DraftKings Inc. (DKNG) Earnings Call Transcript

Okay, everyone. Time to keep going, and this is obviously now our kind of lunch keynote time. So we have a little bit more time for lengthy conversation. It's my pleasure to welcome Jason Robins, Co-Founder and Chief Executive Officer of DraftKings. Welcome.

Thank you for having me.

So we didn't get any informalities on the side. Last year, if you don't remember right before this event, I think you were coming in out of Miami with like a couple of hours of sleep. The loop was like lighting up.

Yes, 1% recovery, remember.

I think that's right. So where are we at on the recovery meter for 2025?

A little better today. I was here for the whole night and didn't fly in on two hours sleep this time. So it was much better.

Awesome. And the other claim to fame was the jacket last year, which...

Not this year.

Not the top drawer, but there was a DraftKings monogram one-of-one series jacket, very subtle, but a great time.

On the inside, yes, totally.

So look, thank you again for spending the time with us. If I could lead off, I definitely wanted to start pretty high level, Jason. And like what -- I field so many different questions around, let's call it -- what I'm going to call, the algorithm of growth for online gaming broadly and online sports betting in particular. We go down these paths as Wall Street modelers of things like handle and hold and promo, and we're assigning all these variables. But I thought it would be helpful just to kind of level set for us a little bit of how you think strategically about the business? Because here we are modeling these KPIs as if they exist in a vacuum. But I doubt that's kind of what you actually do on the inside. And I'd love just like how are you actually talking to the product teams when you're walking in every day about driving the business forward? How are you working on those variables kind of working backwards? Or how do those become outputs to what you're working with on your teams?

So I mean, I think there's really at the simplest level, there's two things. There's how many customers do you have and how well are you monetizing them. So we tend to kind of look at it that way. And for example, I think you're talking about like handle and all the other things you can break down when you look at revenue. We don't really forecast in that way. So our team will have technically, I guess, a handle number that they have in their model, but it's not actually how they're thinking about managing it. They're thinking, I need to hit this revenue and gross profit number, I can do that by driving various levels of handle, hold or promo efficiency. And those are the levers that I basically have to pull to get there, and they're trying to land the plane on that number. So when you are trying, for example, as we have been for the last couple of years to really push up your structural hold rate, driving more promo efficiency, it's going to have an impact on handle the same way as in previous periods, if the market was promoting more and there was more of a focus on that, then the handle will be higher. So they all kind of move together, and we really look at revenue and ultimately gross profit. And then obviously, we look at customer level metrics like how many active customers do we have, how are they engaging? Are they betting multiple sports? Are they playing with the level of frequency that we expect and those types of things.

And when that translates into your conversation with the financial team, when you're kind of getting around the roundtable before the quarter, like what's your level of -- kind of how are you interacting with them the KPIs that kind of Wall Street demands? Like what are -- are you -- as a business leader, are you really focused on sort of that revenue outcome? And then help us walk through the P&L a little bit around just what's the cost messaging that sits underneath that? How are you kind of thinking about growing your business and balancing that with trying to drive some scale and some operating leverage?

Well, I'm actually -- I think, rare that I feel as excited as I am about the cost outlook right now. And the reason why is I think AI is transforming the way that we are going to potentially need or not need to grow our fixed costs going forward. There is some investment. We are going to be hiring more AI engineers. We are deploying some new tools. But the amount of efficiency, I think we are going to generate on headcount and being able to basically replace what would have been human hires with AI agents and also reduce in certain areas as well. I think over the next few years is going to be a big thing for us. So that's something we're very focused on in terms of the cost side. As far as like more on the revenue side of metrics, we're really looking at, obviously, again, making sure we see healthy customer metrics, and we feel good about engagement. We're really looking at how well we can monetize the customer and trying to improve that. So it's a continual push towards how do we make more gross profit per customer and ultimately more LTV. So that's really the way we think about that. LTV, obviously, depends on what time you're measuring it over, so it could be a tricky thing. But we've gotten our models to a point now where for an existing customer, I think we can really accurately predict what we expect that cohort of existing customers to do, not maybe a single customer, for a cohort of customers. So even when we acquire new customers, we very early on have a very accurate projection of them. What's a little trickier and sometimes harder to forecast is the actual number of new customers coming in. As you remember, probably a year ago, we got caught by having really strong customer acquisition, which was good for now this and future years, but made the promo run a little high. So we've obviously adjusted. That's like a little bit less predictable, but to actually know what a customer is going to produce once we acquire them, we've gotten very good at.

