Hilton's AI Trip Planner Shows Early Traction, But LLM Costs Present Wider Challenge
Hilton's conversational AI trip planning tool is generating "slightly higher" conversion rates three months after launch, with the hotel group preparing to integrate it into its Honors loyalty app this month. The tool lets travelers describe trips using natural language rather than selecting a city and dates from dropdown menus.
Michael Leidinger, Hilton's chief information officer, said the first version of the planner had a critical flaw: it would accept one prompt, then immediately ask for travel dates. The team iterated to improve the experience.
The Tokenomics Problem
Hilton faces a separate challenge that extends beyond this single tool. The cost of running large language models across the company-what Leidinger called "tokenomics"-has become a pressing concern as the hotel group expands AI use.
This reflects a broader issue in hospitality: the expenses associated with generative AI and LLM systems can accumulate quickly when deployed across multiple customer-facing and internal applications. Hotels using AI at scale must weigh conversion gains against infrastructure costs.
Employee Education Remains Difficult
Hilton has also grappled with educating staff on when and how to use AI appropriately. Getting teams to understand the right use cases-and the boundaries-requires ongoing effort beyond technical implementation.
For hospitality leaders, the Hilton case illustrates a dual reality: AI tools can improve guest experience and booking metrics, but managing the operational costs and organizational change requires sustained attention.
Learn more about AI for Hospitality & Events and how other organizations are addressing similar implementation challenges.
Your membership also unlocks:
