The Dawn of Autonomous AI. Introducing Large-Action Models (LAMs) into Short-Term Rentals

Why Short-Term Rental Managers need to address the challenges and the opportunities. 

Recently, there’s been significant excitement about how Large Language Models (LLMs) (think ChatGPT) and conversational AI (think chatbots) are revolutionising marketing workflows. Their capabilities in creating content, visuals (think MidJourney), and web designs represent a significant advancement in the field. Yet, implementing these technologies in real-world scenarios is more complex. Integrating the outputs of LLMs into a full-scale marketing strategy, such as developing and executing a campaign, still requires considerable manual intervention. Even advanced generative AI systems can automate only certain aspects of these processes.

 

In short-term rentals and guest-facing activities,  LLMs are promoted heavily, with the tech companies becoming “AI-enabled”, which generally means content, imagery and chatbots. The acronym AI is probably overused and underdeveloped in the rental industry due to its fragmentation of tools and lack of data consolidation—a fishing reel, bird nest scenario for most managers.

 

It’s not very digital or easy to unravel. Data is everywhere but is hard to corral, let alone analyse. Data, not cash, is king these days, and companies want yours. We explore LAMs below and add some caution to your technology decision direction!

We anticipate marketing AI agents adopting a more comprehensive, LAM-style strategy. These agents will leverage LLM interfaces to seamlessly integrate data, tools, and specialised agents, effectively managing complex marketing tasks. The key to this, however, is access to all data, not just specific elements. Building a two-story house is hard if you only have plans for the ground floor.

The current era of generative AI is rapidly evolving, moving beyond aiding in writing emails or blogs and creating visuals to more autonomous functionalities into emerging Large-Action Models (LAMs). These agents are not merely responsive; they are proactive and capable of executing tasks independently in pursuit of specific goals. This transition marks a significant leap in AI development, positioning generative AI as an active partner in real-time work processes.

These systems are, of course, data-dependent and learning all the time. Therefore, all data access is vital, not just bite-size chunks. Taking the fishing line analogy further, the data stream needs to be focused and] aligned and comprehensive, especially in travel, where many factors come into play.

Revolutionizing Daily Life and Business: The Power of LAMs

With cost constraints and extended working hours, intelligent automation can add margin and reduce costs.

LAMs are envisioned to revolutionise individual and organisational tasks by taking over repetitive and mundane activities. This shift allows humans to focus on more meaningful, high-value work. For individuals, imagine a LAM as a sophisticated personal assistant capable of handling complex tasks with human-like foresight. For organisations, LAMs could streamline operations, enhance customer interactions, and drive efficiency at every level.

Unleashing Personal Efficiency: LAMs as Ultimate Holiday Planners

LAMs can redefine personal assistance, making extensive research services accessible to all. For instance, planning a holiday can be overwhelming, but a LAM can simplify this process. Just by specifying preferences like destination, budget, and activities, a LAM can scan through travel sites, curate holiday packages, communicate with service providers, and even initiate bookings. This process demonstrates LAMs’ potential in personalising services and handling complex decision-making effortlessly.

In the holiday package world rental world,  guests may have their LAMs to do all the research work and intermediary communication:-

1. Your LAM would begin by conducting thorough online research, considering your holiday needs, preferences, and ideas. List key factors such as destination, budget, activities, and travel dates.
2. Create a digital note of all the details of its research, organising them along with any relevant links or information gathered from various travel websites and forums.
3. After its initial research, review and summarise the key findings. Highlight the most attractive holiday packages, destinations, and deals that align with your preferences.
4. Use its summarised research to plan the next steps. This information might include drafting a checklist or itinerary, noting additional information needed, such as visas or vaccinations or identifying travel blogs or resources for more in-depth information about chosen destinations.
5. The LAM can set reminders or create a to-do list for booking flights, accommodation, or activities. Ensure all necessary actions are outlined and ready for execution.
6. Consider reviewing your travel plans periodically to refine them. This activity could involve looking for updated travel deals, checking for new activities or attractions at your destination, or reassessing your itinerary.
7. The LAM could be vigilant about changes or updates in travel advisories, airline policies, weather, civil troubles or hotel booking conditions. Stay informed to adapt your plans as needed, ensuring a smooth and enjoyable holiday experience.

