Good morning, everyone, and thank you for joining. I am Mike Tepeli, Communications at Groupon. Before we begin, today's discussion and management's responses to questions reflects management's views as of today, September 10, 2026 only, and may include forward-looking statements. Actual results may differ materially. Groupon undertakes no obligation to update these statements. Risks and other factors that could potentially impact the company's financial results are described in the company's SEC filings, including its most recent filings, Form 10-K and Form 10-Q. This is not an earnings call. Management will not provide new financial information or update the guidance issued on August 6, 2026, and comments will be limited to information already made public. Any non-GAAP financial measures referenced, including adjusted EBITDA and free cash flow, are reconciled in our earnings material at investor.groupon.com. This session is open to everyone, is being recorded, and a replay will be available.
With that, I am going to turn it over to today's host, Nick Nemeth. Nick, go ahead and introduce yourself.
Thank you, Mike. I think you have done a great job, and I am excited to talk to the CEO and CFO of Groupon today, Dušan and Rana. I am Nick Nemeth. I write on Mispriced Assets, a Substack where I talk about small cap stocks. Full disclosure for me, I am a Groupon shareholder, and I am bullish. I hope to get into the details of the investment thesis and allow Dušan and Rana to talk about the business in a way that is atypical to a typical earnings call or analyst call. Without further ado, if we click on Rana, click on Dušan. Welcome, gentlemen.
Hey, Nick. Happy to be here. Welcome, everyone.
Out of Croatia and I believe New York, Rana. You need to unmute.
Yeah. Sorry.
It is actually out of the Czech Republic, not Croatia, but the country which is the most popular vacation holiday place for Czechs, actually.
That is a great entrée because, just like I got it wrong as a shareholder and analyst, a lot of people do not know who you are, Dušan. Some people, certain circles. Can you give a little bit of an intro, talk about your career, as an entrepreneur, private equity guy, now a CEO of a publicly traded company?
Yeah, sure. Actually, the first e-commerce company which I founded was already in university. I was studying engineering, computer science, and math. And within that small company with my peers in the fourth grade of university, we built the first internet banking platform in Central Europe. It was a lot of fun, very wild environment. But the real big business where we got a lot of experience was called [Xactly], and I founded it with my Swedish partner. We had, like, 40 people, but we were able to build dozens of products. We were running the email solution Inbox.com, which at some point was rated as number two behind Gmail. We were running Spyware Terminator security application with, like, 25 million users. Ton of games, screensavers, like, a product factory set up, which taught me that you don't need a ton of resources to build something unique.
We were able to build this profitable without external investment. It was a life lesson for me, and since then, I was building new companies, typically with no or very little funding, which was forcing us to think very smart, find alternative way how to build stuff so that we can compete with much bigger companies.
How did you switch over to private equity? I feel like people think of you as a private equity guy, and in public markets, there's a certain understanding of what private equity guys are.
Yeah.
That intro, people are like, "Okay.
Yeah.
How do you move over to private equity?
Yeah. I do not really see us as a private equity because the core of Pale Fire, you can see it more like a family office. It is me, my partner Jan Barta, and David Holý, and a few other partners. When we sold our last big business, which we built, which was an insurance marketplace in Europe, which we sold for a nine-digit figure in USD. Then we started Pale Fire, and we built part of it, which was investing, typically majority stakes in e-commerce companies. That is why some people were and are labeling us as private equity. But I see us as entrepreneurs and investors. We are building the companies in a way that we can keep them forever. We want to make sure that they are operated well, they are built with very healthy backgrounds, and the systems inside.
We started investing also on the stock market, and Groupon was one of our investments, but we are very strong investors in many other companies, both in the U.S. and in Europe.
It is harder to be activists. It is certainly harder to be an operator. Why not just sit back and let your money work for you? What is your personality that says, "I need to put my hands on Groupon"?
Yeah. So actually, when we sold this insurance marketplace, I switched to the model of mainly investor. However, it is not me because all my life I was entrepreneur, I was building stuff. What I am telling to my peers is that I am short-term impatient, long-term patient. So I need to see that we are moving every day, every week, we do something new. When you are in the investor seat and waiting quarters, months to see things happen, it was really just not something which I would enjoy for a long time. So I am really combining it, and Groupon, for me, was an opportunity to jump into something new because part of PFC DNA is that we want to learn new stuff. We want to see people around developing. So for me, it was a learning experience.
At the same time, it was, and it is, a great turnaround story, which we did a few of those in the marketplace space in Europe. So the first few years, we were really rebuilding the company, pretty much saving it from bankruptcy, because when we joined, the company was not in great shape. Now I see another phase of all this, because when AI came, it became another new opportunity, another game which we can play here, and we are transforming Groupon. So I really get a lot of energy, and I was talking about my history when I was building companies. Anyone who knows it, you know that you are always waiting on something, typically engineering, but it is in many other areas because it takes time to deliver stuff.
