility across properties; and inadequate compute power, rendering them unable to handle the data volume and workload intensity needed to run AI models effectively. 

Modernizing infrastructure

Hotels need high-performance computing, but they cannot afford unpredictable costs or inflexible vendor dependencies — especially not at a time when demand is so unstable. AI costs are also rising, placing further strain on IT budgets. Many popular model providers are raising their prices or limiting accessibility due to a shortage of necessary hardware. 

To remain competitive without obliterating their bottom lines, businesses must invest in flexible, efficient infrastructure. This likely requires shifting away from a single-provider cloud architecture — as is common with legacy systems — to a multicloud, software-supported ecosystem. With this approach, they can compose a scalable, flexible compute framework that synchronizes data and AI model management across all properties.

A mutli-cloud strategy also enables hotels to bring their IT operations closer to their guests — a boon for chains operating across national borders. Running AI at the edge reduces latency, ensuring faster insights and real-time updates to demand changes while improving the performance of guest-facing services. 

This strategy also ensures compliance. Hotels already collect guests’ personal identification and financial data. When personalization expands, so does the potentially identifying data on file. As a result, hotels may be subject to new laws around data privacy and digital sovereignty. A single provider cannot always guarantee sovereign infrastructure, or that the data housed within won’t be duplicated across jurisdictions without authorization. Hosting data within borders on compliant infrastructure prevents costly violations while improving the quality and integrity of service.