Whenever I have some free time, I like helping out the "new blood" of the industry so, whenever students ask me to help them with their thesis, I am more than happy to do so. Yesterday, I had a long chat with Monica Gutierrez a bright student at the Heilbronn University of Applied Sciences, working on a dissertation on the future of hotel marketing and the impact that AI will have in the industry over the next years. The following is the transcription of our chat.
Hi Simone, and thank you for that. My pleasure.
You mentioned that you are writing a book on the future of hotel marketing. Yeah, this is going to be out, if everything goes well, around February '21. The book is all about what I call the "travel singularity." It's the concept of singularity, but applied to to the travel industry: a moment in time when the use of artificial intelligence will free humans from repetitive jobs, so that they can focus again on taking care of guests, rather than taking care of tasks that can be done by robots. The book is more philosophical than marketing-oriented, to be honest.
How's that? I don't know if you're familiar with this, but some thinkers found similarities between Nietzschian philosophy and transhumanism; I start from that assumption: the connections between what Niestzche used to call the übermensch and AI. The book is about that, you know, AI seen in a more philosophical way.
I see that you had how many years in Hospitality, is that twenty years? Twentyone
According to your experience, what is the future of revenue management? Revenue managers today have to take care of so many variables when it comes to finding the right price and the right inventory, that is not a job that can be done by humans any longer. I mean, this has been the case for least the last five years, now, it's nothing new. Pretty much all the big brands implemented some level of AI or rule-based systems to find out what the right price is, and how and where to allocate inventory. If you talk to anyone with some level of intellectual honesty, he won't be able to name one single reason why a human would be better at doing revenue management than a machine. You can understand the implications of that. There's this school of thought that is pretty much against the use of AI in RM, but that is because, you know, the moment you do that, all these revenue managers would be basically out of a job...This consideration is done mainly at the independent hotel level, while branded hotels already started integrating several layers of AI in pretty much whatever they do, you know, from price accuracy to inventory management, to cancellation management...
Cancellation management? Yeah, there are several companies using ML to predict the amount of cancellation a hotel will get at any given time and, based on that prediction, selling more rooms than they actually have, as they know that some of these rooms will be, very likely, canceled anyway.
So, are humans revenue managers doomed? Today, when you look at the stock market, ninety percent of what they do is run by an algorithm. It's really impossible for a human being, or even a team of human beings, for that matter, to analyze all these different variables. If you take a look at the evolution of most of these revenue management consulting companies, you see that they're starting to release software because -deep down- they know that consulting with an Excel spreadsheet is a thing of the past. There is too much space for human errors when it comes to revenue management. But, I mean, this is not even something we discuss any longer, that's just a given: humans can't do that anymore, so that's almost an afterthought, really.
What about advertising? The old school of doing it is that you hire advertisement firms or web agencies, and they run your campaigns using humans. So, basically, you've got humans creating campaigns, choosing audiences, writing copy, etc. but that's no longer doable, especially for more complicated campaigns, where you have a lot of variables, such as metasearch engine advertisement. When you have an OTA selling your room with a 20% discount on the Russian market for 20 minutes, this is not something you can discover with an Excel file, can you? There are so many markets, so many demographics, and so many different devices that it's impossible to make sure that each user gets the best message if you do it manually. Moreover, it's an unscalable business model for advertising firms. It's really not sustainable, and I tell you from personal experience.
So, what 's next? You know, a lot of firms started integrating some level of AI in their advertisement strategies: again, metasearch ads are the typical example, but AI can be applied to pretty much every campaign. If the system understands that, let's say, users in Los Angeles using Chrome browser on a MacBook have a better conversion rate at 7 a.m., it will automatically change the CPC accordingly. You'd need an army of humans for that...
So, human advertisers... ... will end up like human revenue managers, yes. Let me ask you something: what's the square root of 1,342,626?
No idea. Here you are: the first thing you do is grab your phone, and you do it on a calculator, right?
Right And does it diminish you as a human being?
I don't think so. That's the main point. What most revenue managers do today is merely collecting data and waste a lot of time in useless calculations, while this data is already ready if you use machines... When I started, when I wanted to do simple things such as understanding my competitive set price strategy, I had to actually go online, check day by day on OTAs and write it down on a piece of paper... These were the analog days, and, back then, revenue managers' salaries were worth it: they usually spent 7 and 1/2 hours a day just aggregating data and 1/2 hour creating strategies. Today, we can save those 7 and 1/2 hours. There's no shame in using a machine to do that. And, by the way, the square root of 1,342,626 is 1158.7173943632674. I've just checked that on my phone.
And what about marketing? I think we still biased to think that marketing has some level of creativity that cannot be replicated by machines, but, again, this is a myth. Think about Jean-Baptiste Le Divelec's recent Nike commercial, that's a beautiful example. Or generative design, ADI, the list goes on. AI can create multiple different versions of the same campaign to get the best out of conversion, or even create products tout court. I think we have a cultural problem with letting machines being creative, but, hey, "change alone is unchanging," that's Heraclitus, not me. A lot of companies are either releasing their own algorithms, buying AI startups, or partnering with them. It's Darwinian.