LAM to LAM: Travel agents are personalised services with specialities of destination experiences and will, as in most human interactions, have biases. A LAM can extend its reach and be more autonomous and data-driven. LAMS are not time-sensitive, need a coffee, or have interruptions in the workflow when it’s time for bed!

However, the LAM will likely be most efficient when it deals with another LAM. This raises many questions about the influence and independence of LAMs.

Redefining business efficiency: LAMs transforming corporate and consumer landscapes.

For example, when you want your next Airbnb investment property!

LAMs promise to transform various aspects of business, from backend operations to customer-facing roles. An example is researching loans or mortgages for your next Airbnb investment, where a LAM assists in managing interactions. It could transcribe calls, summarise critical points, suggest follow-up actions, and even draft communications with lenders. 

This application illustrates how LAMs can augment human capabilities, enhancing productivity and strategic decision-making. LAMs should identify themselves as AI to ensure transparent interaction with humans. Despite this, it would engage in communication that is coherent, smooth, and human-like, incorporating polite greetings, well-structured sentences, and delivering each message with a distinct request or statement. 

Taking it to the next level, the LAM could facilitate financial aspects by informing the user’s bank about potential loan requirements. As the discussion progresses towards critical decision points, the user would be brought into the conversation for their final consent and approval.

Short-term rental managers? LAMs will permeate the industry.

As you have already witnessed, this industry has become mainstream in a few years. First, we had marketplace rollups, then Airbnb, and then an explosion of tech suppliers with easy entry and hard-to-scale. As investors eye the market opportunities, we have seen mass investment, IPOs, SPACS, and management company rollups. We also had COVID-19 and a couple of economic crises. The problems are legion, but technology can assist in addressing a manager’s competitive challenges.

The net effect of this growth is that guests are more demanding, oversupply is often the norm, and legislative authorities create restrictions and politicise opportunities. How is this related to AI, LLMs and, more importantly, LAMs?  If you are in a destination that is clear to grow, then the focus for all OTAs will be on these destinations. LAMS will become more prevalent, with accelerated intelligent marketing tools using large datasets to target your travellers.

More control than an OTA!

This is where managers have more control than OTAs as a manager has all the relevant information on returning guests, details of the property they stayed in, the issues or not, the group makeup, often all emails, sometimes ages, home addresses, previous questions asked, prices paid, discounts given, bedding arrangements, genuine reviews (not just OTA ones) and more. Top-of-funnel guest delivery will engage LAMs for marketing and work closely with “intelligent websites”. 

Reactive LAMs

As seen above, task automation can quickly become a potential guest tool of choice, which means a business opportunity. At the extreme end of this opportunity, however, it could overpower businesses that must answer daily questions from travellers, provide links to deals, and provide detailed information on properties, formats, extras, etc. Humans are not 24/7, 365 work machines, although sometimes it may feel like it.

Even a simple question on a property would be best answered with more knowledge about the enquirer. A chatbot with an FAQ needs to be more. Have they visited previously? What was the party makeup, what sort of review did they leave, and what are they interested in? Has the property performance been good? Is this an opportunity to get occupancy? What does the portfolio performance show overall in comparison? What are the owners’ preferences? Is there a special event during that period, and much more?

Assume your business has its assistant LAM, which can help counter the enquiry LAM. However, it can only do this if it has the full spread of business data!  This task will be challenging in a fragmented industry, such as vacation rentals, with multiple tech suppliers offering their own AI models. 

Proactive LAMs

Marketing is the lifeblood of any business. Many STR businesses rely heavily on OTAs who are well ahead in AI. A LAM can automate many of your marketing activities “intelligently” using a vast array of your data, write content, monitor responses, follow up, create images, etc. It can select the previous customers to target based on stays, reviews, demographics and many other aspects. The LAM can become your marketing assistant and use third-party services for distribution, advertising, and more.

AI can assist with web design, improve conversions, optimise content and increase direct booking opportunities. Websites need to enhance conversion and information delivery, which is becoming more accessible to facilitate as AI develops and can change web elements seamlessly to improve conversions!

Why is this subject super important?

