ҹɫֱ²¥

Transcript

Introduction to Atomize RMS
 
Hi, everyone. Welcome back to another Matt Talks. I don't even know what episode number we're on. I think we're somewhere in the t's, and I'm really enjoying these conversations.
 
And this week, I wanted to go really deep on Atomize. Atomize is the revenue management system that we acquired at the end of last year. And and as we were looking at the market, there were so many solutions, but we really fell in love with Atomize, the product and then the team. But Ellen works on the product side, and I thought we need to get Ellen on Matt's talk so we could I can do dig really deep on what's the vision for this product and well, you've been with this company for a while.
 
So, hopefully, you can kind of unearth some of the secrets behind this product. Ellen, welcome. Do you wanna quickly introduce yourself?
 
Thank you, Matt. Thanks for having me. So I'm Ellen. I'm a product director for the RMS, and I joined Atomize in early twenty twenty just a few weeks before the pandemic hit. So it was really interesting time times. And I have a background as a product designer and a product manager, so I've actually been doing all of that in Atomize. I've been doing everything from product research, product discovery, even to UX design up to the more strategic decisions made.
 
Love it. Did you did you always know you wanted to be in product and be a product leader? Or is this something that you learned over time?
 
I actually think I almost always knew because I I remember being a child sitting by the kitchen table and, like, someone built this product. Someone worked on this product. And I always liked, like, changing the time on the microwave. Like, that was a very nice challenge. And I was like, how can it be so bad user experience? How is the usability so poor? So it's kind of been part of me forever, I think.
 
So is it what you then went on to study?
 
Yes. I studied something called industrial design engineering. So it's both, like, focused on product design and also I'm an engineer as well. So it covers kind of all of that.
 
It was funny when we started ҹɫֱ²¥ twelve years ago, and we just hired a few developers because we just assume, well, we have to build software, so let's bring in developers. And after it must have been about two years. At some point, I'd read the Marty Kagan book about product management. And this person said, think what we're missing is a product department. And we realized that me and Richard had become the product designers and, like, we've been thinking through it conceptually. And we're like, oh, there's people who actually have this as a job, who have been to school for this. And it was such an epiphany that existed, because, honestly, that maybe the thing that I should have gone to study because I love product development so much.
 
Yeah. I mean, it's a it's a great thing. And it was actually the same in Atomize as well. The founder was really the one doing the the product part until he felt like, no, I need someone else to do this. And this is when I was hired.
 
Nice. And in your intro, you just mentioned twenty twenty, which was a mad year, if we all remember.
 
That was the year the start of COVID, which if you're a pricing solution, I'm imagining your algorithms went completely crazy at that time. How did you deal with that?
 
Well, we had to do a lot of work around solutions because all of a sudden it was actually I mean, of course, when there's no demand, obviously, very few price changes as well. But what was mostly impacting us was also the further we got, we had very unreliable historical data as well, which is one of the important parts. So we had to build work around solutions making this the system not look at the nonrepresentative data that came through the pandemic.
 
It's really hard.
 
Yeah. Yeah. It was definitely challenging times.
 
I can imagine this is just you walking into the business, having to figure this out at speed, I'm assuming.
 
Yeah. Yeah.
 
So if you for people that don't know Atomize k. Who have never heard of RMSs, and we see this a lot in hospitality, the majority of hotels don't actually use a revenue management system. They still rely on Yeah. On humans to use Excel sheets and determine what the right price is based on what they see in the market. What's been the vision behind the Atomize core product?
 
I would say I mean, we started with a very clear vision that we wanted to deliver this sophisticated and data driven price optimization where we could combine automation with exceptional ease of use. So we've invested quite heavily in building this optimization engine, and it uses both internal property data, so looking at your historical data and your current performance, and it combines that with external market signals. And then we we do that to generate accurate real time price recommendations. So the real time part has been really important to us as well because, especially in markets where like city centers where there are big demand fluctuations, it's really important to be acting in real time.
 
