Good morning, everyone, and thank you for joining us for DraftKings 2022 Virtual Investor Day. Today, we'll talk about a number of topics, including our latest TAM outlook, position, investment, and product, unit economics, and long-term EBITDA. I'll begin the presentation, then we'll hear from our CFO, Jason Park. We encourage you to submit your questions during today's event for the Q&A session. I'd also like to remind you that this presentation is subject to the legal disclaimers at the beginning of the deck, including those related to forward-looking statements and non-GAAP financial measures. Before we begin, I want to acknowledge a tragic situation unfolding in Ukraine. Our hearts, thoughts, and prayers are with all Ukrainian residents, as well as those living abroad with friends and family in Ukraine.
We have made contributions to humanitarian organizations working in Ukraine, and we will continue to look for ways to provide support. With that, let's begin the presentation. We want to talk through six broad topics today. First, our updated view on TAM, where we are increasing our outlook from $67 billion for OSB and iGaming to $80 billion, which is comprised of a $4 billion increase in OSB, an $8 billion increase in iGaming, and a $1 billion increase in Canada. These increases are due to impressive GGR per adult trends across live OSB and iGaming states. Second, our share also continues to be strong and in line with our long-term estimates for both OSB and iGaming. The industry structure is healthy and evolving the way that we thought, with a handful of national operators and several other smaller regional players.
We are live in states representing 36% of the U.S. population, more than any other operator, and our major competitive differentiators are shining through. We would also like to touch on some nuances with respect to state reporting and educate you on different ways to look at an operator share across states. Third, our strong share is driven by deep investment in our products, which we have continued to enhance rapidly through the past 12 months. While we won't be sharing much information on future product plans for competitive reasons, we are very excited by our roadmap for 2022. We continue to be encouraged by the underlying retention rates and cohort payback periods that drive our unit economics. We generated $68 million of positive contribution profit in New Jersey in 2021, only its third full year of operation.
Five of our states, including New Jersey, were contribution profit positive in FY 2021, and we anticipate five more states to become contribution profit positive in FY 2022. For purposes of clarity, contribution profit is defined as gross profit minus advertising spend or external marketing and does not include any fixed costs. Jason Park will have more detail on this topic and on unit economics later today. Our multi-year plan supports a path to 30%+ EBITDA margins, and based on our TAM and share outlook, we believe there's a path to $2.1 billion of EBITDA once we reach 65% and 30% legalization for U.S. OSB and iGaming respectively, and 64% legalization of the Canadian population.
Lastly, we'd like to touch on our new verticals, specifically our NFT Marketplace, which represents upside to our long-term EBITDA outlook of $2.1 billion. Starting with TAM, we estimate that the U.S. opportunity for OSB under the assumption of 100% legalization is at least $26 billion. If we apply New Jersey's 2021 actual GGR per adult of $107 per year to the 265 million adults in the U.S., the implied U.S. opportunity is already $28 billion. If we then apply a GDP per adult adjustment relative to that of the U.S. to control for state-level differences in discretionary income, we arrive at a $26 billion TAM estimate. For context, a $26 billion TAM estimate would imply GGR per adult of $98 per year.
Also worth noting, this estimate does not include any assumed future growth for OSB in New Jersey after 2021. Having said that, we do think there is upside to this estimate. If we do the same exact analysis, but apply a 10% CAGR on GGR per adult from 2021 to 2023, the U.S. OSB TAM is implied to be $34 billion using the population method and $31 billion using the GDP method. Investors frequently ask if New Jersey is a good state from which to extrapolate. From this chart, you can see why we believe that New Jersey is a reasonable point of extrapolation, especially when adjusting for GDP per adult.
When looking at New Jersey's first NFL season and comparing it to our other state's first NFL season, New Jersey gross revenue per adult is actually below that of the other 10 states and would be even lower if adjusted for GDP per adult. Using the same U.S. TAM methodology for iGaming as we did for OSB, we arrive at a $52 billion TAM using the population method and a $48 billion TAM using the GDP method. For context, this would imply GGR per adult of $197 per year and $180 per year for each of these two methods, respectively. $197, of course, was the actual GGR per adult in New Jersey in 2021. As with OSB, this TAM sizing does not include any future growth.
It just takes the current size of New Jersey today as an extrapolation point. If we do the same analysis, but apply a 10% CAGR on GGR per adult per year from 2021 to 2023, the U.S. iGaming TAM is implied to be $63 billion and $58 billion using the population method and GDP method, respectively. As with OSB, we are sometimes asked if New Jersey is a good state from which to extrapolate the iGaming TAM. From this chart, you can see that New Jersey is also a reasonable point of extrapolation for iGaming. When looking at New Jersey's sixth full year of iGaming, which was 2019, and comparing it to other states' 2021 iGaming numbers, New Jersey's gross revenue per capita is actually below that of the other three states without even adjusting for GDP.
For context, 2021 was either the first or second full year of iGaming for all of the other states in this comparison. Updating on legalization trends, we now sit at 55% of the country having legalized sports betting, with 44% legalizing online sports betting, compared to 41% and 27% a year ago, which means that 17% of the U.S. population had sports betting legalized in the last 12 months. Of the 44% of the population that has legalized online sports betting, 38% have live OSB, and we believe the remaining 6% will have live OSB in due time. DraftKings is now live in 36% of the U.S. population for OSB versus 25% a year ago. iGaming is now legal in seven states, representing 13% of the U.S. population if you include poker-only states.
DraftKings is live in five of these states, representing 11% of the U.S. population. The last 12 months have seen continued strong momentum with OSB laws passed in states, as mentioned before, representing 17% of the U.S. population. Additionally, two states representing 4% of the U.S. population, Michigan and Connecticut, launched iGaming in 2021. The legislative momentum we are seeing in Canada is also encouraging, and it looks like Ontario will launch both OSB and iGaming soon. We believe that the Canadian opportunity could be up to $9 billion in gross revenue extrapolating off of New Jersey using the population method and $6 billion of gross revenue when using the GDP method. As with our U.S. estimates, this does not include any future growth.
If we apply a 10% CAGR from 2021 to 2023, TAM estimates for Canada increase to $11 billion in gross revenue and $8 billion in gross revenue using the population and GDP methods, respectively. Next, I would like to talk about OSB and iGaming share. It has certainly been an eventful year in terms of launches and new operators entering the space. Having said that, our share continues to be very strong and in most states has been increasing despite competition. We want to invest a bit of time here in helping our shareholders understand state reports. It is important to remember that these reports are tax reports put out by various states. There are reporting nuances across each state report which makes these reports an imperfect measure of operator-level share and makes comparisons difficult.
The four main factors you need to consider when analyzing a state share report are, number one, state reports have a wide range of transparency. Nearly half of the states we operate OSB in do not provide operator-specific metrics, and only four of our 17 states report net gaming revenue by operator. Number two, almost all states report GGR on a cash-in, cash-out basis, which can impact monthly results because of unsettled futures bets. Number three, promotions are accounted for differently on a state-by-state basis, with certain types of promotions obfuscating operator-level GGR. Finally, total opportunity is defined differently by different sources, which I'll explain in a couple of slides. This hodgepodge of visibility makes it difficult for our investors to know our share.