So what is the balance of those two metrics before I now hammer you on the Wall Street KPIs? But before we go there, like what are the balance of these two? Are we still seeing new customers enter the funnel? What's your kind of conversion rate or your target on these people? And do they spend as much as the existing cohorts? Or is that golden cohort 1 and you're seeing some sort of degradation in spend as you get -- as you pick up that incremental customer today?

So the way I would think about that last question is there is definitely a golden cohort in the first year or two. Once you get past that, it's been essentially the same profile of customer that we've been consistently acquiring. And I think the improvements that we've made to the product are actually monetizing some of these newer customers better than ones we were getting a few years ago. But it doesn't look like there's this continual degradation even though for sure, there's a golden cohort in that first year or so that you get. But then it kind of asymptotes out really quickly, I think. And it is hard sometimes to distinguish between the customer and also the products improved a lot, too. But if you're just kind of looking at it as apples-to-apples as you can, that's the way I would describe it.

So again, now let's kind of think about the kind of Wall Street output of this, which is we do kind of break down all these state-level numbers in granularity. But a conversation with actually Alan really sticks out from June where we talked about this balance of -- when we started digging in on handle growth, which was sort of an obsession kind of walking out of the first quarter going into the second as that trend was slowing down. And there was a point made around, hey, look, we can incent these behaviors, right? We can drive a lot of handle if we give XYZ promotion. I believe the discussion we have was all around your sign-up bonus in your -- for Major League Baseball and they kick off of that season, which was very handle-friendly a year ago, might be much more hold driven a year later. Could you elaborate or discuss that a little bit? Because I think it's a really important message for everyone.

Yes. I think early on in the industry and not that we aren't doing this at all still, but much more so like relative to where we are today, we were more focused on handle growth and less focused on driving up structural hold and promo efficiency. Over the last couple of years, particularly in the last year, we have made tremendous strides on structural hold and promo efficiency. And that has come, I think, with some slower handle growth. But ultimately, we -- like I said, we don't even really forecast or look at it that way. We look at it as, okay, I could hit a variety of handle numbers, but then my hold would have to be this and my promo would be this, and we're trying to ultimately maximize that whole equation. And right now, we are finding that we are getting so much value out of focusing on structural hold and promo efficiency that I don't see that really being something at least in the near term that changes in terms of focus.

In each of your shareholder letters, you give a bit of an update on the structural hold in terms of what you've done and particularly in the last year, really the last couple of quarters, major multi-hundred basis point improvement in parlay mix. Is that the key variable? And sort of what's driving it? Just kind of help us under the tent on the product side of amazing to see the outcome, but how are you doing it? Is it merchandising? Is it a certain type of player that you're able to find and really target, retarget with the right promo boost? What is it that's working to be able to drive this? Because we're not young anymore from the standpoint of this isn't 1 year out where you're learning and evaluating. To be this far along with these customers and still moving the entire ship by hundreds of basis points actually seems like a lot.

Yes. I mean I think that -- well, first, it is largely parlay mix. Obviously, we're always improving our models, too. But the big driver of the structural hold increase by far has been more parlays, but also more higher-margin singles bets, more legs added to parlays. So not just purely what percentage of bets are parlays, but all of those things, higher-margin bets, so to speak. And so that is the focus. And I think it's not really one thing so much as like a combination of how you merchandise, market, promote, what types of products we're focusing on. So whereas last year, as you noted in Q2, we ran a very activation-heavy promo that was singles bet focused, which did fine for activation, but didn't -- singles bet focused, so actually didn't drive anything. In fact, it took us the other way on parlay mix. This year, start of NFL, for example, we're rolling out Stacks, Ghost Leg. I mean those are the product features we're launching. Stacks is a narrative-driven parlay builder. Ghost Leg is a promo where if one leg of your parlay misses, it goes away as though you never did it and you still get credit. So that's where we're investing our product dollars, our promo dollars. It's how we're using our marketing channels. So seeing people making picks now that aren't singles bets and are always parlays whenever we do any kind of influencer or talent-based type stuff. And just really orienting the whole thing around that. And I think what it shows is it's not really a type of customer as much as a behavior that you can drive. And I don't view it as there's more customers of this type or that type we need to attract. I think parlays are fun and anybody can get into them. You just have to get them to do it, and you have to give them reasons to be interested in trying it. And so that's what we're doing now. And I think that comes across all those things. But the best way to summarize it is just focus more than anything else.