Yeah. So what are some of the market signals that you refer to?
 
So market signals would be both your sign the signals coming to the hotels, like, the pickup and the pacing, but also the his the data from your competitors. So competitor pricings.

Chapter
The power of algorithms in revenue management
 
And how fast is the algorithm adopting? So one of the things we obviously see, we see quite a volatile macro climate right now. And one day the stock markets are up, then they're down, and that has a downward impact on travel.
 
How fast is the algorithm at adjusting to some of the changes that we're seeing in cancellations and pickup in hotels, and can it adjust the prices?
 
Yeah. Well, so our product runs in two different modes. You either have the regular mode, which is you have three optimizations per day. I mean, if there are changes in the data, it will adjust.
 
And then we have the real time, the real time price optimization, and it performs calculations constantly. So it happens all the time. But I think it's important to point out because we are very strong promoters of real time optimization, so you ensure you're not, like, leaving money on the table. But at the same time, I think some people get scared of, like, thinking that their prices will update constantly.
 
Like, it's not updating three sixty five days every room type, all the time, every second. It's it's adapting to the stay dates that are actually affected about the change shifts in demand.
 
Right. So if you see a short term number of cancellations, then it's only gonna look short term to adjust the pricing. It's not gonna constantly look at, like, a full year ahead.
 
Exactly. It will it will look at I mean, it will, of course, look at the full year ahead, but it will will change the prices according to what's happening on each individual stay dates and the surrounding stay dates.
 
And you mentioned that some people are sometimes a bit nervous about allowing the algorithm to just automatically set the prices. Is that the majority of hotels that do it by hand, or is it the majority that actually end up using the autonomous pricing?
 
Well, what we see from our users is that, like, I would say looking back a few years, it would take properties normally maybe three months, and then they would start trusting the system enough to turn on the autopilot. Nowadays, people we are seeing a shift that revenue managers are more and more starting to trust systems.
 
Now we have people onboarded, and next day, we see that the autopilot is turned on. That's amazing. It shows trust in the system.
 
What shifted, do you think?
 
Sorry?
 
What shifted?
 
I think there are a few things. So one is that the overall trust in systems and that we are getting more used to AI. We're trusting technology more and more. That's one part. And another part, it all is also the work we've done in Atomize with kind of the controlled automation. So what we're doing is that we give the users a possibility to say when the autopilot should be on and whether may they might want to manually adjust new time, stay dates, for instance, or certain periods of time, and they want to have that manual oversight, then they can exclude those. And we're also having a lot of more pricing controls that allows revenue managers to say, for instance, control the minimum and maximum threshold where our prices are allowed to be set within.
 
Right. So I would say, I would don't ever wanna sell a room rate under hundred euros because that might devalue my brand. And then there's a peak of it because the moment someone pays a thousand euros for this hotel, the expectations are of a level where you just can never meet them. So those are the boundaries within which you get prices.
 
And, also, if you know that, for instance I mean, often, I would say that, actually, the system can detect, like, far out events that that's coming. The system will detect that before you do as human. But sometimes it might be that you're actually having some intel that the system does not. Like, for instance, we have a customer that they have when there's a concert in town, the whole crew that comes and build the stage, they will book with them. So even before the the event is announced, they will know about it. And they would typically go in, and they'll increase the minimum price to protect to make sure that we're already at the high a high enough level when bookings are starting to come in.

Chapter
Real-time pricing adjustments
 
Because I remember when Adele came to Munich and I got a so we were in the queue to get tickets. But whilst my husband was buying the tickets to Adele, I already had a hold on a bedroom because I knew that this hotel chain was not using an RMS. So we had booked it at a third of the price of what eventually the price would be. How quickly would Atomize adopt to seeing a pickup in a destination on a specific day like a Taylor Swift or an Adele concert to actually responding with a price increase?
 