Since all states report out at least handle and GGR for the state, we're able to calculate our handle and GGR share with much more precision internally, albeit still with some of the limitations mentioned. We will share these numbers in a few slides. Also, on page 50 of the appendix, we provided an overview on how each individual state that has OSB currently reports OSB data, which should be a helpful resource. We also want to shed light on the analytical difficulty which presents when calculating state-reported GGR share. As we have discussed on prior earnings calls, the promotional mix an operator uses can impact state-reported GGR and therefore obfuscate an operator's true share. On this page, we present a simple example of an illustrative state that has only two operators. Each operator has identical handle share of 50% and identical net revenue.
However, operator one only uses free bets for promotions, while operator two only uses profit and odds boost for promotions. Despite having the same handle share and net revenue, operator one's GGR share is 65%, while operator two's is only 35%. This is, of course, only a theoretical example, but it does illustrate some of the limitations of relying on state-provided data in determining share. We certainly understand that investors will continue to use state reports as a tool in understanding industry structure despite these limitations. We are providing these explanations as additional context so that those who are analyzing state reports can better judge what conclusions to draw and with what level of confidence. As mentioned earlier, industry share is defined differently at different times by different sources.
This slide highlights a few of the primary examples of how we have seen different sources defining the denominator for share. The first example is the simplest, total U.S. share for all states with online sports betting. The second example shown calculates share across all states, excluding Nevada. Unless otherwise noted, this is the way DraftKings typically calculates our share. The final example calculates share in only states in which an operator is live. In the interest of transparency, we have shown our OSB handle and GGR share and our iGaming GGR share for all three methods. For method one, our OSB handle share is 28%, our GGR share is 24%, and our iGaming share for GGR is 19%. Using method two, the numbers are 32%, 25%, and 19%.
Using method three, the numbers round to the same levels of 32%, 25%, and 19%. Note that these iGaming GGR shares include poker revenues and do not include contribution from GNOG share. GNOG iGaming shares for Q4 were 4%, 4%, and 6% using methodologies one, two, and three respectively. Now that we have discussed share, we want to shift to OSB industry structure. Using state gaming reports and third-party estimates from Eilers & Krejcik, the three leading scale operators typically account for about 80% of GGR. In fact, on a population-weighted average basis, the top three operators accounted for 84% of U.S. OSB GGR in Q4. As you can see, there are typically three players who have the marketing, brand, and product capabilities needed to consistently access and drive share across U.S. states and jurisdictions.
As was the case at our last Investor Day, DraftKings is currently operating in more states than any other competitor, which is the result of the collaborative effort across several very strong functions of the company. These include government affairs, business development, compliance, product, and technology. Additionally, as we have discussed in the past, being live in more states will provide significant marketing and thus CAC advantages as scaled operators benefit from cheaper national advertising spend. We are just starting to see the effects of that now beginning to take place. Over the past few years, we have continued to strengthen our nine keys to sustainable differentiation. As shown in the following pages, we will continue to build on these points of differentiation through product and technology as well as marketing and brand.
Broadly speaking, our OSB product roadmap for 2022 can be divided into four main buckets, in other words, what we call the four Cs. These cut across our products. Number one, control. Number two, content. Number three, connectivity. Number four, customer experience. Control is in reference to controlling our own product destiny through vertical integration. A great example of why control is so important is our 99.99% site uptime since migration. This is especially important around key events like the Super Bowl. Content is all about having differentiated offerings. In the following slides, I will share some of the iGaming and OSB content that we rolled out in 2021. Connectivity is about creating an integrated ecosystem for skin in the game fans.
A great tangible example of our product connectivity is having a single wallet across all of our products and across all states. This makes for a more seamless product experience for our customers and has proven to be especially important as more states legalize OSB across the U.S. Lastly, customer experience ties to our obsession with the customer. High-quality products that are easy to use, top-notch customer service, and a relentless focus on always getting better drive our team day in and day out. We are always looking for ways to delight our customers, remove friction, and make each customer's experience personalized and unique. iGaming is a great example of how we've been able to differentiate from our competition when we own our own technology platform.
Our in-house DK iGaming Studio has been able to create 51 total games to date, and collectively, these games have generated about 60% of our handle in 2021. DK Rocket is a great example of a game we have launched that is truly differentiated from our competitors. This is exciting because the DK iGaming Studio not only creates a better customer experience for our players but also drives better gross margins for the business by reducing costs associated with usage of third-party games. Now that we are fully migrated onto our in-house technology for OSB, our main focus over the past six months has been content expansion, where we are very pleased with the progress we have been making.
Some of the highlights from the back half of 2021 include the launch of Same Game Parlays and in-game micro markets, as well as significant expansion of our player props. We also have a very exciting roadmap for 2022, a year in which we believe we will increase our differentiation more significantly than ever before, as it will be the first year that we've had a full year entirely operating on our in-house sports betting platform. One of the most exciting initiatives for 2022 is continuing to build out our OSB product social functionality, including many exciting new features such as integration of user-generated video. Finally, I'd now like to take a minute to bring it all together before moving on to the unit economics section.
Regarding our share assumptions, we are not changing our long-term estimate of 20%-30% for OSB, but we are updating our iGaming estimate to 20%-25% due to the expected completion of the Golden Nugget Online Gaming acquisition. In Canada, where there is already an existing gray market of expected future licensees, we are assuming a lower share in the 10%-20% range. We coupled these assumptions with previously discussed TAM assumptions, plus an assumed legalization rate of 65% for U.S. OSB and a 30% legalization rate for U.S. iGaming, and a 64% legalization rate for both U.S. OSB and iGaming in Canada.
Putting this all together, we estimate a potential range of $6.3 billion-$8.7 billion in U.S. gross revenue and $400 million-$800 million in Canadian gross revenue, for a total of $6.7 billion-$9.5 billion in gross revenue across North American OSB and iGaming.
Thank you, Jason, for that update on TAM, share, and industry structure, as well as on product. This next chapter is all about DraftKings' path to profitability. Our path to enterprise profitability is very simple, moving left to right across the top here. First, we invest to acquire cohorts of customers within a state such that the gross profit generated from those cohorts over two to three years exceeds the acquisition cost. We refer to this as gross profit payback periods. Second, each unit or state turns contribution profit positive two to three years after launch as those customer cohorts stack on top of each other and our marketing within that state decreases to reflect the number of remaining customers that are available to acquire at good paybacks.
We refer to this as states turning contribution profit positive. Third, as DraftKings has more states and, to be precise, more population that are in their second or third year versus their first or second year, then the cumulative contribution profit across our states turns positive and sets us on a great path to enterprise profitability. We refer to this as cumulative contribution profit. Starting with gross profit payback periods, today we will share several things. A, a very clear equation to align on the definition of this metric and where promotions hit the equation. B, empirical evidence of our strong customer and revenue retention. C, a description of why we think we have the best retention in the industry. D, an explanation of how gross margin rate naturally improves for a customer cohort over time, due primarily to promotional intensity for new versus existing customers.
E, a walkthrough of how you can calculate for yourself our gross profit payback periods for each of our LTM periods based on our disclosures for MUPs, external marketing investment, ARPMUP, gross margin rate, customer retention, and revenue retention. In terms of state contribution profit, we will share an update on New Jersey, which beat both top and bottom line versus what we communicated a year ago, as well as the great news that five of our earliest states turned contribution profit positive in fiscal year 2021, and we expect five more to turn positive in 2022. You can now effectively model our business on an annual cohort methodology to accurately estimate our future revenue, gross profit, and contribution profit.