There are industries, such as medical, where AI is already mainstream adopted. Why it's not the case with Hospitality? Most hotels fear the introduction of even tiny pieces of AI in their processes. Think about chatbots, for example: I've seen hoteliers scared that, by implementing a chatbot, they would lose the "human touch," whatever that means. While, you know, the only thing that a bot would do for them is saving the front office guys hours of numbness work spent on answering the same questions over and over again or typing on a keyboard. I always say that the problem with artificial intelligence is in the name itself: it's a semantic problem because we when we say artificial intelligence we are already implying that is something opposed to, or even against, human intelligence, so you got artificial intelligence on one side, and you've got human intelligence on the other. But that's rarely the case. It's not AI vs. HI; it's rather AI + HI. But, hey, if you call a chatbot a chatbot, you are already negatively implying something. But how is that different from using a PMS, right? All hotels use a property management system, and I doubt they're afraid of losing the human touch, just because it's a machine allocating rooms and printing invoices. But that's called a property management system, not a robot concierge. Words are critical.
So you think the main problem is misunderstanding the technology? Damn right, it is. When it comes to AI, you are moving to a territory that most people don't fully understand and, let's face it, many people, whenever you say AI, think of Terminator. '80s science-fiction movies are at least partially responsible for our bias towards AI.
Most hotels today don't even think about applying AI, but do you think it's something that they'll need to do, eventually? Will they be forced to apply AI? Oh yeah, sure. I mean, there's no other way. Hospitality is not only a human industry; it's a data industry. And, probably, the industry creating more data on any given day than any other industry. Think about the amount of information we have about guests: credit cards, passport numbers, what food they like, what kind of rooms they book, and so on. Why do you think there are all these data breaches in Hospitality? If you can get into a PMS, you will be able to get so many unified personal information, and you will have a full understanding of that person, something you won't have by breaking into the management system of, let's say, a clothes store. With all this data, that is the perfect scenario for AI to flourish, but, on the other end, because of all this data, there's no way that you can understand this maelstrom of information without the help of AI. There will be a point in time when hotels will simply not be able to keep up.
And when do you think this would be? Yesterday.
You've mentioned data breach. How do you think guests can trust hotels to keep their data safe? I think we're not seeing the full picture here. In Singapore, they are using a lot of biometrics to check-in, so you get to the hotel, and you don't even have to show your passport. They will scan your face, and that's it. Regulators are freaking out about that, but I am way more comfortable with face scan then that I am with giving my passport and credit card so that the receptionist can make a physical copy... I go to different hotels every day, and I can tell you horror stories: they have piles and piles of folders, accessible to anybody. You can grab a folder and buy stuff on eBay for the next decade. I am more concerned with that than having my data stored on a cloud, encrypted system, with different layers of security. Are you worried about the face scan? Get the hell out of here. Look, I am 41 now, so there have been a few times in my life when I had problems with my credit card. Yet, every single time I had that kind of problem, it was not because I bought something online or because I gave my data to a system, but it was always because I gave my credit card to a human that did something with my card.
Is that even legal? Scanning cards and documents? Nope. Let's say you tomorrow you go to a hotel, right? The first thing they ask you is, "Give me your credit card," and that is usually for a guarantee, right? Now, if something goes wrong, let's say that you leave without paying, or you drink all the alcohol in the minibar, then the hotel should be able to charge you, right? How do you think they do that?
No idea They just go to the basement, take the copy of your credit card, and manually charge you. With PSD2, it will be harder to process a payment where the credit card is not present, but today, if you're a receptionist, you can do pretty much whatever you want with your guests' credit cards. A photocopy is untrackable. Good luck with finding the liable here... Now, I understand that hotels should have some level of guarantee, but that can be achieved, legally, with a payment gateway. Gateways store your credit card information for a limited amount of time, that is exactly what we have in Europe with GDPR. And even though the regulation, per se, is BS, at least we have some guidelines: you can use your guests' data only for the amount of time needed to do whatever you're doing. So if you are checking out today and you leave without paying, hotels can charge your card today, but they won't have access to your card in, let's say, one month. Look, I was a hotel a couple of weeks ago and that they were renovating, right? I've found folders and folders dating back to the late '90s, with a ton of personal guest info. You're worried about AI? Give me a break. If a hotelier reads that and has some decency, he will have to agree with me. Data security in hotels is a joke, and I blame humans for that, not machines. I understand that hotels want to err on the safe side because cashback can account for up to 1% of their losses... but there must be a better way. In a perfect world, checking in and checking should be frictionless, such as getting into an Uber ride: you get in and out of the car, and the payment is mainly an afterthought, yet hoteliers get pretty nervous when there are no humans taking care of payments, even though they are self-sabotaging.