Suddenly with AI, we have at our fingertips so many resources, and we can get outcomes very quickly. So this is extremely exciting time for me and for people around me.
Quite the challenge. There is definitely easier options. Switching over to Rana. Rana, really quick, how do you think that Dušan is seen in Europe and Czech Republic particularly? Because he is a little bit humble.
Yeah. In the Czech Republic, I think he's considered really a builder, a technocrat, and in a lot of ways a leader of, today, increasingly, frontier AI. There are not that many companies in Europe that are pushing as aggressively as we are. We have management and business in both the U.S. and Europe. But for Dušan, I think as an entrepreneur today, I think there's a real opportunity for him, and I see this where the leadership and the voice around what's happening in AI, there is a little bit of a vacuum in Europe from my standpoint.
Yeah. There's a vacuum for technology in Europe, and I think that you're also being humble. From what I've heard from Dušan, he's one of the pre-eminent investors, and would like to get into your history, Rana, because it overlaps with mine a little bit more from finance. Can you talk about the switch from hedge fund into, "Okay, I'm going to be a CFO of a turnaround story, and I got a lot on my plate"?
Yeah. It's certainly been a change, and it was certainly a contrarian move. The business was quite challenging when we showed up. The revenues were declining double digits, free cash flow was negative. The balance sheet was constrained. Most people I talked to when I told them I have this idea to jump, they didn't know Groupon existed still, and it was really a head-scratcher. But I had spent the past decade investing in publicly traded software and internet companies that had fallen on hard times, and I had built a muscle around analyzing the problems they were facing, and the risk-reward around whether those problems were fixable and what the benefit would be to shareholders.
Groupon was one of those that was always from the outside, it just felt like, "Why isn't this doing better?" It's a large and fragmented market, a brand everyone knows, and a highly attractive marketplace business model. The financial results told a different story, and that generally boils down to leadership and execution, and there are different levels of challenge. But this is one where it felt like you needed a full-scale transformation. I always thought about this as we're buying a house in triple A location, but it's a complete gut. It is a rebuild from the ground up. In my experience, looking at businesses that are faced with that situation, it doesn't happen by itself. That is when I met Dušan, and I call him the re-founder of Groupon. You've spoken a little bit about him and Jan's background through PFC.
They are our largest shareholder, and their track record, their skin in the game, and Dušan's readiness to really change his life to jump into this is what pushed me over the edge and convinced me that this is worth taking the leap. Dušan, he is one of these singular entrepreneurs. What I find is he is able to hold an uncompromisingly high bar on standards for everything that he does, while also keeping an extremely high level of intensity and pace. Listen, it has been a wild three years. The company that at one point we were not sure was going to survive, we grew for the first time last year. Making this change, getting off the sidelines and into the arena, it has been one of the toughest but also most rewarding decisions of my career.
Speaking to that, we are going from -25% billings growth when you guys inherited the company. The difference between that and 2%, low single digits, some people might look for, "Okay, it is not growing that much." I think investors are looking at this and looking at the discrepancy. They are like, "Okay, if Nick Nemeth writes about it on Substack and he talks about the opportunity set, why is it not being captured today?" I think that process is an accomplishment, and I congratulate you guys on that. I am sure it probably was just as hard or harder than you guys thought. Looking forward to the next two years, three years, what is the main objective of the company? What are we trying to get to?
I can take it through AI lens first, because for me, having the company which is able to run in the pace much faster versus where we are right now is also an answer on how the company should look like and how it will be able to survive and actually thrive in the future world. Right now, we are pretty much building, you can think about it like company Harness, company operating system, which completely changed how company works because I believe that successful companies in the future needs to be operated in very different way with access to all the data, when people can move very quickly, AI is supporting them. This is what we are building right now with our Project Foundry, which is the project I am talking about quite often. We put all the data on one platform.
We are building the knowledge layer on it, which is connecting all information about our merchant, about our businesses, all the meetings, all the communication, and it provides incredible level of intelligence to Groupon business overall, to our consumer layers, but also to merchant data and information about the merchant life cycle. Obviously, this is not an end game. This is really for me only a path to build an engine. Rana is sometimes talking about it in a way that Groupon, when we started, was definitely not a speed car or race car. But with the stuff which we are building right now, we want to have a car which is able to compete in Formula One league, which will be very fast, able to speed up, accelerate, decelerate, take sharp turns.
And with that car, we will be able to pretty much cover all the needs of our customers. Because when you look at it, we are not constrained by not having demand. We are operating on the market, which is over $1 trillion opportunity, but we are constrained by our capability to execute. And with what we are building, we should be able actually to execute much more projects to cover unmet needs of our merchants, unmet needs of our customers, cover all new ways of working, because AI is right now epicenter of everything. We don't know what will be future surfaces for AI, how people will be operating, and our role is to be ready for pretty much everything. We want to have a modern application. We want to have a chatbot which will be able to answer everything.