So we are on the eve of football. We're going to get into some of the product innovations shortly. But it also reminds, Julie and I as we start writing all of our notes about weekend outcomes because individual games start to go up in matter, it's going to -- there's going to be a little bit of that Sunday night [indiscernible] about what the matchup is, which I think is a quite big one on this Sunday. So what this brings to us, though, is volatility, right? As parlay mix goes up, and we're talking about driving hold as a major important outcome, when parlay mix goes up, there is the inverse of when those hit, especially if there's concentration into a handful of bets on parlays, volatility also goes up. So how do you think about that? Like is that something you even care about? Or you're like, Shaun, this is math. And when I look at expected values, what will happen or should happen in mean revert over time, we're super comfortable here. How do you think about kind of managing that volatility as a bookmaker, which is a very basic part of your job?

I mean the only reason I really care is some of our shareholders care. But my point of view is more what you were saying, it's math, and it makes sense to do whatever generates more long-term expected value as long as the volatility can be managed and the risk can be managed. So never in any situation, are we taking on risk that is not manageable. We have lots of controls in place for that. That's why even when we have a horrible quarter in terms of sport outcomes like what happened in Q4 last year, it's not like we held 0. We still held pretty close to 10%, right? So I mean it's -- I think it was like 8 or something percent. And so we're not going to be in a position where we're going to risk capital that we can't afford to risk or do anything like that. And once you kind of cross that threshold and you're like, okay, like I'm managing the risk side of it, I think you just want to maximize EV and maximize long-term value. And I also think part of why -- and this is part of why people like parlay is part of what makes betting great is there is volatility to it. Customers like the volatility because even though it makes them lose sometimes, they actually can go on some win streaks, too. And so that's part of what makes it fun. If it were just the same thing every time and you were just winning or losing at exactly the same rate, I think it would be less interesting from a customer perspective. And so we're trying to balance all those things and obviously do care that some shareholders are asking questions about volatility. But I find that once -- there's almost like kind of -- at least when we were getting asked a lot about this before, there was this underlying, I think, question that wasn't really being asked that I think was behind it, which is like, okay, do I really believe you it was bad outcomes. And I think that was as much the question is like, hey, are you cool with the volatility? I think if everybody is like, yes, I feel really good that the structural hold is right and there's going to be some volatility, then sure. But when you have a couple of bad quarters and just trust me, it was the outcomes, I think naturally, you're going to get some questions. So we did a lot of work to try to assemble factual data on like the rate which favorites one and other things to show. And even though people then saw that, they still don't really know, okay, but like should you still have done better or worse? I don't know. I'm just taking your word for it that this was outcomes. I think last quarter, being able to come back and say we had a material positive outperformance because of good outcomes, I think that's something that actually quieted a lot of these volatility questions because it's one thing if you're coming back in 2 quarters in a row and saying, yes, we got killed and then like, all right, well, volatility should swing both ways. So I think as time goes on and investors see it's normalizing to where we thought it was going to be over any real period of time and the volatility swings are both directions, and it seems like it's normal, I think people won't care as much about it.

Again, the natural gravitation to parlays theoretically should increase volatility to some degree, but not if they're independent outcomes. But one question we've had is that concentration risk around influencer-based events, right? And I think this is something that -- or the prepack parlay, right? It's -- you get in and ultimately, people also want to bet the same -- socially kind of want to bet the same thing that their friends bet. And then you get these almost a herd mentality. So is that accurate? Do you expect over time as the product suite evolves, that people do more different things? Or do you see some of this herd mentality? And again, as the big event, we can get into the Jack -- the Powerball, right? But Powerball is like as it gets big, it then gets really -- gets a lot bigger because everybody sees the huge opportunity and then kind of, well, I want to be part of the crowd. So how do you just see it evolving from a behavioral perspective? I mean you can have some influence here, too, I would assume, in terms of what you -- kind of what you show the customers.