This is really interesting because exactly this scenario happened with one of our customers just last summer. So they would always need to be on the real time price optimization, which these ones luckily were. So there was a, artist announcing that they were coming to the city in mind.
 
And and at that time I love how you're not disclosing the artist.
 
I won't say which artist was coming.
 
It's not even a secret because it was here in Gothenburg where we have our head office, and the artist is a very local Swedish artist. His name is Hakan Hellstrom, but he he's so he sells out our biggest arena, like, three nights in a row, like, instantly. So it it's very local event, but still Right. People traveling for all of Sweden and maybe Norway, potentially Denmark.
 
So it was like he announced that the concept he will be releasing the tickets in a week's time. And instantly, people started booking hotels. And for our customer, they had nothing on the books because it was, like, nine months away. And then the system, it understands that this is not anticipated demand.
 
This is not a regular pattern, and it instantly increased the price.
 
And then so it increased the price a bit, and then more bookings came in, and it kept on increasing the price until it And the first person who books probably got a good deal.
 
But the moment you start seeing the pace, then the price goes up. Right?
 
Yeah. Like, actually, the price was increased, like, a week earlier, which could have been that some other hotels in the market had known about this event coming. So thanks to us using competitor rates, we were already increasing the price one week earlier. But, yeah, it was the first booker, Probably did a good fairly good deal, but it then it raised it quite significantly. And it goes up, and then it realizes, okay. I'm at a quite high rate at the moment. Might not be having more coming in, but it's still so far out that it's worth to keep it to ensure that you're actually making these all of these incremental revenue increases all the time.

Chapter
Do small hotels need an RMS?
 
So and I think that this is leading to my next question. So an RMS might not be for every hotel. So if you're a tiny little hotel with five rooms, would it make sense to have an RMS? Because in that scenario, the rooms are already booked before you pick up, that there's an actual trend because you've only got five rooms?
 
Yeah. Actually, five rooms, that's a very small scenario. I would say you would probably benefit from having an RMS, but maybe you'll benefit from having a more simpler system than atomized one that is more rule based, for instance. Whereas, we're saying, like, twenty rooms and above, that's when you really start to see the value because our system also requires data. And with five rooms, we we simply don't get that much data.
 
Yeah. You get the market insights there. Yeah. So you can glean something from what's happening in the market.
 
And concerts don't happen every day. No. Exactly. So, generally, you won't see this massive pickup.
 
So there's definitely some angle to it, but it probably gets really beneficial when you have more rooms.
 
Yeah. That's it.
 
So how does Atomize adapt to different types of hotels? So I can imagine a downtown city center hotel has a very different pattern than a a resort that has much longer booking times. How does it optimize for that?
 
Well, I think this is a very good thing with our system that it adapts to the unique context of each property because it learns from the the historical data and the current performance of each individual property. So in fast paced, like, city hotels where booking windows are shorter and demands can shift quickly, the atomized real time engine would help hotels respond this, like, we were talking about, like, in instantly and also capture value from short term spikes, which could be driven from group bookings, business travel, local events happening. Whereas looking at leisure resorts, that typically, as you said, they have longer booking windows and maybe also more pronounced seasonality, then Atomyze would tailor the pricing strategy by analyzing the seasonal demand patterns, group booking behaviors, and cancellation probabilities, allowing the system to optimize rates over more broader time horizon and capture value earlier in the booking curve.
 
And I also think that we have one quite recent feature that we added that is especially valuable for resorts, which is the ability to factor in estimated ancillary revenues. So you could include spa treatments or dinner packages. Nice. And that would allow the system to optimize more for total guest value, not just rooms revenue.
 
So you talked about needing historic data so that you can get these patterns based on which you do pricing. If you were to deploy Atomize in a in a hotel today, what historic data are you looking for? How many years or months? And and what level of data are you are you requiring?
 
Actually, as we built the system today, we are able to start with hotels that have no historical data. So for instance, if it's a brand new ҹɫֱ²¥ property, you are still able to go live with Atomize.
 