But at the same time, if the future world will be working through personal assistants, for example, Groupon needs to be ready to connect to all these personal assistants of the future and simply serve it. Because I don't think that our customers on the merchant side, small businesses, that their goal and motivation is to be AI expert and connect everywhere. We want to be a gateway, and we want to be sitting on the intersection of this AI future, crazy paced world and the regular small businesses who are fighting every day just to build some service, some product, and quite often just even survive.
Let's start from first principles of what Groupon is. It's a two-sided marketplace. What would you describe the value prop to be?
Every marketplace has two sides. On the side of our merchants, we want to be there for them and help them run their businesses better. If it means that they are starting the business and they want to bring new customers, Groupon has supply in this. If they are already up and running and they have great Fridays, Saturdays, but some days of the week, the utilization is not there, we can help them to cover their needs. In the future, we want to be closer to them. We want to be part of their operating system. And ideally, it should be pretty much like a slider in our ecosystem where we will tell them, yes, you can increase the price because you are sold out. And based on the data which we have in your neighborhood, similar services are sold for more.
If they are not there, we will help them also to just fill the empty slots. It should be super easy for them in the future. On the other side of the marketplace, our main belief in the company that the best things in life happen offline. We want to bring beautiful, great, huge selection of what people can do on our platform. We want to be a trust partner because quite often this is the main role of the marketplace, that you are going to buy something from someone who you don't know, but you trust the platform. Just think about Costco for experiences. You know that if you get products in Costco, they are good. They are at great value. There will be always this value piece because we want to provide great local experiences at unbeatable price, great value.
The Costco samples is actually a good thing I wouldn't have thought about, the experience in the store itself. So we have a value prop where businesses are trying to get customers are trying to get deals. That is a value equation, slightly different on both sides. The way I see it, I'm not sure if you guys agree with this, is if you want to get new customers as a startup small business, I know you guys do enterprise stuff, but you're typically trying to develop a strategy on Google, right, or some sort of ad strategy. The benefit of Groupon is you know the customer's coming in the door if you're going to pay. How do you see the competitive dynamic for that value proposition leading into, you can do it in this question, or I can follow up, what is the platform?
How does it maintain itself? Where does it plug into AI, and how does it go? Let's just start with the first one of what is the further value prop. It might not be another coupon deal value related platform, but what are the alternative options and how is the competitive dynamic today?
Rana, do you want to take it or should I?
Yeah, I would say a few things here, Nick. On the merchant side, this is one of the things when we came here, we had some views at the top level, but when you get into the details, different categories have different unmet needs. What we see is that as we invest in solving some of those unmet needs, they respond well. For example, I saw a question in the chat, how do we know North America's getting better? How I look at it is we have a cross-section of cities and categories that we compete in, and you look at Things to Do. We have been growing Things to Do quite strongly for strong double digits for several quarters now. The unlock there for us was we needed to really improve the purchasing experience and the redemption experience.
We enabled that through doing connectivity and bookability. Making those investments allowed us to serve those merchants better, which allowed us to get more of them onto our platform and stay with us longer. Then we were able to offer that selection to the consumer side. This is where these marketplaces get to be really exciting businesses. But it is tough because for us, to solve the marketplace, you have to solve it at the level of where it happens, which is the customer and the category, and the merchant. What I see is there is inconsistency. We have some cities doing great. We have some categories doing great. The other side, we have some cities and some categories that are struggling.
This is where building the marketplace model, that executes at the level of the neighborhood on what the user and the merchant is trying to transact. This is sort of how I see how we can unlock more of the value proposition. I think on the AI piece, I would probably ask Dušan to jump in.
Yeah. Let me reframe the question for AI. The bear case is AI is going to eat software. You are just going to talk to a chatbot. Chatbot is going to know everything. It is going to be the example. Groupon as a marketplace today, how does that become the infrastructure for the future of AI, where chatbots are certainly going to be used, whether on Groupon or in-browser. Potentially there is going to be agentic, next level Alexa, robots. How does Groupon, just wherever AI goes, maintain its value prop, and what is the company doing?
Yeah. I actually think that this is extremely exciting time because the development and pace here is insane. In the end of the day, I believe that it will actually strengthen the demand for experiences and for the deals which we are selling. Groupon simply wants to be everywhere. We will keep building our own application, our interface for the website, desktop, touch. We will have, and we are already piloting and testing the chatbot, which will be available across the board in our interfaces for people who prefer to talk or to just chat about what they need for Groupon. However, we have several initiatives for the underlying technology because in reality, we are kind of a system of record for these deals. We have unique deals with merchants.