In terms of cumulative contribution profit, we will share how our cumulative contribution profit was still negative in 2021 due to us being only roughly three years since the overturn of PASPA and the population-weighted average of our states as of year-end 2021 being just 1.5 years, but that in 2022, the cumulative contribution profit flips positive and will very likely stay positive from this point onwards. Once the population-weighted average age of our states is greater than three years, you would expect us to have meaningful positive contribution profit. At the end of this chapter, we will also share an important recent observation that is changing how we think about state launches and should impact how you model our business. Put simply, we are acquiring more customers faster in recent states relative to what we saw in New Jersey.
The net effect of this is that we are acquiring customers more efficiently on a per-customer basis, and the total marketing investment earlier in a state is deeper today than versus the New Jersey prototype we introduced at our 2020 Analyst Day. We are effectively pooling forward marketing to invest more for a bigger business. Let's invest a few minutes into aligning on the definition of gross profit payback. Put simply, gross profit payback is the number of years before the gross profit generated by a customer cohort is greater than the investment to acquire that cohort. The upfront acquisition investment for an LTM cohort is easy to know, as we can calculate that factually for any given LTM period.
It is simply the amount of external marketing we invested in a period, which has historically been in the high 70% to low 80% of total S&M expense. This year, we expect it to be in the low 70s% as we continue to invest in our fixed sales and marketing and mix into more efficient national media spend. The future gross profit generated by that customer cohort is a function of our strong customer retention and revenue retention, as well as the gross margin rate. Gross margin rate is impacted heavily by promotional intensity. Typically, we see gross margin for year one of roughly 35%, moving to 45% in year two, and then roughly 50% in year three. We will get into why this gross margin expansion occurs later in the deck.
As a reminder, we count promotional activity in the gross margin portion of this equation. Lastly, of course, player LTVs are much higher than just three years of gross profit, but we manage customer acquisition for the business on a three-year basis. On this next page, you can see that our customer and revenue retention is stellar. This page shows that after we acquire, say, 100 customers in a quarter, that 83 of them are still with us over their first full year. 88% of the 83 are still with us in their second year, and 96% of the 73 are still with us after their third year. This trend of improving customer retention rates is fairly intuitive.
If a customer has decided to continue to engage with this product category and has chosen DraftKings as their app of choice, after two years, they are very likely to stay. The right-hand side is even more important and shows that the revenue of the retained players more than makes up for the customer attrition. If those initial customers we acquired generated $100 in their first full year, the entire cohort of players, including those who churned, generated $122 in the second year and $143 in the third year, meaning that the revenue per retained player grew 39% and 49% in year two and year three, respectively. For the avoidance of doubt, the $122 and $143 is only from the retained players. This trend is also fairly intuitive.
You are seeing the effect of decreased promotional intensity impacting net revenue as well as general enjoyment driving frequency. We often get the question of how retention differs in OSB-only states compared to OSB and iGaming states. This page is the same analysis as the previous page, but for OSB-only cohorts. Admittedly, a smaller sample size, but we wanted to share the data that is available. As you can see on this page, both customer retention and revenue retention are remarkably similar across the two pages. In fact, both metrics are slightly better for OSB-only cohorts. Retention is very important to our gross profit paybacks. Investing in acquisition only makes sense if you can keep your customers. We think we are uniquely strong at retention due to a variety of factors, most notably the quality of our products. It starts with acquiring the right customers.
Our marketing team uses data science to target customers that fit our desired profile instead of just acquiring customers for the sake of acquiring customers. Once the customer is on the DraftKings B2C platform, we aim to provide the best experience on their end-to-end customer journey. This includes great customer experience for any questions they have as they are getting familiar with our products, as well as a myriad of factors that are consistent with our internal value of putting customers first. We are also always looking for opportunities to cross-sell our acquired customers, both across products, for example, from DFS to OSB, from OSB to iGaming, from Marketplace to OSB, and across sports, for example, from PGA to NASCAR, from NFL to college basketball. This drives frequency in play and player engagement. Last, but most importantly, we are always innovating on our products.
We believe product will always win in the long run, and DraftKings is dedicated to listening to our customers and anticipating their desires and building what they want. Gross margin rate of the retained revenue is very important as well. Promotional activity impacts gross margin rate, and promotional reinvestment is highest when a player is new to the platform and naturally declines for a customer cohort over time as other retention factors drive their decision around where to concentrate their play. Much of our cost of goods sold is tied to gross revenue, including promotions. Therefore, as promotional rates decrease, gross margin rate increases. Additionally, we are achieving gross margin rate improvement from operational initiatives, such as getting volume discounts from certain platform providers and building more in-house games.
Finally, in this section, around gross profit payback periods, we want to help our investors see what we know to be true about our relatively short payback periods. By using our disclosures on MUPs, external marketing, ARPMUP, gross margin rate, customer retention, and revenue retention, you can validate that each of our last four LTM periods had payback periods of two to three years. For our last four LTM periods, you can model our payback period for the customers acquired in that period. By taking our disclosed MUPs for that LTM period and comparing to the same period from the prior year and applying a customer retention rate, you can get a year-over-year change in MUPs, the numerator here. Dividing by a frequency assumption, how often during the year an average player engages, you can arrive at a number of new actives acquired in that period.
By taking our sales and marketing expense for that period and applying a percentage to account for the relevant variable acquisition spend and dividing by the actives acquired, you can get our CPA or CAC for that LTM cohort. Note that over the last four LTM periods, our variable acquisition marketing spend was 77%-81% of our total marketing spend, and on average, was 79%. It has historically been in the high 70% to low 80% of total S&M expense. This year and going forward, we expect it to be in the low 70s% as we continue to invest in our fixed S&M and mix into more efficient national media spend on an adjusted EBITDA basis.
Finally, by taking our ARPMUP and applying the same frequency used in step number one and applying our reported gross margin rate for the LTM period, you can arrive at an approximate annual gross margin generation for that cohort. From that, you can compare the CAC for that cohort to the gross profit generation for that cohort and arrive at an approximate payback period. It's worth noting that we do not disclose the frequency of our players, but this variable cancels itself out when calculating gross profit payback periods. Whatever frequency you use for MUPs needs to be the same frequency you use for ARPMUP. This page applies the equations from the prior page to our last four LTM periods. As you can see, our payback periods for an acquired cohort are consistently between the two to three year payback periods we target.
The great thing about this framework is that it allows you to consistently monitor the efficacy of our marketing spend by estimating our payback periods for an LTM cohort. Right now, we are investing at payback periods that are greater than three years for the New York golden cohorts. This spend is with the hope that New York will rationalize its tax rate going forward. Given this, you can expect the math outlined on this page to slightly exceed three years in Q1 and Q2 of this year. If it becomes clear that the tax rate in New York will not change, we will adjust our acquisition marketing spend in the state accordingly. For a more detailed view of this analysis, refer to page 46 in the appendix.
Now that we have the building blocks for customer cohort gross profit paybacks, it should be fairly simple to understand how we expect states or units to turn contribution profit positive after two to three years. As we stack those customer cohorts on top of each other, and very importantly, as total marketing expense declines to match the number of available customers that can be acquired at the two to three-year gross profit payback threshold, states turn profitable. On this page, you can see that in fiscal year 2021, we had five states that were contribution profit positive, and we expect five more to turn positive in fiscal year 2022, for a total of 10 states representing 18.7% of the population, while the remaining 17.4% of the population will still be in the investment phase in 2022.