Well, I mean, I think, one, this is part of what I was saying on the risk mitigation is we can't and don't allow for amounts of concentration that create risk for us. We have controls and thresholds in place that if we get to a certain level, we'll protect against that. But I think for the most part -- and there are very rare exceptions where that does come into play. But for the most part, it just never happens because at this point, and this is, I think, a competitive advantage we have over most of the market. At this point, there is such a variety of things and there is such diversified betting going on in the app that even if you do have a promo or an influencer or whatever doing that, that drives volume to one, it's still not that concentrated relative to how much is going on throughout the entirety of the app. So we really haven't had an issue in recent memory where that came into play. But we do have controls in place just to make sure we did, for example, a few years ago, have a promo that went viral that was for a really long shot parlay and it had virtually no shot of hitting. I mean it was like knock of the century, it wasn't going to hit, but we still shut it off because we're like, look, we -- at that point, we were up to about $100-something million of risk or $200 million of risk, and we were just like we're not going beyond that. But that was the last time I can remember that happening. And I think the thresholds are a little higher now, too. The other thing I would say in terms of the volatility in investors and all that, that I think is going to help is we've changed a little bit how we're doing our guidance, too, in light of this. I think we kind of, in some ways, exacerbated maybe the perception of how much volatility there really is by focusing on a guidance midpoint and moving it any time there was a dollar shift one way or the other for outcomes, which this time around, we said, look, even though we had positive outcomes, it's still within the range, and we're not going to change the range just to -- and we're trying to focus people away from a midpoint and more towards a range. So I kind of think we shot ourselves of the foot a little bit like guiding to a precise midpoint essentially, which no business can do. I mean every business provides ranges for a reason, right? So I think that will help also put us in more of a light of like, oh, yes, this is just normal volatility like you'd see in any business because we're not going to frequently be adjusting our guidance anymore if we're doing it more that way.

So I want to switch gears a little bit. Of the big topics that have come up in our conversations, I don't know if there's been a bigger one in the last 3 months than prediction markets. So I kind of want to lead here a little bit. I know since even last public comment, a heck of a lot has evolved here. So FanDuel has announced a partnership with CME. Underdog announced sort of a unique relationship with Crypto.com. So -- and Polymarket, as of yesterday, from what we can tell, is officially back in the U.S. I don't have a product slide, but they are able to operate here more independently than they were prior to that. So a lot going on in this vertical. We -- you had a very interesting sort of open to the idea of always need to be around the hoop, but maybe don't want to or need to be first mover in some way, shape or form, I may have paraphrased that poorly. But unpack that for us a little bit. And just as you started to see some of the strategic moves around you, where is DraftKings at right now?

Well, I mean, I think the most important question is where is the space going to evolve? And then I'll come back to kind of where we're going. But right now, you're right, there's a lot of cards being turned over in real time. And so we're getting a lot of information on these things like -- and it continues to appear that at least at the federal level, there is going to be -- this is going to be here to stay. And so then the question becomes, okay, like what happens with the states. And obviously, for someone like a Polymarket, they don't have to worry about that, and they also don't have any revenue risk in the states. But I think for somebody like us, we would view both the revenue opportunity as well as the pace where we could not have as much risk is largely focusing on places that did not have online sports betting for anything that resembles sports. And I think for other types of things like non-sports predictions, which was, I think, what Flutter announced they were doing at least initially. I think that's probably a little more widespread, but also depends, and we have to sort of see how that goes, too. We've always kind of been -- and this applies not just to this, but I think it's especially important for this of the mindset that why make an announcement before you have to on anything. And so I don't think we feel like we have to announce what our plans are until if time came, we were ready to actually act on those plans or pretty close to ready to act on those plans. And I think that's just the best way of kind of seeing what happens without being committal,, but also doesn't mean internally we can't be preparing and doing things and slowing down at all there. So I don't know if it's as much I'd say like we don't think that we should be an early mover or anything like that as much as it is. We have to make sure we have the right thing and we are prepared to do it. And when that happens, if that happens, we'll announce something, but we don't feel like we need to announce anything ahead of that, which is a different strategy than I think others take, but there's pros and cons to both, I think. But that's just for whatever for better or worse, how we've always kind of operated in these types of situations.

