Boston Consulting Group takes stock of AI’s advance in hospitality

Direct bookings nearly match OTA sales

A new report from Boston Consulting Group (BCG) sets out how “AI-first” hotels are gaining in visibility, revenue and operational efficiency from trip search and pricing through to property design and construction.

According to BCG, a new generation of AI-first hotels - reimagined around artificial intelligence from the outset - is emerging: faster to develop, leaner to operate and richer in guest and employee experience. 

Produced with experts from New York University, the study identifies three major shifts: AI-optimised distribution, lower costs through automation and intelligent resource orchestration, and, further upstream, design decisions supported by generative tools.

The most visible shift is how AI is changing whether and how hotels are found by prospective guests. Generative AI is reshaping marketing with finer, more personalised targeting, but discoverability no longer depends solely on search ranking or advertising. 

It will hinge on whether a hotel’s content is machine-readable by large language models. BCG estimates that 37% of travellers now use LLMs embedded in travel platforms to plan or book. 

Without a coherent digital footprint, a property risks becoming a “digital ghost”, absent from the top three recommendations. 

This shift has consequences: in 2024, digital direct bookings almost equalled online travel agencies, at $262 billion versus $266 billion, buoyed by loyalty and personalised offers.

Once the guest is engaged, AI continues to work behind the scenes. Pricing can now be truly dynamic, adjusting in seconds to demand, weather, events, air capacity or sentiment on social media. 

The report cites AI optimisers that have generated more than 15% additional revenue per available room. 

On the ground, chatbots and multilingual concierges handle a growing share of requests, while automation and the Internet of Things (IoT) streamline inventory, energy consumption and the allocation of human and material resources.

The study highlights striking use cases, including a 20% acceleration in room cleaning at The Ritz-Carlton, San Francisco (through optimised staff allocation), predictive models at IHG, and an approximate 50% reduction in food waste at Four Seasons Peninsula Papagayo.

AI can also intervene earlier, at the design stage. Generative tools enable multiple layout options to be created and tested, anticipating people flows, energy performance (including sunlight exposure) and carbon footprint. 

This foresight can extend further with geo‑intelligence platforms that assess flows, demographics and accessibility to select sites and accelerate returns.

Impressive though AI’s optimisation potential may be, one figure in the study stands out: fewer than 10% of hospitality professionals currently approach their activity with an AI-centred mindset, with many others experimenting piecemeal to convince themselves of AI’s benefits.

To harness AI across all aspects of a hotel’s operations, the business needs to be rethought end to end, rather than by adding “AI tools” to a legacy stack. 

It is precisely this inheritance of standalone systems (PMS, CRM, F&B or sales management) and fragmented data that BCG identifies as barriers to unlocking AI’s potential. 

The data must be freed from silos and the operation reconsidered as a whole. The required investment, the report notes, is often less appealing than a visible renovation with a more immediate, demonstrable return.

BCG cautions hoteliers that this shift in philosophy needs to happen quickly. The window of opportunity to catch up will not remain open for long.