On a population-weighted average, our states were 1.5 years old at year-end. When our average exceeds three years, we expect significant contribution profit. Here's an update on New Jersey, our oldest state. New Jersey generated $68 million of contribution profit in 2021. Net revenue grew $86 million, while contribution profit grew $60 million, beating our top and bottom line estimates from last year, and contribution profit rate improved from 5% to 28%. In 2022, we expect to see continued meaningful top-line growth as we still have young cohorts, and we expect contribution profit to grow at a faster rate than revenue as we continue to realize operating leverage in the state. More specifically, this operating leverage comes from gross margin improvement and from marketing spend decreasing as a percentage of net revenue and on an absolute basis.
While you have New Jersey state data from the prior page, and we still believe that it is useful for explanatory purposes, it is important to note that we have learned a lot since the New Jersey launch, and newer states are behaving differently. As you can see from this page, the pace at which we are acquiring new customers has massively increased from August 2018, when we launched New Jersey, to today. This page compares New Jersey OSB to Arizona OSB customer acquisition on a population-adjusted basis. We are acquiring more customers faster than we had originally thought approximately two years ago. The effect of this is that we are acquiring customers more efficiently on a per-customer basis, and the total marketing expense earlier in the state is deeper.
The great thing is that as we look at quality and early retention indicators, these cohorts are just as high-performing as others and are very much on track to pay back in two to three years. By way of comparison, over our first two quarters in Arizona, we acquired approximately 3.5% of the state's adult population. In New Jersey's first two quarters, we only acquired about 1.3% of the state's adult population. This explains why our sales and marketing investment in fiscal year 2021 may have looked big. However, the spend was more than justified given that more customers joined the DraftKings platform faster than ever before, and the foundational customer and state economics were excellent. On the left-hand side of this page, you can see that we had negative cumulative contribution profit in 2021.
Yes, we had states and products that had positive contribution profit, but those were outweighed by newer states where we were still investing. On the right-hand side, you can see that we expect our cumulative contribution profit to be positive in 2022, which is an amazing milestone. As I mentioned on our earnings call, we expect to generate positive contribution profit for the entire 2022 year across all of the states where we are live today. I really hope this was a helpful discussion on our customer and state economics and gives you further confidence in our path to profitability. Now I will turn it back to Jason Robins to discuss our long-term enterprise economics.
Specifically, he will discuss our cost structure at scale, including our long-term gross margin, external marketing spend, and SG&A, and why we have conviction in a greater than 30% EBITDA margin at maturity.
Thank you, Jason. This next chapter is about our long-term EBITDA, which we define as five years after the final state launches, representing 65% OSB legalization and 30% iGaming legalization. It is worth noting that at that point, we expect that our top line will still be growing at an attractive rate given ongoing positive revenue retention, and there may still be additional state launches in the future that provide more upside to the TAM and the top line. This slide updates our expected gross margin rate for OSB and iGaming. You can see that we now estimate an at-maturity gross margin rate of 56% versus the 58% we presented last year. The driver of this reduction is the current tax regime in New York, as the methodology we use includes all legal OSB states with multiple operators weighted by population.
This was partially offset by a 1% improvement in each of two other categories, payment processing and revenue share. These reflect the same actuals using the same method. Please note that there is upside to this gross margin rate if other large states legalize OSB and/or iGaming at a lower tax rate than the average shown here, and there is downside if large states set tax rates higher. While we will continue to focus on incremental improvements in other categories of COGS, state tax rates have the largest impact on long-term gross margins. That said, there are other levers we can pull outside of COGS in any given state to drive a long-term EBITDA margin that is consistent with what we have shown previously.
We understand that it may be counterintuitive that our external marketing spend at maturity will be lower than it is today, but it is important to understand that marketing spend does not scale with revenue or active customers but is instead a function of customer acquisition. At maturity, we will be acquiring fewer customers than we are today, which is something that we have already experienced when we only offered a DFS product. Over 90% of the decrease from current external marketing expense to at-maturity external marketing expense we believe will be due simply to an assumption that customer acquisition at maturity will have slowed from the current early phase of the industry. The remainder of the decrease is attributable to optimizing CAPEX. Most of this CAPEX optimization comes from a mix into more efficient national advertising from local advertising.
Additionally, as the industry matures and customer acquisition growth slows, we believe new customer LTVs will be lower than early cohorts, which will naturally lead to cuts being made in our highest CapEx channels while leaving more efficient channels in place. DraftKings is constantly testing and optimizing across the business, and with more data and more time, our analytics team is better able to optimize marketing spend. That said, we have assumed very little benefit from this, given that over 90% of the planned decrease between now and at maturity is expected to come simply from less customer acquisition. Our fixed costs or SG&A have grown at a roughly 40% CAGR since 2019. This was part of a multi-year plan to rapidly scale certain functions to create competitive advantages in key areas such as product, engineering, data science, analytics, and marketing.
After FY 2022, we expect the growth rate of SG&A to begin to ramp down. We will of course continue to invest in our product as well as in new initiatives, but we believe the core OSB and iGaming businesses will not require the rapid addition of personnel to expand into newly launched states and provinces. Lastly, it is important to note that our long-term SG&A structure is very much in line with global at-scale online gaming players. We are currently planning to spend a mid to high-teens percentage of revenue on fixed costs over the long term. Bringing it all together, we are increasing our long-term adjusted EBITDA guidance to $2.1 billion from $1.7 billion.
The increase is due to the conviction we have as a result of our revised TAM estimates and our continued strong position in both OSB and iGaming. As a reminder, these long-term economics are estimated for five years after the U.S. reaches 65% OSB legalization and 30% iGaming legalization across the population and 64% legalization across the Canadian population. Additionally, this estimate for long-term EBITDA does not include any revenue or costs associated with projected GNOG synergies or any upside from any other new initiatives such as Marketplace and media. Achievement of any projected GNOG synergies is subject to the closing of the previously announced acquisition of GNOG. One example where the company may have additional upside is our NFT Marketplace, which launched in the back half of 2021. NFTs are a rapidly growing category that crosses over nicely with DraftKings' core demographic.
As you can see from the stats on this page, we are very encouraged by early results, validating our belief that non-gaming products can eventually become a meaningful component of our CAC to LTV flywheel. While it is very early in the development of the NFT industry, the rapid growth cannot be ignored and potentially points to a very large TAM. It is also important to note that both the users and sales numbers on the bottom half of this page underestimate the current industry size, given that CryptoSlam, the source for this data, only tracks owner to owner secondary sales and does not include initial sales from the content creators. Some industry research suggests that the total NFT industry in 2021 was north of $40 billion in GMV when accounting for all sales. Thank you for your time and interest in our presentation.
We will now take a short break and return for Q&A in about 15 minutes. Welcome back, everyone. We're ready to begin the Q&A session. Please, remember to continue to submit questions throughout. We'll try to get to as many of them as we can. Joe, we ready for the first question?
First question is, can you comment on the status of collecting signatures for the California ballot initiative in 2022?
As a reminder, we have a California ballot initiative that we formed a consortium with seven total operators and if we're able to gather the required number of signatures, we'll be on the ballot this fall and hopefully it'll pass. Right now, we're doing great against our target. We continue to exceed our goal. For eight straight weeks, we've been ahead of our weekly goal. You know, feeling like we're going to get there and very excited about the prospect of hopefully being on the ballot this fall.
Could you update us on your DFS business? How large is it today? Is it still growing? How much of that is customer driven versus spending?