We do the transaction, and we just need to be in places where people will be buying stuff. This is not only about AI, because there are employer reward programs where people are going because they have some benefits, and Groupon is not there right now. It's AI personal assistants. There are credit card loyalty programs, and this is all I see as a new opportunity for Groupon. Right now we are building a platform which will be completely open and in, I don't know, today or tomorrow, for example, we will announce access for AI builders and for people who want to build smaller local communities. AI-enabled system where you can, with your AI, build small website for your community, local club, which will be showing just deals in New York. It will be affiliate-based so that you can make some money.
But the same technology allows us also to distribute our deals to other platforms, whatever the future platforms will be, and also connect us to future AI personal assistants or whatever the surface for people to place transactions will be.
If we're looking at the airlines industry, it's pretty consolidated. The deals, especially for small businesses, that value needs to be uploaded somewhere. Is it going to be uploaded to the company website and then AI's going to burn a million tokens figuring out what is the best deal for this person? Or is there going to be a home for it? I think that's the long-run opportunity set. As we look past, like this business is still, in my view, priced like it's dead. And the fact that it's not dead should be acknowledged. But the opportunity set going forward is where I get really excited. As an investor, as an analyst, seeing something that is so far off the valuation for what I see as the probabilities going forward. When you're talking about that data infrastructure, can you possibly give some color?
Do you recognize the same things? There's got to be some infrastructure for deals and value, whichever way the AI train goes.
Yeah. I'm absolutely on the same page. I see it really as a huge opportunity because many other companies, they are just selling same deals or same products as what their suppliers have. For that piece, I see AI ultimately quite a threat, while Groupon is building different deals, different value proposition, especially for smaller businesses. We have a very unique product, and by opening our system so that we can serve pretty much whoever will be driving demand in the future, it's actually a huge growing opportunity.
From 2019, 2020 nearly killed the business. Obviously, it's going down. But if you go back to 2012, 2010, everyone had to be on Groupon. A return to 2019 levels, any investor today would probably be excited about that outcome. What is the marketing push on the road if it's correctly identified everyone needs to be on Groupon, everyone needs to check Groupon. If it's not going on Groupon's website through ChatGPT, which has a Groupon app, I'm assuming is part of the potentialities as well as just integrating with these model providers. On the road to that, if we're shifting to marketing, what is the new game plan? Anything you've shared publicly on performance and strategies there. As Rana opened up and said, everyone knows Groupon. They just think it's dead or whatever they're thinking.
I would caveat one thing I've heard, and I think you guys recognize, is there is a trust issue specifically after 2020. Opening it up to the marketing strategy, what do you see as that strategy and how is it driving financial performance today?
Nick, maybe I will do one step back and before I will be talking about the marketing strategy, I would also like to mention that one of the reasons why we need to build this Formula One car engine is to have better capability to cover unmet needs. We have products right now on Groupon which are serving well to some merchants in some categories, and we have categories where we still have an opportunity. One of the nicely growing segments is Things to Do, where, for example, what we are quite often doing is that we are connecting our merchants at the best possible price which they offer, but it's not a unique price typically. We just connect to whatever they operate with. Then we understand what are their needs really, and we build special products.
For example, we find out that they have empty afternoons, so we build something unique on top of their proposition. This is a muscle and capability which we need to bring across the board because this will give our merchants a reason to be on the Groupon all year long, even in the periods where they simply don't need anything special. They will be part of our inventory, and then they will have time areas, days of the week, or simply they will need some change in the pace of the business. We will be where we can boost the business for them. This is still the piece which we are building, and I expect that we will be bringing many new features, options, products for our merchants. On the marketing piece, there are several elements. We were talking about influencers recently.
Actually, it was also one of the theses which you were coming with to Groupon. Everyone knows Groupon is true only for people over a certain age. I am not sure that it applies for people who are 20 or 25. The reason is very simple, because Groupon was not really growing last few years. That huge boost you were talking about was really over 10 years ago. The new generation is simply searching and discovering products on the internet in a different way. One part of our marketing effort which we need to do and we are working on is to come back to influencers, build a strategy for TikTok, Instagram, and other social networks and show them the product.
Even here it means that we need to be working on the product and improve the product because young generation doesn't want to pick up a phone and call to do the booking. We need to have much better connectivity, and that's why we are investing into connectivity to our partners to make it very seamless, very simple. Then in the middle of this is the piece, and you also touched it and I consider it super important, and that's trust. It's not directly marketing, but our experience is that marketplace works only when there is trust on both parts. Merchant trust us and consumers trust us. Let's be honest, it was not always the case. With Groupon, the motivation was quite often just to generate revenue, get the deal sold, and that's pretty much it. We are changing it.
I was talking on earnings call and in several other communications afterwards about our bet which we are making on trust and quality. From the short-term perspective, it may have negative financial impacts because we are taking down deals from our marketplace. We are actually refunding people more than before when they are not happy with the service. Obviously, we prefer to refund in Groupon Bucks when they do another purchase. We are going to merchants and telling them, "Well, you need to change something here because this is an issue for your customers. That's why they are not coming back." But with this, I see, and I measure it in very simplistic way on top of what the team is showing to me.
