"The Anticipation Layer" and the Future of Hospitality - Arnold Amrhein

Josiah: You open this article talking about how much has stayed the same for a very long time. I'm going to link to it in the show notes. I want to walk through it because I think you've structured it really well — it helps people ground their planning in these dynamics, and you've explained it through the lens of your career. I want to walk through from the perspective of what has stayed the same, and what is changing.

So I wonder if you could pick this up wherever you'd like, but in the article you open going way back thousands of years.

Arnold: I'll take a step back before I get into that and talk about the genesis of the article. As I mentioned, I'm being very deliberate about understanding technology in my ecosystem. I don't have a technical background, so I'm looking at it through an operator lens, through a capital side perspective. I think that can be deeply introspective about what's happening in the industry.

When you look at hospitality from a first principles perspective, you can remove everything else and it becomes this patron-guest relationship. How to make people feel welcomed at a space. That is fundamentally what this business is. Hospitality has been an incredible laggard in terms of adopting new technologies. We have never once been at the forefront of adopting new technologies, and a lot of the processes within operations have been fractionalized. You had a PMS, a CRS, a CRM, a revenue management system. On the asset side, a building management system, a housekeeping system. Everything was deeply fractionalized.

As I've been deliberate about understanding technology in this space, I stumbled upon an article about the death of the PMS — how everything collapses into one thing, with a layer leveraged by AI where memory becomes the asset. I thought that was very thoughtful. At the end of the day, data will become the superior asset, and the ability to recall that data becomes important. But I thought memory stops short.

Memory is reactive. There is something missing, and that's when I thought back to first principles: anticipate the guest's needs. Memory is the floor, anticipation is the ceiling. That becomes the barometer of success for whatever we're building with AI in our industry.

Josiah: When you say memory is the floor — this is table stakes, a baseline expectation. That's what you mean by the floor.

Arnold: Exactly. Everybody can collect data, put it together, and bring it out. But when I say anticipation is the ceiling — how do we utilize that? How do we think through that interface? That's the genesis of the article. I started it by going back two millennia, to the inn in Rome. It's the same concept as today. It's become about scale, but the concept hasn't changed.

Josiah: You have this line in the article — this isn't a technology problem, it's a human one. I appreciate that because I'm based in San Francisco, surrounded by AI. I've spent a lot of my career building technology companies, so I think through a technology lens. At the same time, I'm recognizing there are real limits to that, and there's a human experience that is timeless. When you talk about first principles, it seems like you're going back to what humans fundamentally want and need — and maybe that's the best lens through which to think about strategy, guest experience, and technology.

Arnold: One hundred percent. I am a techno-optimist. AI and broader technology will enable humans to free up their time and focus on that first principle. AI will allow us to liberate time to deliver that intangible service.

In the hotel business, there are two important cost centers. One is labor, and everybody knows the OTA and commission structure. Labor is probably the single most important cost structure in a hotel P&L. That speaks to the human side of the business — you are delivering an intangible service. When you give your team the capacity to remove themselves from the minutia, from the processes of bringing up information, and actually go to the human contact — that first layer, that interaction — you free up a lot of that capacity. I'm not saying you'll be able to reduce the labor force. I think you will be able to reinvent what that labor force can do with those tools.

The second thing I broadly mention is the channel your guest comes from. Everybody knows the OTA argument around commission structures, everybody knows your rate gets diluted. But nobody is speaking about the data handicap that comes from OTAs. When someone doesn't come from a direct booking, you have a data gap — what that person is expecting from their stay, what their purpose is in coming to your hotel.

What AI enables is: as we create experiences for both new and recurring guests to our properties, the transition into direct booking becomes a seamless trend. You'll be able to understand what those guests' preferences are, what their intentions are with their visit, and be able to anticipate how to deliver a service. You'll know what that guest likes in terms of wine, in terms of food, how they travel when they're on business versus with family. Does a guest come every April to celebrate their wedding anniversary? Why don't you capitalize on that during the stay and think through offering some form of economic incentive for them to return, knowing they usually book three months before they come back in April? Why don't you create an offer or an experience that inspires them — that's incredibly thoughtful.