Well, we haven't broken out specific DFS volumes, but in the past, we've stated it's in the hundreds of millions of dollars of revenue. It's a very high margin business, and it is still growing. The growth is coming from a combination of spend and new customers being added.
How do you expect Canadian betting habits to differ? Do you account for that in your projections for the Canadian opportunity?
Well, Canada is a very similar country in many ways to U.S. Obviously, there's differences as well. Being our largest trade partner and very close in proximity, very similar culturally, we expect it to be a similar type of market. Right now, we haven't accounted for any differences. You know, obviously, once we get some real data, we'll adjust up or down appropriately. Right now, we don't have any real data or evidence to suggest that we should be accounting for Canada differently than any of the U.S. states.
Looking at your TAM analysis, with New Jersey being the reference point, can you comment on expectations for growth in New Jersey OSB in 2022, especially considering the launch of New York?
Do you wanna take that one, Jason?
Sure. You know, we think underlying that question is a question around cannibalization between New York, New Jersey. I'd mentioned that we are one of the few operators that are also live in Pennsylvania and Connecticut. We stood up multiple teams early on to measure cannibalization with different methodologies. The punchline is, so far, the cannibalization has been quite minimal. We think New Jersey will continue to be a healthy grower and we'll continue to track the cannibalization.
How does profitability across iGaming versus OSB compare long term in your view? Are you seeing anything within states now that inform the view of profitability long term for iGaming versus OSB?
Well, from a gross margin perspective, I should say, iGaming and OSB are remarkably similar. Very similar cost structure on the COGS side. You know, I think in terms of profitability, we don't really break out marketing spend in that way because, you know, we look at ourselves as a B2C platform. We acquire customers on using whatever the most efficient means is. Sometimes that's iGaming, sometimes that's sports betting. And then we try to cross-sell across all the products that we have once we have a customer engaged with the platform. We really look at it from a gross margin perspective, it's quite similar. That said, as we noted in our presentation, we believe if there were full legalization or equivalent legalization, the TAM for iGaming is bigger, and therefore it would be a higher profit product overall.
You know, right now, we're not projecting iGaming to have the same level of legalization as OSB in the short to midterm.
The gamified NFT collection with the NFL seems like a really significant opportunity. Can you talk through how this will work? Is it traditional fantasy football where you can only play with players you own?
It's very similar to that. You know, right now we are diligently working on having a product ready to launch in the next few months, so I don't wanna upset my product team by giving away too much. The question, you know, is it similar to fantasy sports? It is. It does require ownership of NFTs in order to play certain players. Should be a game that's relatively familiar. There'll be some differences, but should be relatively familiar with the added twist of you know, collect NFTs and, you know, those NFTs can be used to play in the games and you can win prizes using them.
What do you think are the biggest drivers of faster customer acquisition in new states? Is there a halo effect of legalized gaming becoming more mainstream, or is it marketing tactics?
You know, I think this is a somewhat speculative answer, and, you know, I wanna kinda caveat that we don't really know all the underlying drivers. Certainly, there's some optimization we've been able to do on the marketing that's allowed us to go to market in a more effective way. I think learned a lot since the first few states that, you know, has increased our level of performance at a higher level of scale. I do think there is a lot of momentum. As more states legalize, there's more press coverage, more people have friends in various states that are betting, and I think that's really helped get early adoption up.
Then lastly, I think that, I you know, ironically, 'cause we always get asked about this as a negative thing, I think the influx of spend from competitors has actually helped boost awareness in the overall market, and we feel that we have a very strong conversion funnel. We know how to target and to convert effectively with our advertising. I think the halo of increased awareness from all the competitive advertising and the surrounding coverage has actually helped ramp markets faster.
Can you provide an update on your plans to accept crypto as a form of payment?
Right now, we don't have a definitive update. It's certainly something that we're exploring and looking at. I understand it's something that many of our customers want, and also with our NFT Marketplace, it's something that's very relevant there as well. It's definitely something we're looking at, and when we have more to share there, we'll do so.
Can you talk about your EBITDA burn? Do you have enough cash to get to cash flow positive?
Yeah. You know, one, if there's anything I want people to take away from today's presentation, it's three things. One is, the playbook is working. Two is that if we execute that playbook, we are able to get to profitability. Third is that we have more than sufficient capital on the balance sheet in order to execute that playbook. The short answer is yes, and we have a multi-year plan we're executing against that should allow us, with cushion, to be able to get profitable without changing our playbook at all.
Can you remind us what is included in contribution profit, and how do you delineate external versus non-external marketing expense?
Well, external marketing expense is essentially advertising. You know, external marketing expense includes things like the people that manage the campaigns, the creative folks who design the creative tools and other sorts of, you know, services that we use in order to conduct analytics and optimize the marketing, things like that. That's the simple way to think about it.
How do you handicap the competitive environment in Canada for OSB?
Well, obviously, Canada has had a gray market for many, many years. Ontario has indicated that their intent is to license those who've been operating in the gray. I expect it to be a pretty competitive market from the get-go because there's so many that are already in there, and those are the very same people that we're competing with in the U.S., with the difference being in the U.S. when a state launches, everybody's kinda starting from the same place. You know, that's the reason that we projected lower market share. We projected a 50% of the market share on the low end and about 2/3 of the market share on the high end that we are getting in the U.S.
The reason for that is that while we feel we will have a very competitive product, we think that we do have a really strong brand, a database due to our DFS presence in Ontario for many, many years. The fact of the matter is that Sportsbook is not a new product. iGaming is not a new product in the market, and the very same operators have been doing it, will be able to continue to do it as far as we see. You know, we're being a little bit more cautious about what kind of market share we can project taking. Hopefully, there's some upside there, but we won't know till we get in.
When do you think New York will be profitable? Can you offer any color on that state?
Well, what we've said pretty consistently is that we are targeting a two to three-year path to profitability in every new state. Some may be closer to two, some may be closer to three. Hard to say 'cause each state has its own dynamics, but we believe New York will fall in that same two to three-year time horizon.
Looking at your retention rates and your investments in customer acquisition, are you able to determine how many apps your customers use, say, after a year?
Well, if we're talking about competitor apps, you know, we certainly have survey data talking about that. The indications that we see are that during the trial phase, especially with a lot of the, you know, very generous offers that some of our competitors have put out there's certainly been a lot of trial. What we found is that customers tend to stick with where they get the best experience, where they feel the best product is. You know, I think our CRM and our data science that allows us to personalize and retain at very high levels is effective as well, and it shows up in the numbers. As we showed in the presentation, we have very high year one retention, very high year two retention, and once you get past two years, it's almost 100%.
We've seen very strong retention numbers, and I think even though it's survey-based, it's not, you know, quite the same as if it's our own hard transactional data. We do have indications to suggest that while there is trial in the early phases, people tend to focus and stick with one app. You know, that seems to be consistent despite any state differences, despite any differences in competitive environment. That sorta effect seems to be very consistent.
Can you talk about the impact to your 2022 guidance if MLB cancels the season?
Well, MLB is an important sport for us. You know, from a customer acquisition standpoint, it's a decent sport, not one of the biggest, but given so many games, it is an important monetization sport. That said, what we've seen in the past is that when there's certain sports that, you know, we saw this with some of the schedule shifting around, you know, during the early days of COVID, following it, you know, people who want to bet tend to find things to bet on. Are there people that would maybe prefer to bet on MLB? Of course. As long as there's other things, I think generally speaking, most customers find things to bet on. That said, there will be an effect.