I'm counting number of emails which I am getting every week from our customers who are complaining that, "You did this, you did this, and I don't think it's fair." When I started in Groupon, and I think that number was very similar even last year, it was 30+ conversations every week. Giovanni Lozano, who is running this bet with us, he was asking me, "Is it true that last week it was just two?" Yes, it's really tenfold. We decreased the number of people who feel that they are so desperate that they need to find a contact on the CEO of publicly traded company to just show what we do wrong.
That's part of our bet, part of our mission, and it will take time, but it will bring this compounding effect that people will see that we are doing good job for them, we are delivering value, and then they will be coming back, and it will be helping the purchase frequency and marketplace flywheel.
Hey, Nick, I want to make a few additional points here because I think this is an important point that allows us to communicate a few different things within the example Dušan just went through. I think one is you commented about 2019, 2012. I think that the big picture from my standpoint and the seat I'm in is Groupon came around at a time when small businesses were dealing with a shift in the way that consumers went to discover what they wanted to do offline. At that time, it was offline print, TV, radio, to online, and internet. They didn't know. It's what Dušan said earlier. Typically, the small business or experienced operator, they get into the business because they love that experience. Or they're trained to deliver that, or they're hospitality-oriented.
They're not digital marketers, so we played a role for them. Then our business, we got defocused, we got big, we tried to compete in a lot of different playing fields, and we stopped focusing on that customer. We started making decisions that were reducing that value proposition. We didn't evolve with them as the market has evolved. You fast-forward, the same thing is happening now, okay? Small businesses everywhere, they're some of the smartest people you will find, and we consider them our partners, and they're reading, they're like, "How are we going to solve this AI thing?" They know that traffic is moving, and they want to make sure that they can be discoverable. To me, listen, I don't really look backwards. This is one of Dušan's gift.
He's always about like, "Let's go forwards now," so how do we improve from here? I think the big picture from what made Groupon great in the past, I think those conditions are on the playing field today. Then it's a question of what are we going to do about it, right? How are we going to execute? This is where I would just say two small things, okay? On this trust and quality piece, which Dušan's hitting on, it is a great example of how we're deploying the company Harness at scale, where every day now, and the team that's running this under Giovanni is one of the most AI-native teams we have. They are able to see every morning, here's all the customer data, here's all the complaints, here's all the feedback.
I went to get an oil change, and that guy charged me $5 extra. It wasn't in the fine print. Right away, they're able to decision, okay, what do we do about this to solve this for our customer? Doing that at scale across tens of thousands of small businesses, across millions of orders, it's not something that our company's ever been able to do, and it's one of the things that AI and AI-native experience operating teams are able to solve. The second thing is performance is the marketing funnel, and this is more of the DNA of our company that we're building, which is we want to go step by step here. A different CEO may have come in and said, "You know what?
We're going to go put $100 million on a big brand campaign." But the way that Dušan has always built businesses is from first principles, step by step, building the foundation. For him, he wanted to make sure we optimize the people that are lower in the funnel, who are already searching for these things, so we can make it easier to convert and have it be a more profitable ROI. As we improve the trust, as we improve the product, then we are frankly earning the right to spend investor capital in mid-funnel and upper-funnel strategies to unlock the latent potentials we see in this brand. I think this just gives you some flavors for not just how we're executing on these topics, but just how we think about transformation here bigger picture.
Small businesses spent 15 years learning how to use Google Ads, and you are saying that now they are like, "Oh, there is a new paradigm," and it is the same sort of wave. Now to the we, you guys came out in 2023. It was lefty for a while. The business, you are recognizing that you have to own the brand, despite if people dropped off thinking about it in 2016. So you are talking about sort of a more full-stack approach of thinking about the value for businesses. When it comes to consumers, you mentioned a $1 trillion opportunity, Dušan. Can you just identify what that is? Is that experiences fulfilled online?
This is the third-party data around what the local economy for experience and services is globally.
Okay. Awesome. On the marketing piece, I want to get into some of the financial stuff. Actually, stop. You mentioned Giovanni. What is the talent acquisition like? Obviously, you have mentioned you are cutting costs, but also, I am trying to do the math on the model, right? And where the cost is going. You mentioned, but we are using this money to hire new people. I think specifically, you said talent density. So what does that look like over the past two years?
Yeah. When I joined Groupon, we were not able to hire people in the U.S. because everyone thought this is like a dead horse. So we were hiring people in Europe, mainly based on a Pale Fire reputation, my reputation here, and we were able to build really decent team who did great job and really saved the company. It was not Czech Republic only. We have really great team in Madrid, in Spain, in Valencia, in other countries, in all over Europe where Groupon has offices. Then when we switch to the new phase with AI, focusing us to be AI native company, coming with different stories, it unlocked also the different level of talent across the world, which we were not able to get before.