Josiah: So I'm hearing implications on both sides — customer acquisition, distribution channels, how we acquire guests and retain them. There's a play there linked to the guest experience — the stay is so good you want to return, but you also have mechanisms for interacting with guests to get them back, so the economics can improve. And the economics around labor can also improve.

You had this line in the article about the historical trade-off between scale and soul. Smaller independent properties have leaned toward that soul, that bespoke experience. Then there are participants more focused on scale, and those often end up acquired by one of the big hotel brand companies. There could be a third path here where that trade-off disappears.

I was recently talking with Burak Yilpaci, who is the general secretary of Les Clefs d'Or, the worldwide association of hotel concierges. He's a concierge himself, won Best Concierge of the Year. He was excited about this trend because he said time is luxury. His implication really stuck with me — it's a luxury for the people providing hospitality. To be a great concierge, you want time back to have that meaningful guest interaction. He wanted more time to just have coffee with his guests. But to do that, he needs the busywork handled.

One of the really interesting components of what you put forth in this article was around guest data becoming an asset class of its own. You're an investor. You've been an operator. What do you see happening there?

Arnold: I love the concierge example. Every great hotel has always had someone running what I call the anticipation layer, but in their head. The question AI answers is: how does this look at scale? The concierge, to your point, has been doing pattern recognition. They've been activating their network of contacts. They've had behavioral inferences to recognize patterns and provide a personal recommendation. The problem is that breaks after your 50th guest. It breaks at scale, at volume. And it limits the capacity for someone to truly sit down and have a coffee with their guest.

Back to the data point. Every GM, every hotel owner, every operator knows there are thousands, if not millions, of data points across their entire infrastructure — whether recorded in folio cabinets, in PMS systems as preferences. Somewhere, all of this data is stored. What has traditionally happened in a lot of hotels: some brands have been able to mechanize this into consolidated preference patterns. But everybody knows hospitality has one of the highest turnover rates of any industry. That brings a tremendous amount of brain drain from your top team members — because you have people who have been at a property for a decade, seen the recurring guests and the new guests, understand the business, and carry a lot of these preferences simply in their head. Then they leave, and take a tremendous amount of this data with them.

Hotels need to not only understand where their data is sitting and aggregate it, but beyond that, process it, identify insights and patterns, and anticipate a service that is inevitably intangible.

How am I thinking through this from the capital perspective? How does this create asset value? Let's step back. How do we traditionally underwrite or create value at hotels? NOI — a multiple of direct cash flow the property produces. When I'm looking at two similar properties with similar cash flow, one can assume they're similar value. But not quite — there's location, the age of the asset, the physical building. What I briefly touch on in the article is that I think that's changing. There's an opportunity here for assets — whether independent, whether portfolios — that have been able to consolidate their data, organize it, structure it, so that any new owner can come in, interpret that data, take from it, and develop insights. I think that will create additional value when underwriting a property from the capital side.

The data has existed — it's always been there, nobody was looking at it. We retrieved it when a guest came in: "Look in our preference book." Now we have technology that enables that. But back to the memory issue: that's just recall. Our ability to anticipate will be what actually creates value.

Josiah: You have a great line in the article — the owners who recognize this early are going to start treating their guest history the way a good restaurateur treats a wine cellar. It's something that compounds, increases in value. It goes back to what you mentioned about the interconnected nature of a hotel: this data allows anticipatory service, which creates loyalty, which spreads word of mouth, which has benefits across the business, including team retention. It can be a reinforcing cycle that gets more valuable over time.

Another implication you get into — I'd love to hear this from the capital side. You talk about margin getting unlocked. Do you see this leading to more interest from institutional investors in hospitality as an asset class?