Will it be the same as saying, let's just deduct whatever money we would have made off of MLB? No, because as I said, there will be some transfer into other sports, into other products. There will be some effect if there's a cancellation. We don't have any reason to believe there will be a cancellation. Right now, there's only been a couple series that per team canceled. Now, if that's all that happens, we don't expect to have any material impact. We think that the guidance we've issued is unchanged. We'll just have to wait and see. Our hope certainly is that there is not a long-term cancellation, and I think it would be a very unfortunate thing for baseball and for the players if the whole season got canceled.
I'm confident that they both know that, and they'll make sure that that doesn't happen.
Can you discuss your ramp in product and technology investment in 2022? What are the key buckets of where that investment is going?
Well, the bulk of our investment is going towards our OSB and iGaming products. That's where our largest growth and revenue sources are. While we're very bullish on some of our other products that are new, like media and Marketplace, we're managing those in a way where the investments are not going to get ahead of the revenues. You know, we're running those at or close to breakeven, and we'll see how things go and if more investment's merited. Certainly, we are putting resources against them, though. As noted, we expect meaningful revenue from Marketplace and from media this year, so we are putting some investment there. The very large bulk of it is going to OSB and iGaming. We're still investing in our DFS product.
Now, that product is also generating a lot more gross profit than by far the marketing expense and fixed expense we're spending on it. We are still investing there. We've added two different features and different game types, and that's part of why that product has continued to be growthful.
Can you share any thoughts on Pennsylvania still being contribution profit negative, and what may turn that around? Is it just the tax regime?
Well, there's a couple things there. I think it's less about the tax regime and more about the timing. We didn't launch Pennsylvania OSB until after the start of the 2019 NFL season. You know, we missed the most important customer acquisition phase. Obviously, Super Bowl was good, but you know, that beginning of the season customer acquisition phase. We also didn't launch iGaming in Pennsylvania till middle of the following year, till middle of 2020. You know, in many ways, if you look at sort of the average start date by, if you average two products together, it's really a 2020 launch. We feel like Pennsylvania is on track to be in that two to three-year timeframe.
Some states, as mentioned earlier, will be closer to three years, some will be closer to two years, some might even be less. Pennsylvania, like every other state that we're in, we believe is going to be in that two to three-year timeframe.
If you are acquiring more customers faster, can the two to three-year payback period you target be shorter?
It's a great question. It's certainly possible. That's not something we're ready to say just yet. I think that, you know, logic would imply that if you're ramping faster up front and there are more new customers coming on, that that's certainly something we might see.
Two-part question on TAM. Should we assume it will take at least a few years longer to reach the iGaming TAM projection versus that of OSB? Can you update us on what your views are regarding the iGaming legalization pathway?
Well, it's really hard to say. I mean, state, you know, governments, any government, you know, I wish I had a crystal ball, but I can't predict. Certainly, what we've expected and said from the beginning is that OSB would lead the way and iGaming would be a second wave, so to speak. We continue to believe that's the case. I don't know if it's a few years or if it's one or two years, but I do think that iGaming will be a little bit behind sports betting. You're seeing that play out already. Some states certainly have chosen to do it all together, like Connecticut and Pennsylvania.
There's been, you know, in the case of West Virginia, a state that did it one year later, and we know there's other states that have sports betting that are now considering iGaming. Much of it depends on the economic state of the country and of different states, too. You know, a lot of states that are flush with cash might not be in as big a rush, whereas others that really need the tax revenue might accelerate. There's a lot of different dynamics that make it hard to pinpoint an exact year. I think the general comment or the general question of will you see a little bit of a lag in iGaming versus sports betting, yes, that's something we've always said and continue to believe to be the case.
As far as the state of iGaming legislation, there's been some activity, not as much as sports betting, you know, which as I noted, is to be expected given sort of the sequence that we see things happening in. I do expect that you'll see an acceleration there over the next few years. I think there's a lot of states that are considering it now. Sometimes it takes a couple years or even three years to pass a bill in a state. Our one of the government affairs consultants that we work with likes to say, "It's like the Hall of Fame. You have some first ballot Hall of Famers, some get in the second and third." I think it's very similar with legislation, and that's what you're seeing play out with iGaming.
With respect to the higher penetration of the Arizona population, do you believe that the overall population penetration is increasing due to customer awareness and/or better product?
You know, I think that could be the case. It's really hard at this stage to say, is it just a faster ramp or, you know, is it something that's unique about a state, or is it just a macro effect that, you know, this, as the question noted, of just, you know, more awareness, better product driving deeper penetration. It is certainly notable that on a per population basis, Arizona is ahead of where New Jersey was, or where New Jersey is, even many quarters into New Jersey and only a few into Arizona. Hard to say. Is that something about Arizona, or is that something about, you know, the state of the market, or is that just a ramp type of thing? We'll find out, though, in the next couple of years.
How do you think about offsetting inflation in your marketing and advertising expense and/or your other fixed costs in SG&A?
Well, inflation, I think, where it sort of, you know, seems to be, at least from the economic analysis I've seen, people seem to be, pinpointing inflation, it's not really an advertising cost. It's more in things like oil and other sorts of goods. That doesn't really affect us. We're a digital company. You know, people are still mostly working remotely, so they don't have to drive to work, as much anymore. You know, I think for us, the real big impact, if any, of inflation would be cost of capital, but fortunately, we're very well capitalized. We ended last year with over $2 billion on the balance sheet. We feel like, you know, we don't really have to worry as much about a cost of capital thing right now.
Given that we have enough capital on our balance sheet with cushion to get to profitability, we may never have to worry about it. We can wait out this environment as long as we need to, and I don't really think inflation will have any impact on our business. From a consumer standpoint, the gaming industry has certainly shown over many, many years that any economic conditions it's quite resilient to, whether that's a recession, whether that's inflation, anything. Typically, that does not have an effect on the type of business that we're in.
Are there substantial benefits of partnering with traditional media companies to drive more efficient top of the funnel marketing support for new users?
Well, certainly, I mean, we already do that. We have great partnerships with many media companies, and those have been very helpful in driving a lot of our early success. So, absolutely. I think media companies have the eyeballs. They have the live sports rights in many cases, and that's something that, you know, I think is very complementary to us, not only because it's an opportunity to collaborate to drive customers to us, but also all these media companies are very aware that the more people that engage with fantasy sports and sports betting, the more their appetite grows for the content that the media companies are putting out on sports, whether that be the games themselves or other sorts of supportive programming around the games.
Is your TAM estimate based on NGR or GGR? What is the assumption for long-term promotional percentage?
It's based on GGR. As we've talked about in the past, we've assumed that we have roughly a little over 20% promotion percentage of promotions at maturity that's consistent with other mature markets, and obviously very much within our ability to control. So that's how we did the analysis. We based it on GGR. That's the most accurate data that we have available, and then we used an assumption of low 20s% promotion rate at maturity.
I would add that as we talked about the TAM for OSB and iGaming, we did not apply any growth on the 2021 numbers, so that should give comfort around any concerns around promotion.
Yeah. I think that's a great point, Jason. We were quite conservative that, you know, we use New Jersey as a benchmark, which as we demonstrated in the presentation, is a very reasonable benchmark. We didn't assume any future growth. I know as Jason Park just noted, we've gotten questions, "Hey, do you think we've accelerated, you know, the market a little bit because of the rate of promotion?" Perhaps that's the case. I'd be very surprised even with that, very surprised if New Jersey doesn't continue to grow. You know, we wanted to be cautious here. We wanted to hopefully leave some room for upside. We didn't include any assumption of future growth.