Suddenly we are able to hire people who have Uber experience, people who have Klarna experience, people who have DoorDash experience, and it significantly helps. At the same time, it fits into the strategy Groupon. If you work in Groupon for one year, I think you get more experience in any other company in five years. Internally, what we are talking about is we want to have speeding tickets, no parking tickets, which means that we are unlocking paths, growth to people everywhere. This is ideal environment. If you are young, hungry, smart person, come here. In one year, you can have five different jobs, work in three different departments, and end up with finding something which will be your passion, and you will have their P&L responsibility, which you would never get in traditional corporation.
I love this part because when I really see someone young who is new here and I am coming with some idea, and I think, yeah, this is pretty good idea, and that person comes and tells me something better, I say, "Yes, this is what we want to build here." To have people who are coming with great ideas are learning. Now how it moves to talent density. I think this company culture, performance culture is and will be attracting more and more people. Not only people who are senior, but really we want to have more with hungry people who want to grow, be with us, be part of this experience. This environment is also attractive for senior people who are hungry. I don't want to adhere just the team of people who have experience from whatever great company.
At the same time, they still need to have this high agility. They need to have a mindset that they want to learn something new every day. I see that these people start coming, and they are coming not only in Europe but also in the United States. Giovanni Lagasio we were talking about, he is in Europe, but he has Klarna experience, Uber experience. He brought several people, was able to hire several people, onboarded and find some internal talent in Groupon. What they did with customer support in few months, it's amazing. They completely reset the expectations, raised the bar, and is one of the best performing AI teams in the company.
On the other hand, we have Adi Rajkumar, who joined Groupon as COO, DoorDash, other great experienced guy who is bringing life experience, and at the same time, he is hungry and bringing that piece because he was part of the DoorDash growth. So very relevant. Local marketplace, successful company playbook, and the person who wants to do something similar or better at the same or bigger scale with us. So amazing. We have more and more examples like this.
Nick, as from the investor hat, you may say, "Okay, well, why does this matter? Why are you guys spending money attracting best-in-class talent?" To me, this is where I think in order to leverage the full potential of AI, you need best-in-class AI native operators. What we see here is the speed and let's say the standard deviation of team speeds is increasing as people adopt AI. There are some engineering teams, there are some operations teams, there are some marketing teams. The speed they're moving is incredible. Then we have other teams that are frankly behind our expectations. The single ingredient that unites this is talent. For me, it's great. I've been really I track our talent cohorts every six months, and I see ours getting better and better at this.
What I see as a common thread amongst all the people who are joining, it's exactly what Dušan said. They see ride-hailing maturing, consolidating, food delivery maturing, consolidating, hotels maturing and consolidating. They know how to run experienced marketplaces at global scale within the complexities of solving it at the local level. They're looking where can we go take this into the next chapter? The areas we operate in are still secularly under-penetrated. There's some third-party research that says for Things to Do, only a third of Things to Do experiences are still booked online, and that's crazy to imagine, but that's the sort of the ingredients that could allow for secular growth and, for us, market share growth because we know that we're not doing our best.
We found that that resonates with really ambitious people who are talking about, "Well, I've seen a company grow multiple size times the GMV." Today, we're not a huge company. We're $1.7 billion of GMV, but the market potential to take that multiple times higher and then the marketplace business model, which has great characteristics, that is what gets people really excited to come join us on this mission. So, I as a finance guy, I'm very much excited about underwriting talent density as one of the key themes for us to build this platform to make us execute and realize the opportunity we see in front of us.
Nick, maybe I would slightly double down on what Rana said because you can read it quite often, but we really see it, and it's really amazing. In the past, you could use brute force because your best people were 2x, 3x versus the average population. That's completely gone. The landscape is different. Now it's really time when you see 100x-ers, and I am super convinced that successful companies of the future will be around these people. Our motivation is to create an environment that they will be able to grow. We are here, some top management, to unlock all the blockers.
Quite often, we have people who are coming from other companies and telling us that they were not allowed to use AI fully, they didn't have access to data, and when they see what we allow to do them, they are just happy in this environment, and this will be one of key drivers for us going forward.
The costs have gone down, headcount has gone down, average salary has probably gone up. If we're looking at the marketing component of it, that's $40 million in the last quarter. Where have you said publicly that you can say? Do you see that as a percentage of revenue going? If you could just describe it and color it in as sort of what. Because I see a mix shift. I see you're cutting SG&A and other costs, but that is growing, albeit down from three years ago significantly, but it's starting to grow again. What I see on the financial side is, well, and you see this in startup companies a lot. You guys are acting and appealing like a startup company that's publicly traded with $225 million of cash. If you are buying growth, it's not necessarily growth.
