By Adam and Larry Mogelonsky
Chatbots have been around for a while, but ChatGPT has brought them forth into the spotlight, while large language models (LLMs) now present a host of different utilities for evolving both how various hotel systems talk to one another as well as how guests book.
The important term here to remember and break out at your next cocktail party is ‘technological determinism’. Because chatbots are all the rage right now, this will propel more travelers to start to try them out for planning their next trip, which in turn due to recursive learning will help to augment the accuracy and usefulness of these new platforms to draw in more users. It’s an exponential growth curve where we’ve yet to reach the elbow.
First Some History
Even considering any new technology, a good exercise is to look back at what came before in order to better evaluate the boundaries for what’s possible in the near future. For this, think about how you discovered your chosen hotel before and after OTAs like Booking, Expedia, Priceline and Agoda became household names with extensive, worldwide distribution. Travel guides, travel magazines and travel agents likely all had a much stronger influence on where you decided to go to back in the day. As the OTAs gained network effects, brands with a smaller footprint suddenly gained a new channel through this to frictionlessly sell inventory, while travelers benefited from a portal through which to conveniently discover new products in new destinations.
If you think of the OTAs as a simple line market, by bringing buyers and sellers closer together, they worked to expand the total volume of hotel inventory sales. Concurrently, TripAdvisor and other review websites came to provide a baseline of trust for consumers in a positive feedback loop – that is, more reviews have meant more trust in the platform and more trust in hotels with lots of good reviews, thereby incentivizing more travel to new hotels and more reviews on the platform.
Next, consider the sharing economy. Airbnb and its ilk have further propelled global travel numbers to astonishing new heights starting in the 2010s by incentivizing growth within the vacation rental market. The response from traditional hotels has largely been more hotel brands to serve different customer niches, more technology deployment, better amenities and a shrewder focus on quality service. Similarly, one could point to the growth of social media travel influencers and the opening up of previously unknown destinations as well as a new channel through which to reach lifestyle consumers.
The point throughout is that technological innovations that attain global scale inevitably reorient consumer behavior, with a few common threads from all development to keep in mind:
- More hotel product diversification and segmentation
- More destination access, distribution and travel volume
- More expectations for enhanced service and flexibility
As further fuel to the fire, both Expedia and Booking Holdings have plans to incorporate ChatGPT into their platforms as an additional incentive for customers to stay loyal to the OTAs versus opt for the hotel’s direct booking incentives. This puts the ball in our side of the court to both improve the interfaces by which we facilitate direct bookings as well as improve our offers to encourage these channels.
Further Democratizing Travel Products
With these broad strokes as a framework, it’s far easier to predict how the hospitality-specific chatbots and robotic process automation (RPA) tools that are leveraging machine learning and generative AI will change the ways that travelers discover new destinations and book their accommodations as well as how hotel guests will want to purchase additional experiences from their host hotels.
Newer hotel systems developed within the past decade have been able to use advances in cloud-hosting services and APIs to outcompete legacy software by offering nimbler platforms at drastically cheaper prices. We often label this progression as ‘democratization’ because the lower costs enable ever-smaller businesses to reap the benefits from using the technology.
To start, ChatGPT-esque clones will become synonymous with other elements of a hotel website like the top navigation and booking engine, so much so that such a feature will move from value-add to customer expectation, where eventually more bookings may come through the chat thread than through a search initiated through the classical booking engine. Just as user-friendly website development tools like WordPress and Squarespace democratized website creation, these chatbots will only further enable small hotels or other tourism operators to maintain a great digital storefront without the high costs of human-based customer service.
From there, it isn’t hard to imagine a curated AI built with OpenAI’s GPT being used, not just for basic guest inquiries and facilitating bookings via interfaces, but to natively power direct reservations and back-of-house workflows for properties of under, say, 10 rooms, supplanting the need for a PMS, booking engine and ops platform. You may be skeptical that such a leap will take place sooner rather than later, and yet a recent example of this is Jurny’s announcement of an AI-enhanced PMS powered by GPT-4.
Further, diminishing tech costs without any major sacrifices to service, such systems might empower more small hotels and vacation rentals in little-known destinations where the economics currently don’t work. Moreover, the tech cost savings might be redeployed to finance other more capital-intensive operations, enabling additional product nichification to serve emerging segments like digital nomad, glamping-adjacent or wellness-primary travelers.
Over-Tourism Breeds Travel Redistribution
With an increasingly vast network of travel products and various incentives for people to go abroad, we now have enough macroeconomic data on this self-perpetuating cycle to see how technology has accelerated the modern problem of over-tourism, with knock-on effects being unaffordable real estate, disruptions to local culture and contributions to global warming.
Despite the damage that has been brought by over-tourism, under-tourism is a far worse problem because of how it can impair a destination’s economy, diminish trade and increase crime. By democratizing access to technology for SMB operators in second-tier cities, rural areas and developing countries, AI can enable tourism redistribution and thus act as a solution to this challenge.
These sorts of changes may seem to occur at a glacial pace and not necessarily affect your hotel. And yet, they still do. The easier it is for someone to shift a house into a viable vacation rental business, the more likely it is to happen. And while the growth of travel over the past two decades as supported by technology implies an ever-expanding total addressable market for worldwide travel, this says nothing about the products that get left behind because they didn’t adapt fast enough and got caught in a death spiral of flatline revenues and no capital leftover for a brand refresh.
Amongst financial circles, we often say and have heard is, “The hotel industry isn’t overbuilt; it’s under-demolished.” To close, here’s a broad, sequential list of what to investigate with regards to AI so that you don’t end up on the latter end of this adage:
- Start with a chatbot or enhanced AI assistant to facilitate basic inquiries on the website
- Expand this chatbot to handle on-prem guest requests via SMS or within the branded app
- Deploy a conversational AI to do the same for the voice channel and replace the IVR
- Investigate how RPA can further reduce manual work to free up the team’s time for other tasks
- Recruit a semantic analysis engine to guide operations from prearrival and post-stay remarks
- Connect the chatbot and conversational AI so that they can handle reservations on the website
- Connect both so they can handle ancillary bookings for restaurants, spas and activities
- Use an AI tool for dynamic pricing, channel management, upselling and attribute-based sales
- Roll out an AI-powered platform for loyalty and one-to-one return stay offers
- Look out for new systems that are using AI to do all this under a single, unified interface