But during the first year, there will be a higher uncertainty in in the demand forecasting. Now we're actively actively building a solution where users can help out to kind of, adjust these overall price levels. Because currently, it is our optimization engineers that helps with kind of creating this historical data, looking at similar hotels. So it's not your actual property data, but something our best guest almost.
 
But then you also have, like, the historical data is just one part. How the performance is at the moment and how the competitors are pricing is also very important, and we'll lean more towards that when we are lacking historical data. Nice. So but if you have three months or or a year, ideally a year, then the system works very well.
 
Nice. And and so, like, we always know that Christmas is always the twenty fourth, twenty fifth. Easter, however, seems to shift. Does your algorithm know this?
 
Do you have to tell that in some way?
 
Yeah. We have this very basic, but still important integration with Google Calendar. So we are aware of all the regional holidays in each market and when they are happening, and then it the it maps to the relevant historical data.
 
So if I have, for example, a lot of Orthodox Easter guests coming, which is a different week than Christian Easter, for example, the calendar of that hotel would be aligned so we would actually know you're likely to get a pickup of of these travelers?
 
Yes.
 
That's incredible. Oh, sorry. I'm just looking for through my questions. Another with ҹɫֱ²¥, we have an open API, so we have a lot of data available. Is an onboarding with ҹɫֱ²¥ easier nowadays than it was a few months ago, or or will it be in the future?
 
It's already easier. So it goes it's out of all the PMS integration that Atomize currently supports, news is definitely our easiest ones. And we're currently enhancing this, So it will dramatically simplify the onboarding for ҹɫֱ²¥ customers with the goal to for for ҹɫֱ²¥ customers to be able to go live with Atomy's in under sixty minutes with minimal setup required.
 
So what does it what does it take today? Like, just to put the sixty minutes in perspective.
 
Today, it's still we're actually having a manual step in Atomize. So we have an onboarding engineer going through your data. So you'll have to fill in a sign up form after which the onboarding engineer starts looking at your data and the things you've input, and then it takes them around maybe two hours to do the actual onboarding. But it's also requires then some back and forth with emails and stuff. So so in in total, I would guess it takes a couple of days before you're live.

Chapter
ҹɫֱ²¥ and Atomize RMS
 
Which is still very fast.
 
Yeah.
 
Yeah. We want technology to be so smart, especially by bringing ҹɫֱ²¥ and Atomize together, that those data flows are so seamless that we we don't need to do any any more validation of the data. And I think that's the goal that we have where we can just trust the data flows between the systems, and we can now really accelerate the onboarding. Yeah.
 
So we set some really spiky goals because onboarding within sixty minutes, no one else does this in the marketplace today. Yeah. So if you were to fast forward, because you're a product leader, so I'm sure you're already thinking about next year and the year after. What are some of the things that you would love to build?
 
And I I'm not saying that these are the things we will build, but what are some of the things that you would love to build?
 
Well, I think, I mean, looking at the RMS space, I think I'm very excited to see it evolve from central from this kind of room centric model, which is where we're at with Atomize today as well, to one model that optimizes total revenue across the property. So it could be meeting spaces and other services. But, ultimately, we want to get to total profit optimization. And I think within the ҹɫֱ²¥ ecosystem, we're kind of uniquely positioned to lead this shift because you have so much data.
 
And we could leverage that, like, rich guest data, for instance, and we can combine that with the pricing intelligence. So that's something I'm really looking forward to. And I think also in a in every RMS, forecasting is a is a crucial piece, and it's kind of the foundation. To be able to set the right price, you need to do a proper job in forecasting.

Chapter
How to build more impactful revenue strategies
 
And these forecasts, they can be used not only to influence the pricing, but we can use them to inform housekeeping schedules. Or we could do it to target guest marketing, for instance. So I think this is where we'll see, like, real transformation happening. And I think it it will unlock unlock smarter cross functional collaboration at hotels.
 