We just said, "What is New Jersey doing today, and what would that look like if we projected it to the rest of the U.S. on a GDP-adjusted basis?
In your last Investor Day, you used New Jersey, Australia, and the U.K. as goalposts for your U.S. OSB opportunity outlook. Do you feel that New Jersey has reached the level of maturity where it is a better reflection of the entire U.S. market?
I think so. I mean, you can look at a lot of different things and triangulate, and they sort of all get to the same answer. But we felt that, you know, especially given we got the question earlier on, is Canada different? You know, the most reasonable and acceptable way would be to use a U.S. state, and as long as that U.S. state seems representative, which as we showed in our presentation, it certainly is maybe even a little bit conservative of a choice, then that should be a great benchmark. You know, as far as where New Jersey is on the curve, as noted a moment ago, I personally believe New Jersey will continue to grow.
If you look at the iGaming market as an example, we're, what, seven, eight years in now, and it's continuing to grow at a very healthy pace. It almost seems unfathomable that, the numbers that we've reached in OSB are the peak. You know, we felt like it was mature enough, and it was, you know, reasonable enough as a state to compare other states to that using today's New Jersey numbers would provide, a very reasonable and achievable TAM estimate. Whether there's upside or there, you know, I believe there probably is, but we'll have to see.
Two-part question on SG&A. When you say mid to high teens SG&A as a percentage of revenue, what are you including in fixed marketing? Second part, why is it growing so much this year?
You know, if you look at the end of our presentation, the slide on EBITDA, we break out external marketing separately. You know, as per kind of that earlier question, that's the advertising. Everything else, you know, people that run the marketing programs, any third-party tools we use are all included in fixed cost, and that all is included in SG&A. Also, as noted in the presentation, the multi-year plan that we're executing against entailed us hiring up fairly quickly in certain functions such as product technology, marketing, analytics, data science, all of which we felt were very important in order to outcompete the others in the industry. We now feel like we have a team that is well set up to do so.
While we're certainly gonna continue to grow our head count, it's gonna be at a much more moderate pace given that we feel we have reached a point of scale by 2022 that, you know, allows us to be not just competitive, but to outcompete the others in our industry.
Customer retention is roughly unchanged since the update on your last Investor Day, but revenue retention has improved. Can you talk about any drivers?
Well, I think product is the biggest driver. We've added so much to the product since last Investor Day, in particular, since we migrated in the late summer of this past year of 2021. We've added a ton of new features to OSB. We've released many new games on the iGaming side. Some of those have outperformed any of the existing games that we have previously released. I think as we continue to improve our product, we'll have better and better monetization. I also think it'll help on the retention front. You'll start to see that show up as well.
One of the reasons I think we've retained so well despite, you know, having more competition in the market this past year than the year before, is I do think we have a superior product to the vast majority of others out there on both the iGaming and sports betting side.
Why was Connecticut able to be contribution profit positive so quickly versus a state like Michigan, which also has iGaming?
Well, Connecticut only has two iGaming operators and three sports betting operators, so naturally we're going to have a larger share in that market. I think that's the primary reason why it was able to get profitable faster.
Looking at NFTs, is this a margin or EBITDA drag in 2022? How should we think about the long-term profitability or margin of your Marketplace product?
You know, as we noted, our new growth initiatives, Marketplace being included, are not, you know, areas that we think will have any material negative or positive impact on EBITDA in 2022. Obviously, we have high hopes for these products, and we hope that they will contribute positively in 2023 forward. What we've said so far is that for 2022, we do not expect them to have any material impact, positive or negative on the bottom line.
Do you see a consensus forming among states about the quote, unquote, "right tax rate to maximize state revenues?
Not really. We have a saying here, every state's a snowflake. They're all a little different. The dynamics are different, the way decisions are made are different. The issues that, you know, people are looking to utilize tax revenue to help support are different, and the politics are different. We haven't really seen that happen. It's been pretty different by each state and, you know, that's consistent, I think, with what we've been seeing since day one.
I know you talked about promotional rate a couple of minutes ago, but what gives you confidence that promotional allowances will be reduced to 22%? Based on your data in early states, will this achievement be later than you originally projected?
Well, there's two answers to why we're confident. One is we've benchmarked against a number of other markets around the world. Now, one of the nice things about our industry, even though always, you know, gotta be cautious, sometimes past is not always the best predictor of the future. But unlike a lot of other industries that are new, where in those cases the products are new, the product market fit is not as clear. This is a product that's been around for many, many years, hundreds of years, and has been on the digital side present in many markets in Europe for many years, as well as other parts of the globe.
There's a lot of places that you can benchmark, and they all tend to, you know, generally settle around that low to mid-20s%, and some are in the teens. You know, we kind of took where we felt like the average was, and I think that's one way of doing it. The second way of looking at it is we looked at just the pure effect of, as mentioned in previous presentations and I think today, new customer promotions are much richer than existing customer promotions. The way that you attract people to sign up for a platform is very different than the way that you retain those people. The attraction tends to be promotional, the retention tends to be because of product and customer experience.
If you just look at where we are today, the amount of promotions or the percentage, I should say, of revenue spent back on promotions for new versus repeat, and you sort of fast-forward to, you know, a future state like that we presented in the end of the deck, where a percentage of revenue coming from new is much, much less than it is today. A lot of that is what's going to bring us down to the low 20s% on its own.
Looking at New York, how long do you plan to invest in customer acquisition in the state?
Well, we'll always continue to invest in customer acquisition, but I think the question is what are our CAC targets and what are those based on? Right now, there's certainly been some chatter about a potential of seeing a lower tax rate or promotional deductions being allowed or both in the state. They're in the midst of the beginning of their session right now. I think we'll get clarity on whether that's likely to happen in this session over the next few months. You know, because such a large percentage of our New York spend going forward this year will be in NFL season, we have time to see that play out. The plan is to adjust accordingly based on what the ultimate tax rate is.
If it stays the same, we're obviously gonna have to have lower CAC targets than if it goes down, simply because the gross profit we'll generate off of the players we acquire will not be as high as if it were at a lower tax rate. Just like every state, we look at the tax rates, we look at the LTV projections of the players, we look at what different products are in the state, and all the other factors you would consider in modeling out the value of a player and the expected gross profit of a player, and we target two to three-year paybacks, and New York will be no different.
How should we think about the SG&A ramp between 2022 and 2023?
As noted earlier, we expect there to be a meaningful slowdown between 2022 and 2023. A lot of what you're seeing in 2022 is because of hiring that was done in 2021. We've already moderated the pace of that a bit in 2022. The expectation is that the exit rate in 2022 will not be at as much growth year-over-year as the exit rate in 2021 versus 2020 was. I think you'll continue to see that trend. You know, our strategy was to staff up quickly to try to make sure we could out-compete everyone in the industry on product, technology, data science, analytics. We feel we have a really strong team now, and while we'll continue to add to it, we can absolutely slow down the ramp and still achieve better than expected results when we first started.
Is there an opportunity to add live streaming within the app? How do you think about the benefits of potentially doing that?