What do you mean with that? There is a study made by Cornell University saying that, for Americans, a five-minute wait in line during check-in negatively impacts reputation by 47 percent. And this waiting in line is usually due to check-in and check out processes, what I call the "logistics" of travel. All of this can be automated: self-check-in kiosks, biometrics; you name it. Do you see my point? Not only you are not doing the guests any favor, but you are increasing human error, working in a way that is entirely illegal, and wasting precious human time and resources. I mean, wake up! Why should we let humans run this part? It's proven that they are not good at it.
So, are you suggesting that, in the future, there will be no human receptionists? Before technology became an essential part of Hospitality, a receptionist was somebody that was taking care of guests. What a receptionist does today, by contrast, is mainly taking care of boring tasks that can be done by a machine. I've been there, both as a consultant and as a receptionist myself, back in the late '90s/early '00s. It's pretty common that you go to a hotel and the receptionist is barely talking to you, because it has to do so many things like, you know, checking you in, making a copy of your passport and credit card, computing you in on the system, generating the key card, etc. All these things do not enhance the guests' stay by one bit, they actually make it worse. So, smart companies started giving guests a choice, like in supermarkets: you've got five people in front of you waiting in line, and then you have that little automatic machine, where you can scan your barcodes, pay, and get the hell out of there. I am not saying we should get rid of humans, but what's the added value, both for the guest than for the receptionist, to have a human being scan a passport? Being a front office guy used to be fun, now it's alienating. Most people don't understand that they are more robotic now, with their broken processes that they could be with proper automatisms. Paradoxically, the more technology you put into hotels, the more human the experience will become, because you are freeing your staff from all these repetitive tasks and putting it back where the added value is: taking care of guests. I don't care if it's a human, a robot, or a chicken, checking me in, but I'd probably prefer talking to a human being for tips on good restaurants (unless humans are taking a commission out of it, but that's a horror story I'll save for another time...)
What about AR and VR? There's a whole advertisement potential that is completely unexploited there. Now, if you want to move from A to B, I'm pretty sure you use Google Maps, right? And, on Google Maps, you have a layer of augmented reality, so you can see on top of the "real" world, information from Google like, you know, turn right, turn left, etc. You may remember the Google Glass fiasco but, if you want my two cents, the problem with Google glasses was not the technology, but that people were not ready. It was too much, too early. A failure in timing, not design. Case in point: Facebook is now working on its own version of glasses, and these will not be a geeky, nerdy version: Zuckerberg signed a partnership with Rayban. So you will have this amazing pair of Wayfarer and, on top of your lenses, a layer of augmented reality. Do you understand why there's so much potential? Let's say that you are looking at a street, and, on top of the "real" reality, you can see restaurant menus, prices, and reviews. You can reserve directly from Google Maps. If you want my two cents, it won't be google.com/travel to win the travel game, but Google Maps... And the moment you can show your property in front of the others via paid ads, you'll get to a complete virgin advertising territory.
And VR? I doubt it will be as disruptive in travel as AR, yet, it may be used for a more top-funnel part of the traveler journey. I am working at a complete hotel rebrand now. This property is very MICE-oriented and, instead of printing the usual old-school brochures, we created a VR video for each meeting room. We sent companies a free card box visor, so they're able to see the inside the meeting room before that's even built. But you know, there is a cost involved, because you need to create a video, and you have to do it in a certain way, and you need to buy the visors, etc. I think it's great for some hotels but not for all. Again, my winning horse is AR: Google closed a lot of travel systems lately, like Google Trips, and they started converting all this technology into Google Maps. I see the near future (and I'm talking about maybe five years), where Google Maps will be pretty much your main point of contact for everything travel-related.
What other trends should hotels look for? 5G and blockchain distribution. The first creates the perfect infrastructure for technology to flourish, while the second can fix the rate leakage issue once for all, making metasearch engines obsolete. But don't be fooled: technology doesn't happen in silos. That is why I talk about "Singularity," because there are several different technologies that are pointing all to the same direction. On one side, you have 5G; on the other, you have biometrics, open API adoption, voice, AI... All these things are happening at the same time, and they're happening so that the guests can, eventually, have a frictionless experience. The top of the iceberg is AI, but there's a lot that we don't necessarily get to see. Yet, in order to get to this level of technology, we need to make sure that all the systems used by hotels are connected with each other, and that is another big blocker we have in on the industry today. A lot of these systems were built in the '80s and '90s, without the need for integrating third-parties in mind. One blocker is our cultural suspicious towards AI, but the other is that most hotels are still running on obsolete, on-premise systems so, even if you want to integrate some level of AI in your operations, you can't, as your taken hostage by that spaghetti-code, on-premise pieces of cr°p. If you're implementing an RMS, you may want to know, for example, the speed of pickup, and that information is in on the PMS. If you cannot access it, then there's no AI. Not being able to access the data means that you are not able to add any level of automatism, let alone AI. So, if you want to move to the next level of Hospitality, the first thing you need to do is to auditing what kind of software you are using and getting rid of any possible legacy systems working on monolithic structures. Tech agnosticism first, amazing technology later.
Thank you so much for your time My pleasure and good luck with your thesis