Yeah. I think I think if we look at Atomize as the part of this today, it is so good as a stand alone solution. What we try to do is bring it together with ҹɫֱ²¥ so that we can really drive the road map because ҹɫֱ²¥ had capabilities far beyond just room pricing. We launched, I think, flexible product pricing. During COVID, we built this.
 
Okay.
 
And, you know, tons of spaces and bicycle rentals and co workings. And no one in the marketplace was able to keep up with it. And we thought, if we bring this in house, we can really build out a stack that genuinely thinks beyond just bedrooms. Because bedrooms are the core of a hotel, but there's so much money left on the table.
 
If you think about booking a parking space or booking, a bicycle rental, that those are prices are always stable. And I get excited about, like Yeah. Thinking beyond what we've always done. And I think that's really where the powerful powerful partnership will come from long term.
 
Yeah. I agree. And and we're already good at doing the room room revenue. So kind of like, it's really time to take on the next challenge and Yeah.
 
Build a stronger product and a stronger ecosystem.
 
And one of the things that is shifting, and I've I've talked about this before where the the traditional revenue manager was always in charge of room revenue, and they were in charge of setting the prices. And then they did that manually. So they, you know, that was the job that you looked at room room revenue. And I think it's shifting towards total revenue where that revenue mention needs to be less day to day managing Excel sheets and determining pricing, but much more strategic, which makes the job way more fun because suddenly you do get to think about, well, what upsell products will work and test out new upsell products and try out different pricing for different parts of the hotel. So I think we're gonna see a massive shift in the role of the traditional revenue manager to a much more fun strategic role, but it does require a visionary hotel manager to see that and to allow for this testing because AB testing is real, and not everything works for every hotel.
 
Exactly. No. I think you're absolutely right. And I think, like, automation will become the default mode of of operation for these more tactical execution parts. So that exactly what you're saying that revenue managers and general manager can focus more on the strategic parts.

Chapter
AI in revenue management
 
Love it. AI, let's talk about it. Yeah. What what are you guys doing right now with AI in the product?
 
So our product is built on smart algorithms. So we do that to build our demand forecast, and we also have, like, different models for cancellation probability, for group conversion, and for the demand forecast. We're also using partly using Gen AI, and we're making use of that to to explain to our users why we are suggesting a certain price or why we changed it. And because this is the biggest challenge, I would say, with the an RMS system that is AI driven.
 
It's not as clear as a rule based system for a user. Like, it's not as easy to understand why the decision the price change was made. So we realized that we need to give this understanding to users somehow. And so we let Jen, look at kind of our decision making and provide a clear explanation to the hoteliers.
 
Yeah. And it's it's it's bridging the trust gap, I guess. So where you know, we would love everyone to just adopt the autonomous pricing where you just let the machine take care. Initially, we have to win over saying, this is why we set this price.
 
So whilst it goes against really deeply explaining the algorithm is really, really hard because there's so many factors playing into it. Exactly. Giving the revenue manager a sense of, no, no, no, we got you. Like, we've seen all of the things that you've seen and a little bit more.
 
And I think it's the a little bit more that's really critical because revenue managers are phenomenal. These people have so much data and they are doing such incredible work to understand the market dynamics. But we also see a lot of those things, but we also see more in the trends and the patterns. And I think it's this that little bit more that has quite a significant impact on the hotel performance.
 
And I love those tooltips that just tell you what's going on so that they can move on to the next pricing decision and trust that that we got their backs.
 
Exactly. I think I think the main benefit is in situations where we are actually seeing that it's a high probability of cancellations, for instance, or where we see that, you know, you have some big group on the books that might not and and in the end, it might change, and we kind of have intel on that. These are the things where, our explanations really help them understand, our price decisions.
 
Nice. And I just talked about the little bit more. So, you know, when you when we make pricing decisions, sometimes hotels will uncover something through the data. Have you have you seen that in a hotel where they they learn something from atomize that they hadn't seen before?
 