What are investors looking for when it comes to the percent of marketing, and what are you seeing on the ROIs of your paid channels, and how do you expect that to develop over the coming.
Dušan, you want to take the philosophy, I'll do the numbers, or how do you want to do it?
Yeah, so start with numbers, and I will give them the big picture.
Nick, I think the way I see the business is if you look at it at a customer level, and we have approximately 16 million customers that have bought with us in the last year. When we got here, what we see is that we have a core base of loyal customers that are, we call them our champions. We also have regulars. The champions, they are buying, let's say, five times plus a year, and they really are getting the value from the proposition. What we observed at the time was we were not really doing new customer acquisition. We had to fix a lot of things. This is where you are building the house, or the other analogy is fixing the plane mid-flight.
We will be going back to first principles, what do we want to solve under our CMO, Josef and Dušan, we have built really, in many ways, a best-in-class acquisition machine in the lower funnel. Over the last couple of years, we have largely solved acquiring customers. Now, there is a lot of opportunities for improvement, don't get me wrong, but we have shown that we can acquire new customers at really attractive cost of acquisition relative to these champions' and regulars' lifetime values. Where we have struggled, and we are very open about the struggle, is that first purchase converting it to that, and that new customer converting it to that second, third, fourth customer purchase and becoming those regulars.
When you go, "Okay, well, why can't you just solve that?" It starts to go back to the underlying issues around systems, processes, people, and really, what do we need to solve? We need to solve the customer experience, the onboarding experience, make sure that they are able to properly get what they want, and not be inundated with noise. This is for us, when you are asking the question, where are we in marketing spend? Largely, it is a dial, and we can spend what we want in terms of choosing what we want in the current Pareto curve. What Dušan and I are most focused on is how do we push out that curve?
The unlock for that is building better life cycle management around that first purchase to that champion. For that, once we get to that, if you think about it, just what you see is the spend as a percent of full revenue. The second and third and fourth purchase, the incremental margin on those is much higher because you are not spending that same acquisition budget. A lot of times they are coming to you directly or through your managed channels. That is the way we get leverage on the marketing. I think from an investor, it is a great focus area because we have a lot of operating leverage in our business. The way we unlock this operating leverage and get scale on our marketing dollars is fixing this area, and it is one of our highest priority projects that we are running AI natively.
We have some stuff out there in social around what we are doing around personalization, amongst other things, and this is something that we will be updating investors every quarter. I think we have a lot of experiments. I am not ready to declare victory on it, but I am very encouraged by the progress, and we will keep you updated on where we go with that. Dušan, do you want to add anything else?
Yes, I would like to. As a part of fixing this acquisition engine, when we started, we did not have even visibility into data to really understand what is the difference between one and second customer about the deals. The engine which we have right now, we understand which campaign is acquiring what profile of customer, what is their lifetime value, what is their purchase frequency. So finally, without AI being able to dig so deep, we would never be able to build it. We have tens of thousands of campaigns. We understand whether we should be spending more on this campaign because the customers have 2x, 3x purchase frequency in a year or 1x, for example. So we optimized it for short-term ROI. I really think that this is, if not best in class, one of the best performance acquisition engines.
Now we are switching it that this is not just acquisition engine working with short-term 7-day window, but this purchase frequency element, which we are talking about everywhere now, internally and also externally. We already have it as a part of marketing, and for me, this is extremely important. I was explaining in the beginning that when I was building companies, I was not getting external money. It was bootstrapped and all that stuff. I think that I have quite a good feeling when we are ready to invest. I would never invest and do channel splash campaign if I do not believe that we have unit economy and we will create a value. So I will not give you a percentage number, but when I will see that this acquisition purchase frequency engine works, then I would like to invest as much as possible.
Now we need to run it with some hygiene. The number is moving slightly because I really believe that the trend which we have here is good. But once we will be ready to come also in front of shareholders and actually talk about huge increase in spend, it will be, for me, one of the best moments because my internal trust that we crack the purchase frequency at that moment will need to be close to 100%.
That is one of the things I went activist on. As an investor thesis, I am not talking about user growth. I forgot to make the comment that I dated myself saying everyone knows Groupon, and now I am on TikTok, so messed up a little bit there. But if we are talking about 16 million users, just make them go further. Right? In order to do that redesign you guys are doing, recommendations. Recommendations were awful, no offense. Now they are starting to get better. And the friction to the purchase. All of that, solving all of that, simply you pay $30, $50, whatever the number is, for a customer. The first purchase maybe you lose money, maybe you break even, depending on the channel. But it is really about the second customer because you have 90% gross margins, right?
Yeah.
Then we can talk about growing users, we can talk about potential unlocks for businesses being a value CRM of sorts. But can you talk about the operational leverage before we wrap up, Rana? Talk about when this business grows, not low single digits, but 10%, 15%, what does that mean for financial performance? Do you think we are seeing the operational leverage today? What should investors that are trying to do financial analysis pay attention to?