We have been preaching the total guest automation research and booking experience for several years. AI, LLMs, LAMs, etc, are now facilitating this. A LAM will, for some time, give the final booking decision to the guest. Still, all the pre-work is undertaken, and even post-decision payments and organisation are complete.

If you want to see a snapshot of where the deliverables and rapid creep toward LAMs in travel, check out Rabbit-R1, an independent handheld device. This is just the beginning, and your smartphones will undoubtedly align with this type of activity soon.

The world of accommodation is pivoting. Based on search and money, global OTA corporations have increased their opportunities, developing into apps and brands. However, they still require supply and try to pigeonhole these into simple categories. However, these are not hotels, and each accommodation is unique in some way that plays to suppliers’ strengths in a LAM world. LAMs can work 24/7, 365 and have the internet at their fingertips.  Despite their size and money, OTA corporations need help in a fully integrated LAM environment: they need comprehensive supplier information on properties, prices, deals, previous guest feedback, stay issues and local activities and information. Suppliers can benefit from this:

Price and Information: a guest LAM with Google’s capacity to identify direct booking as part of its instructions and communication protocols is a case in point. Add in “Google Pay” for trust, and hey presto, increased direct bookings!  

This change in corporate direction does mean the need for great converting informative websites; however, LAMs increase the opportunity for information sharing, speed of communication, and direct booking.

Your Tech Stack is the facilitator for the future success of your business.

Until recently, we have witnessed a significant increase in tech tools for the STR industry. These can be categorised into:

• “Source of Truth” tools such as a PMS
• Dynamic pricing and revenue management tools 
• Operational activity tools 
• Guest apps
• Owner apps
• Payment solutions 
• CRM and marketing tools 
• Websites 
• Distribution channels
• Review management
• Eco and other audit tools
• Social media tools
• Data services

Everyone is now adding AI as either a marketing play or a functional instrument to these categories.

The Problem: There are 13 types of tools listed above that will have 13 AI systems, but a proficient LAM needs integrated and pan-data access across the whole system from booking to re-marketing and, very importantly, reporting on all activities at the management level and with reactive actions at an operational level.

The future looks to be a consolidated tech platform that can access, combine and interpret data intelligently with productive outcomes.

What does your technology stack look like?

Adopting best-in-class via API connections would have been the best option three years ago. The market is changing, and perhaps that is why investors continue to seek “Source of Truth” companies that collate the initial booking information. These businesses (mostly PMSs) are not yet at a level of LAM AI competence or have the vast array of tools that allow a single AI model or company to work effectively. 

Adopting a best-in-class service, e.g., a website or comms with a real AI interpretative function, is a good step. Embracing a system where all functions are integrated would be the ultimate goal. Cross-supplier and brand data sharing and cross-platform AI will no doubt see commercial agreements, but this comes with many challenges in data privacy, costs, price and control.

A word of caution

Control of and access to extensive data is the gold rush of the 2000s. Sharing data helps everyone take snapshots of the market and competitors, but you need to control all your data in a composite fashion. This allows you to perform a siloed and pan-business AI assessment and a proactive approach based on complete information. 

 

There is little point using tech tools where a third party AI monetises your data, albeit anonymised, and re-charges you. In addition bolting together multiple systems and expecting a refined, efficient and productive AI enhanced business would be akin to building a car from many different manufacturers. All of whom have different designs, metrics,  shapes and sizes.  It won’t be the Ferarri you’ve been promised!

 

 

Choose a tech partner with cross-service integration tools and single embedded AI for the best results.

 

Conclusion: Embarking on a New Era with LAMs at the Forefront

The potential of LAMs is vast, marking a new chapter in AI evolution. With a focus on human empowerment, LAMs are set to enhance productivity, simplify complex tasks, and revolutionise personal and professional domains. In fragmented, multi-disciplinary environments with many stakeholders and moving parts, such as short-term rentals, adopting LAMs will be an advantage. Those who can use LAMs working from complete data sets will have the edge.

 

 

The image above is from MidJourney. The author is also a board advisor to BoomNow and select management companies, as well as a short-term rental technology business consultant. Yes.Consulting

If you have questions or want to “chew the fat” on your tech decisions, please email me or hook up on Linkedin!