I mean, you made an interesting point around just like -- where I would go -- like if I uplevel this a little bit, I would just start with, is this a TAM that you can't ignore, right? So when you think about size of what's happening because there is a debate in the industry of, oh, we look at the U.K. and we say that this is -- pick your number, high single digit of mix, and it's a product that's always existed. But then you see a product evolution, and there were a heck of a lot of people that did similar U.K. analyses to the U.S. and were like, oh, sports betting is probably not going to be that as big of a deal or as disruptive either. Kind of how do you kind of come out on -- is this a -- people talk about right to win in this vertical and then people also talk about, oh, that's probably not as big of a deal as you think. Where are you out on it? And I need to pay attention here.

I mean I think it's pretty simple. I think the TAM opportunity is likely to be very significant in states that do not have legal online sports betting. And I think it is likely to be fairly small in states that do have online sports betting. And I give you two reasons. One, evidence-based U.K. example, that is a market where both products coexist together, and it's a single-digit percentage of the total revenue, which I think is an indication that when both products are available, customers overwhelmingly prefer the traditional online sports betting product. The other thing I would say is just if you think about it conceptually, it's going to be very difficult to ever have as full featured an offering in a prediction market setup as you could in an online sportsbook. One of the chief reasons being risk management, when you are putting as a market maker, a market up on an exchange, you just have to be comfortable with anyone taking that liquidity. Anyone can fill that order versus we are able to place limits on Sharps and other people. And that is the only reason we're able to offer the variety of bets and things that we can. If we offered all the different bets that we offered, and we weren't able to do that, we would get picked off and destroyed, right? So I think that is going to be a limiting factor for not just the variety of bets, but parlays and other sorts of combinability and things like that. For example, if you are putting out -- we were just talking about this earlier, actually, a very high leg parlay, you have to collateralize the potential winnings on that every time you do that. So it's pretty hard even if you know you're going to win the vast majority of these parlays and you have enough variety out there that you're not going to lose on -- you're not going to -- even if you somehow knew that and you could manage the risk, which is also an if because you don't know who's taking your bet on the -- or who's taking the order on the other side. you also have to then collateralize all that, too. So I just think it's virtually impossible under that type of regulatory framework to ever have something that could be as rich and varied as what you see in an online sportsbook. But no doubt, there will be in a state that doesn't have an online sportsbook available, at least the legal one, there will be people that just want to go and do game lines and other simpler stuff that's available on it. So that's kind of like the way I look at it. I think where you have both, it's clearly, in my mind, going to be challenging for that product to compete. But I think where you don't have anything else, it's pretty good.

One thing that also evolved differently here in the U.K. was the kind of verticality of the tech stack. Here, it became just absolutely required that you own your own tech. You guys were super early and a lot of foresight around SBTech, which I think has turned into a huge amount of your platform. You've brought in further technology, Simplebet and other things that you've built around that beyond that. But it's become, again, day rigor, and that's what you do in American online sports betting. Do you think the same requirement is going to be there for prediction? Because right now, one of the things that has allowed for faster move to market for the public announcements that have occurred has been owning -- has been some sort of partnership or B2B framework. It has not been owning a vertical tech stack. So do you think that's going to be a necessary part of this landscape? Or is it -- is the exchange-based model different?

Hard to know because it's really not even a question of do you want to own your tech, but what parts of the tech are absolutely essential to own and what's not. And I think it's too early to know exactly what the answer to that will be. But I think as a general strategy, controlling more is always in almost any case I can think of in a technology product better. But it doesn't mean everything. Even today, we don't have -- we use plenty of partners for different parts of the tech stack. We use Amazon Web Services. I mean there are a lot of pieces. So I don't think I have enough understanding of this ecosystem to know that yet. And I think what we did in sports betting, it could end up being a template where we actually did start off partnering on the B2B side with Kambi, and it took us a couple of years before we really realized -- or maybe a year before we realized like we got to own this thing, and that's when we did the SBTech deal and went public and all that. And I think part of that was the notion of owning or having our own tech initially wasn't even realistic at that point. And I think if we did, it would have had to been an acquisition of a level we weren't comfortable making. So like those are all factors, I think, that you have to consider. But generally speaking, I think owning more of the tech is usually better in almost any consumer product than not.

