The future of AI in RMS and the hospitality industry

Revenue Management Systems have become one of the most vital tools for hotel revenue departments because of how much they can facilitate data management, allowing for better decision making. Now, thanks to the imminent advancement of Artificial Intelligence (AI), RMSs can enhance their usefulness and data management capabilities. This week we talk about the future of AI in RMS with Rafael Gómez: Director of Revenue Management at Minor Hotels & Resorts.

Welcome, Rafael. To begin, please tell us who you are and what you are passionate about.

I am from Bilbao, Technician in Business and Tourism Activities at the University of Deusto, Canary Islander by adoption and a person with many concerns. Every day I try to learn something new. I can't conceive life without applying a certain passion in everything we do.

In my work, in the hotel sector, I live intensely the world of Revenue Management, where I started in 2001 at the hand of the American multinational Starwood Hotels & Resorts, which was then leading the change. They had a very innovative and attractive learning program with very innovative strategies that are still being applied today. In this sense, I have collaborated at a regional level in projects for Resorts. In addition, and at a private level, I founded in 2019 www.revenueresort.com, a web oriented to training and opinion that embodies my professional concerns.

I also collaborate with opinion articles in the magazine TecnoHotel and I am part of the faculty of the MBA Master of the Canary Islands Institute of Tourism. And of course, all this combined with my duties as Director of Revenue Management at Minor Hotels & Resorts. Although it seems that everything has been invented, the truth is that we do not stop evolving in this world. I always say that we keep pedaling because the day we stop pedaling we will simply fall.

Let's talk about Revenue Management Systems (RMS), what are they and how do they work?

An RMS is the heart of any hotel from where many decisions are made. To try to define what it does, let's imagine all the raw data that exists in each booking transaction, which we call "INPUT", for example: date of arrival and departure, number of people, segment, room type, date of booking or advance booking, day of the week, price, conditions, F&B spend, competitors' price, etc. An infinity of data that we could never sort by ourselves and that, however, an RMS compiles by transforming all these "INPUTs", returning us ordered "OUTPUTs", such as demand forecast per day, per room type, segments, price recommendations based on a competitive group, etc.

What are the advantages of RMS for hotel revenue management?

They are fundamental. When we talk about Revenue Management we always talk about data that are essential to make decisions and establish strategies. That is why we need a system that organizes all this information. As with any type of application, we need to make our lives easier. It would be unthinkable today to perform our tasks efficiently without correct data. Among the advantages I would like to highlight:

Save time.
More realistic forecast.
To be able to optimize demand and prices.
Anticipate strategies to help generate income.
Create reports on demand.
Anticipate times of lower demand to take the necessary actions.

What differences do you consider between the RMS created by an international hotel chain and the existing generic RMS?

There is a big difference. A generic RMS is designed to be integrated into different hotel systems. However, a hotel chain's own RMS focuses on its own systems, so they can make more advanced developments.

In this regard, I had the opportunity to participate in 2014 in the creation of an RMS in Boston that was revolutionary, with a million-dollar investment and that after a year was implemented in more than 1,100 hotels around the world. It included parameters that even today I have not seen in other RMSs.

For example, a generic RMS usually allows you to set up a competitive group of hotels, but usually considers them equal for price recommendations. We had a system that compared each competing hotel differently, something that makes the price recommendations more realistic based on product-quality and actual market positioning.

Another example is the more complex algorithms we added to determine the price sensitivity of demand. It considered the historical impact of the pick-up after price variations based on similar demand parameters, giving each future day a level of risk to the price variation, which was shown in several levels, low, medium or high price sensitivity. In my opinion, these concepts are essential when it comes to establishing a pricing strategy.

It is not more important how much we raise the price for the recommendation of an RMS, but the impact that this price increase generates in the demand or even in the different segments, since it could have the opposite effect to the desired one. And if we are talking about resorts in tour operator destinations, the flexibility to change the price is rather scarce, but this is another debate.

How does the application of AI benefit revenue management?

What we know so far about AI is just the tip of the iceberg, it is just beginning to take its first steps and algorithms will soon incorporate data from AI, which will surely help us in decision making. To give some practical examples, demand prediction algorithms could incorporate some of the following parameters:

Socioeconomic factors: Consumption trends and future employment forecasts, CPI forecasts, economic situation and product price trends for each issuing country.
Impact of demand due to climate change factors or policies carried out by the different emitting countries in this regard. Airline strategies, associated costs, new technologies to reduce the carbon footprint and their impact on consumption and the future of air transport.
Air route changes worldwide, and especially due to international conflicts that vary the demand between destinations, anticipating possible route changes due to the resurgence of conflicts. Today we can observe pick up variations for this reason, but we seek to anticipate accurately and globally with data provided by the AI.
Even behavioral models for the generational change that is affecting the hotel industry, where the use of social networks exposes our business. We are looking for a model that associates the RPI (Revenue Performance Index), against a competitive group based on measures of early adaptation of this generational change, considering factors from the online reputation.

How does AI benefit demand forecasting?

No doubt we will have much more accurate predictive demand models based on all that additional information that we cannot manage now, but the most important thing is that we will anticipate possible changes in demand for external reasons before it is noticeable in our pick up. This will help us to be more agile and anticipate strategies. And if, in addition to incorporating internal data, we add external data, the demand forecast will be more accurate and we will be able to make better decisions.

What challenges might be encountered when implementing AI in these systems?

The challenge facing AI is the veracity of the information that floods the networks, which is where it is nourished. In our specific case, it could be unverified data. Perhaps the search for those "INPUTs" that provide value is the greatest challenge for the RMS. To the extent that everything can be sorted out, the future of the RMS is promising. I am convinced that, if I were involved in the creation of a new RMS today, it would certainly include data from AI.

How could a small hotel, not a hotel chain, apply the use of an RMS and AI? Is it possible?

Everything is possible. This is always the big question I have in the courses and Masters I teach, but not only when we talk about AI, but also about RMS or Revenue Management strategies. Sometimes a small hotel does not have a revenue manager who can optimize the GOP (Gross Operating Profit), because from the Revenue we no longer talk about revenue, but analyze from the GOP or even from the perspective of EBITDA to not only sell correctly, but to do it in the most profitable way.

My recommendation would be to go for Revenue Management, to have an RMS and a person who can make proper use of the system, it is the basis of revenue generation. It would not make sense in a small hotel to have an RMS or any other system if there is no one to manage it.

Finally, what advice would you give to other hoteliers on this topic?

The advice I can give is to never panic about the evolution of systems. In this sense, AI is here to stay and is a very important additional step in our industry as well.

On the other hand, I believe that we must attach great importance to training. It is essential to continue training in order to stay up to date and maintain a global business perspective. Constant learning is vital in all sectors and even more so in today's technological world.

This Thursday, August 8, we chat with Rafael live to talk about the future of AI and RMS. Don't miss it!