Yeah. I mean, this is such a good question. And I guess there are so many things. And, I mean, just being a data driven system may and being able to process so much more data than a human can do, that that makes the biggest difference.
 
But I would say one thing that is a bit that has been a bit surprising is the impact of quite, like, local, more niche events that might be happening far out. So you wouldn't typically discover that. They wouldn't show up on your radar through a traditional analysis or or, in in an Excel model, so to say. But we could discover that this is there's something happening where the demand is somewhat out of ordinary, and the system has been able to react instantly and raise price ahead of competition and in capturing kind of incremental revenues.
 
Users have really said, like, this made a difference.

Chapter
How RMS helps hotels earn more
 
Love that. So, the majority of hotels today don't have a revenue management system. What would you say to convince them that they should have one?
 
Well, I would say to them that it's not only that it I mean, the obvious ones are that we have clear proof of increase in revenue and RevPAR. So so that would be like I mean, because the argument is something sometimes around that, you know, the RMS cost, but it it pays off instantly.
 
So for me, that's kind of a no brainer. But it's also another important part is that it gives you frees up your time. So we see hotels spending, like, up to twenty, thirty hours less every month because they can focus on other things. So so to your point, you can actually focus on more challenging and more fun tasks than actually sitting doing these very tactical price decisions.
 
What's been one of your favorite hotel experiences ever?
 
One.
 
Sorry. This is a curveball question. This is not revenue management related. I just wanna talk about hotels for a little bit.
 
Have to think about one that is well, I have to say I have to mention one, which is a common customer of ours where I went with you guys to do an event recently, which is Etem in Stockholm. They are so unique. I haven't I've only been there for the event, but still, it was one of the most favorite times I mean, hours I spent ever because it's such a unique place with this really I mean, everything, attention to details, the interiors. And in Swedish means a home.
 
And that's really you feel like you've entered someone's really luxury home. You don't feel like it's like you have the chefs standing by you cooking the food. It's like you're you're really in someone's home. They do that, and they have all these different types of guests.
 
So you have, like, an American hip hop artist sitting next to some older couple that that's saved up to come there.
 
They live on the countryside in Sweden, and that's really I think that's a great example of great hospitality where you manage to bring these people together and deliver such exceptional Yeah.
 
It's really experiential travel. In the way, I think Airbnb was thinking about doing something local where Airbnb said we will have local hosts and who will tell you about the city. And now they've just professionalized, and and I don't really love booking an Airbnb anymore because, really, these are professional apartments, bit soulless that I'm moving into. And then you get to this hotel Ett Hem, and you walk in, and you're greeted like, an old friend who's walking in.
 
And I remember when I arrived at that hotel, and and I said, are you still serving lunch? It was three in the afternoon. I said, I I'm so hungry. I haven't had lunch at the airport.
 
And she's like, any allergies? And I said, yeah. I'm lactose intolerant. And she just walked into this kitchen, which was literally there.
 
And then she just she came with food. And it wasn't a menu. It was just, like, at home. Like, they just had a nice lunch, and it was a beautifully prepared, all local ingredients lunch.
 
But it was such it was so smoothly done. It wasn't, like, you know, here's a menu. Here's you need to take payments, and it wasn't such a, you know, chopped up experience. It was just fluid.
 
And I think the fluidity of what Ett Hem has achieved is, in my opinion, the future of travel, but we'll see. But I think it's one of our my favorite hotels, and I'm going with my husband in July. I told him about it, and I said, you are coming with me to Stockholm. We're gonna go back to experience this in full.
 
Oh, that's you're lucky. I'm I'm gonna go some some day as well.
 
Love it. Okay. Ellen, thank you so much for spending time with me. I really enjoyed digging deep into some of the, the product features. Thank you.
 
Yeah. Thanks for having me. It was really nice. Thank you, Matt.