Yeah. I would say two things here, Nick. One is that I think you described some of it, is that just understanding our business model. We have 90% gross margins, we have around 50%, 55% contribution margins, and we have a fixed cost base which we can leverage. Now, when we got here, we took down those costs so we could get the business healthy. But from there, our view is we still have quite a lot of opportunities on the cost side, but we also see a lot of things we want to invest in to accelerate. For me, the big picture, how do we create a lot of value here and are all of us around the table, this is our singular focus, is we see that if we can get the engine running and what you are talking about is how we see it.
This is a large market opportunity. This is not a time for optimization. This is a time for driving billings growth in the right way with customers who are onboarding and engaging and repurchasing. That business model has the opportunity to leverage a fixed cost base. When you start dropping that incremental purchase at 90%, or it's not exactly because we may have some remarketing costs associated, but it will be very accretive. You will see that EBITDA growth will be faster than revenue or billings growth. For me, this is where it. It's tough for investors because it's like, okay, it makes sense. Well, show it to me. It goes back to first principles. You can't just do that until the marketplace is healthy and operating with the right flywheel. For us, we're not ready to guide to 2027 yet or anything like that.
Our principles, you look at how we've managed the business over the last three years, and that gives you a sense for how we will manage it going forward. We are people who will invest against where we believe that we'll have conviction around there's a good investment opportunity. We're not always right, but we're also not spray and pray type management team. I think that gives you a little bit sense of the operating model.
Yeah. I also think they don't understand how much more EBITDA or free cash flow goes up given 10% billings growth. Free cash flow goes up higher than EBITDA.
Well, Nick, like I said, you can look at our past. You talked about the negative growth rate when we joined. You also look at our EBITDA. Over the last three years, we have shown not just improving growth rates, but we've also taken our EBITDA up as well. That's where finding the balance for us is over time. Our number one KPI is really inflecting billings, orders, purchase frequency, and over time, active customers, like you said. That's the primary goal for us. But we're highly confident knowing the underlying business model of how marketplaces work, that when you solve that, we can turn this into a cash machine, and it is already generating cash.
Yeah. As a percentage of market cap, for sure. If the company grows 10%, people should model it out, figure out the probabilities. There is risk in every investment. You are dealing with guys that got into the fire. I think the financial performance says they righted the ship, and the question is, can you execute? I am very excited to see this process through you. I really appreciate you guys talking to me. I get some questions from hedge funds like, "Why are they doing this?" I just point to the Goldman Sachs TMT Conference, and I look at the JP Morgan Healthcare Conference, and the thing is that those sell-side analysts, they write milquetoast vanilla. They do not put their money at stake, and quite frankly, it is usually not worth reading.
I think it is forward-looking for you guys to engage with a Substack author, which some people might pejoratively say, but I am quite proud of, and I hope that investors, potential investors, other entrepreneurs, enjoyed this call. I think that it was great. We clearly went way over. I really appreciate it. Thanks.
Hey, Nick, thank you for joining us. Listen, I cannot speak badly about my sell-side partners or institutional investors. We enjoyed the opportunity to come tell the story. We have really enjoyed the feedback we received from you and other engaged shareholders. Like Dušan, based on how his Pale Fire business is run, this is really for us, we look at this as a partnership between all of our investors. We just happen to be active managers, but we are really shareholders along with you guys, and so we are always open to feedback in any channel and a conversation in any channel. So we really enjoyed that. Anyone who has feedback, please hit me up on X, and we will continue to improve from here. So thank you, guys. Dušan, anything you want to say in the end?
Nick, really thank you for inviting us here. It is a great format, and it is great to hear and get feedback from small shareholders, all shareholders, actually. I would just mention on top what Rana mentioned, it is a great communication channel also for our team, because this is really the first time when we are getting ton of direct feedback from you, from so many other shareholders. Our teams internally are really trying to show what they are building. So you actually started with the wave of openness also in Groupon, when we are proudly showing what we have. We have ton of great stuff inside. We need to do better job in selling it and presenting it to everyone else in the outer world. You really ignited that, and thank you and all the people who are participating and who are shareholders of Groupon with this with us.
Thank you.
I appreciate it. It's a different shareholder base. It's not so focused on the next prints and what real-time data says. Last question I have for you guys. I've gotten, like, a dozen questions about an internship.
We've got to run, Nick.
Is there an internship policy? I mean, program.
Well, listen, we get ideas every day. We do not have one formal, but we have a new Head of People here, Nick Walker. He is great. I am sure I am happy to take it offline with him and you and see if we can. The story for us is clear. We want ambitious people who are super passionate about the mission we want to serve and want to become AI native, best-in-class, experienced market operators. If you feel like you fit that boat, we would love to have a conversation.
Awesome. Thank you, guys.
Thank you.