And maybe kind of last on this area, but just be zooming out, like as a CEO, there seems to be some nexus of financial technology that's entering this discussion here, whether you look at -- I mean, the primary name that comes up increasingly in our discussions is Robinhood as sort of an onboard or obviously on the brand or broker side of this. But I kind of want to ask this from DraftKings' perspective, is there a world where you could be a meaningful platform in sort of a non-sports context? Or how do you -- how important is the sports overlay to DraftKings as we start to think about some world where convergence of financial products and again, what we'd consider today and truly a sports offering start to come together?

It's a very interesting question you're asking. And I think that if things continue the way that they are trending today, there will be some convergence there. And it's actually really for one simple reason, which is if we do end up going into something like predictions or sports predictions, we're going to have to build out or acquire or partner to get basically the -- build out the regulatory framework to operate the same types of products that you talk like financial instruments of a variety, right? So once you kind of have that, it was sort of similar to when we were just doing fantasy sports, it was a huge dig to think of getting into online in any way when it was smaller in the U.S. because it was just a lot of effort. And at that time, I think it was just New Jersey doing iGaming. And -- but once sports came, we're like we're going to make the investment and we built out everything we needed to do sports betting, yes, we might as well do iGaming. We already have a lot of that, and we know that there's a market opportunity there. And so I kind of think of that as similar. Now hopefully, that becomes successful for us. But like if I were approaching, for example, the other way and saying like will we just jump into financial products absent of anything like this, I'd say, I don't know if we have a right to win and if that's the place we want to focus now. But if we're going to build out the other stuff, then you might as well at least explore extending into there. And I think that's sort of what you're seeing is that the Robinhoods, the financial products are starting to dip into sports. And I think potentially, you could see the other side happen too, once this all gets built out, if it all gets built out.

So let's pivot to product. And you mentioned iGaming. So let's just stay on that for a minute. A lot -- a number of companies have sort of changed their strategy here. I'm thinking FanDuel has been pushing in this area. Caesars has done this and pushing kind of casino first, right? Where do you kind of sit or where do you think DraftKings sits today on pushing on that relative to kind of more of an OSB cross-sell lens? Because I think, again, that's an area where you were super early and successful with a product like Rocket, but it led you towards a customer that was looking for a certain experience, whereas when we think casino first, we think different demo, possibly different even branding and certainly a slot first experience, which requires maybe a third-party set of product around that as well.

I think -- so first of all, I think you kind of summarized it. We were early and I think did a very good job focusing on the cross-sell side from sports with table games and games like Rocket. And I think that we didn't maybe appreciate and go as hard after as we are now, the opportunity for a slots first customer. So I think it's a big opportunity for us. If you look at our share of the table games market. It's orders of magnitude higher than our share of the slots market right now online. So there's clearly a lot of low-hanging fruit there. I think there's some element that's product, but most of it, I think, is the marketing and who you're going after, like you were saying. I don't think it's like our products always can be improved, but we have all -- we have the product. We have the games, we have the setup. I don't think historically, we've been actually -- maybe to your point on like -- I don't want to say branding, but like the way that it looks and like what games you're featuring, I don't think we've been orienting that way. And I certainly don't think we have been marketing that way. So that's something that we are adjusting now. And hopefully, we'll have some real opportunity there.

Can you care to put a timeframe around sort of like where this sits on the whiteboard? I mean we are also -- which was my next set of questions, going to NFL launch. So it's not like your engineers and product people aren't super busy. Where does iGaming kind of fall on the priority list right now? And kind of what timeframe should investors be thinking about?

I don't necessarily want to put like a timeframe for anything like share or anything like that because there's so much that goes into that. But I think now we are making dramatic improvements and also focusing more on the right thing, which is that slots first casino-first customer. And I don't think that was true a few months ago. So when that will start to really show up in numbers that you guys look at, I don't know, hopefully not too long. But I know that we are making that shift now, and so I expect results to follow.