Arnold: Hospitality as an investment has seen an evolution that is night and day over the last few decades. Two decades ago, the people touching hospitality were touching it as a hobby. Hospitality assets weren't seen as wealth compounders. They didn't have the economic incentives to pencil an investment into them. That has slowly but surely evolved. It is now very rare to find a diverse portfolio that doesn't have alternative investments in real estate, with some degree of exposure to hospitality.

The reason for that has been a combination of institutional players coming into the business — being able to truly understand what running an institutional operation entails, being able to deliver margin where the independents, the mom-and-pop shops, didn't have the sophistication or the processes to do that. Once they did, you could create budgets and expectations, underwrite for years, and really pencil in a hotel. From the other side, capital markets had been typically inefficient in our side of the business. Banks and lending institutions were very risk-averse to hospitality as an asset. That has completely changed.

Now we have this incredibly interesting asset class that has been institutionalized and has the capacity to use technology in ways that go beyond just the economic incentives of NOI. Data becomes very interesting.

To give you an example beyond the acquisition perspective: I'm also thinking through the CapEx cycle of an asset. Owners know that physical assets have CapEx cycles — the building, the roof, the infrastructure systems. But not a lot of people are thinking about technology. Everybody gets the base case — let's have the best Wi-Fi. But typically, the conversation around a PMS or a CRM happens once. You get someone who sold you on a piece of software, you implement it, and you run it indefinitely. I think that's going to change. CapEx will now not only be focused on the physical asset, but also on the technology stack.

Hotel owners and investors will start thinking like technology operators — asking how they aggregate this data, how they can use technology, who's building the best systems. As I've been deliberate about understanding how this ecosystem is unfolding, I keep thinking about it like a gold rush. The people who made sustainable money during the gold rush were the ones selling the picks and shovels, not the people actually digging for gold.

When I first started to understand this ecosystem of technology in hotels, I was in shock at the number of tools, startups, and software I was completely unaware of. There will be an important moment in our industry where, as happens in other industries, most of it gets commoditized. Some of it will aggregate to the big players. But it's fascinating what's happening. Operators will have an interesting time being thoughtful about what their tech stack looks like today and what it looks like for the future they're building.

Josiah: There are so many elements to this that are fascinating. I wish we had hours to talk. This idea of operating margins increasing, this whole system playing out as you've described — that is a great thing. More money moving into hospitality means more innovation. If these businesses work, it means we can pay people better, create better guest experiences. Everybody wins.

My last question before we go: where are you getting inspiration from? What are you watching and finding valuable in learning about all this?

Arnold: I think it comes from a combination of things, and it's a great way to conclude. The more AI automates, the more physical presence becomes the luxury. It makes the human moment more scarce and more sought after. I call this the return to the analog. Everybody will go back to those analog experiences because everything else — our inboxes, our commutes, our content feeds — is already structured and optimized. So let's go back to what started this conversation: first principles.

As I'm listening to podcasts, hearing operators — including people who have had tremendous respect over the years, like Radha Arora, the president of Rosewood — everybody is slowly but surely speaking about what makes this business the business. When a guest comes into your hotel, when a guest steps into their room, how do they feel? It is an emotion. Do they feel at rest? Do they feel at peace? Do the people at the front desk genuinely see them? It's that irreducibly human moment, and I think it will be profoundly enabled by AI.

A lot of the articles I'm reading come from technology operators who have never operated in the hospitality business, and they lose the point of what first principle they are actually delivering. For me, I've attempted to spend time in silence, thinking and understanding the feeling. I go back to that moment as a kid thinking, "I feel so welcome — what is this?" And I bring that to the present. This is an emotional business. Let's go back to that. Let's solve for that, and I think hospitality will be incredibly enabled.

Josiah: Arnold, I have learned a lot from you in this conversation. I'm going to link in the show notes where people can read your article and follow you on LinkedIn. You are clearly thinking about all the different elements of what's going on and the implications of it, and I would love to have you back on the show sometime to explore this further. Thanks for taking the time to record today.

Arnold: Thank you, Josiah. It's been a wonderful conversation. You're a great host.