You know, it's an interesting question. It's a very theoretical one because we haven't been able to test it yet, and we're very much a test and learn company. You know, we always have a hard time answering questions like that without having any real data. I do think that you know, whether it's in the app or elsewhere, having a low latency stream available that will allow people to watch things that don't have delays will absolutely make the live betting experience better. Right now, one of the complaints we get, whether it be fantasy stats updating too quickly or live betting, is that the video feeds that people are watching have delays and are behind where our betting feeds on the data side are. I think having some sort of low latency feed would be helpful.
Whether that needs to be in the app or not, I don't know. I mean, one of the things that you've recently seen in iOS over the last few releases is, you know, now when you have a window open with video, you can put that as a thumbnail in the corner, you can easily expand it, and you don't have to actually. You know, it used to be when you closed an app or when you switched over to another app, I should say, the video went away. Now it's in the corner and it can be expanded, it can be watched as a thumbnail alongside the betting. It almost doesn't matter whether it's directly coming from our app or from another app. I think the ability for people, if they want, to watch a stream while they're betting is there today.
The real thing is how do we get those streams to be low latency, you know, regardless of where they live?
Do you have any plans to add an online poker product for the states that have legal online poker?
You know, again, that's something we've looked at. We're not right now announcing any plans around poker and, you know, if we do have something to say there, we'll update. Just like every other product that's out there that, you know, we're always looking to get better or add new products around, poker is one that we've talked about and considered and, you know, if we have something to share, we'll do so.
Do you have confidence that you can renegotiate with existing skin providers? If so, how soon can that happen?
Well, we have a lot of great relationships, and, you know, certainly I think some of those early on, you know, we thought that, or maybe our partners thought that we weren't gonna do as well as we have. Since then, we've become a real premier partner of choice. You know, we'll see what happens. I don't really wanna comment on the future state of any contracts or any sort of relationship with partners. I will say, if you look at our last year's Investor Day deck, the percentage of revenue share, you know, in our gross profit build-up, gross margin build-up went down by about 100 basis points. I think that's an indication of where things are going or at least where things were going from, you know, last year to this year.
Really don't wanna comment on future.
This is a question on hypothetical M&A. Are there revenue synergies when you look at OSB M&A opportunities? Is there a rationale that cumulative market share would go higher?
Well, if you look at what materials we put out when we announced our Golden Nugget Online Gaming acquisition, we did point to that. We do think that and obviously, it really depends, right? I mean, it depends on what the acquisition is. Generally speaking, you find that there's best of both world that comes with the products, and by plugging everyone's products together, revenues go up, and also I think just knowledge sharing and other sorts of things can help there. That's been pretty consistent if you look at Europe, and there's been a lot of consolidations happen in Europe. That's been a very consistent outcome. You know, we'll see what happens with Golden Nugget, but certainly that's our expectation based on the analysis that we've done.
You know, I would say generally speaking, that would be true of any other sorts of, you know, consolidation that might happen.
Clearly, you are comfortable to spend within your two to three-year payback periods. With operators messaging lower competitive intensity, I would imagine the cost to acquire customers will come down. If this happens, would you be more likely to target shorter payback periods or allocate more dollars to sales and marketing?
You know, it's a very interesting question because, as I noted earlier, one hypothesis would suggest that competitive spend actually lifted awareness and boosted response and allowed us to acquire more at scale more efficiently. I don't know for sure if that's what's going on. So just like everything, we're gonna test, and we're gonna see. We'll let the data do the talking. That's the nice thing about the way we do things. We don't speculate. We test a lot, and then we analyze and based on what the data is saying, we make decisions. As far as will we actively reduce, you know, I think right now we're very comfortable with the playbook. As I mentioned earlier, one of the things that I want everybody to take away is the playbook's working, and we're well-capitalized to execute against it.
I think as long as we continue to believe that's the case, and I don't see why that would change, we're gonna continue to stick with our playbook. When competitive spend went up, we stayed disciplined. We didn't extend our payback periods. We did, you know, exactly what we've been doing since the day we went public. You know, I think people, investors hopefully appreciate consistency. As long as the data and analytics are saying this is working, we believe being consistent is the way to go.
Next two questions are on product. First, how are you thinking about integrating more social aspects into the product?
Well, as we noted in the presentation, one of the very exciting differentiators that we've been working on are social features in our OSB product. We have a lot of cool stuff we've done already, but we're still very much at the beginning days. Things like video integration for user-generated content are on the horizon, and there's a lot of other cool stuff that I won't talk about that we're working on there. I think sports betting is inherently a social activity. People like to discuss what they're betting on with their friends. They like to debate it. They like to share wins and, you know, lament together over losses.
I think inherently it lends itself very well to a social experience, and our objective is to build as much of that, you know, or to facilitate as much of that in, by what we build into the platform. We have a very talented team working on this. They're solely focused on building out the social features. We expect to see a lot more there in the coming years.
The second product question is more regarding your vision. We often hear about the product of the future as being an integrated platform of streaming video, integrated data, e-commerce, et cetera. What still needs to be figured out to get to that vision?
Well, I think we'll have to see how things play out, to really understand what the hurdles are right now. If you talk about video integration, a lot of the hurdles are existing contracts and, legacy sorta, you know, elements of those contracts that prevent this type of thing from happening right now, at least in a way that really works for the consumer. You know, as far as e-commerce and other things go, those are things that certainly can be integrated. It's just a matter of when the right time to focus on them is. You know, in a way, what we're doing with Marketplace is a form of that, so I think that's a good example.
You know, in the end, the way we look at it is the market will evolve in a way that has a lot to do with factors that we don't even know today. We wanna be very, you know, high level in our vision, and very consistent about that high-level vision, but also be nimble and adaptive. I think what makes the best technology companies are those that can see trends and go after them. A great example of that is the NFT Marketplace. Two years ago, I didn't even know what an NFT was. You know, I think that the fact that it's grown so much in the last year. It was less about us figuring that out beforehand and more about us identifying a trend that we did analysis on and saw it was very much overlapped demographically and with our audience.
We were able to quickly pivot and build that product. It wasn't even on our roadmap at the beginning of last year, and in a matter of months, we were able to build it and launch it. I think that speaks very much to the culture of the company being very nimble, innovative, always looking around corners and also able to pivot and adjust and reprioritize in order to go after something as soon as that opportunity is identified and vetted.
Regarding GNOG's 4% share, which you referenced in the deck, do you plan to alter the product to more DKNG in-house games? And do you believe you can maintain that share level? Lastly, how long will it take to transition to the DKNG platform?
Well, our hope is we can actually increase it. You know, if we're successful at generating revenue synergies, then we will. It all comes down to that. As far as what we'll do, certainly there's a lot of platform costs. That's a big part of the synergy that can be taken out by integrating DraftKings PAM, integrating our OSB product, integrating our iGaming games. As far as, you know, whether we'll be pushing more of the DKNG games or whether we'll be relying more on other games we create or whether it'll be third-party games, just like everything, we're going to test. One of the hypotheses that, you know, we had around the GNOG acquisition, is based on it reaching a different segment of the iGaming audience, the casino-first audience.
If you look at their demographics, it's very different than DraftKings, where, you know, about 80%-90% male. They're 50/50 male/female. We have an average age. It's almost a decade younger. Our table games are more of a mix than our slots. They're about 75% slots to table games. Definitely, you know, that hypothesis is rooted in real data and analysis. We have to understand that if that's the case, then we might have to create a little bit of a different experience. It could be, you know, small things like the way that games are ordered and presented to the customer or larger things around branding and, you know, what things we push in market.
We're gonna test just like we always do, and we're gonna learn and, I think that'll allow us to really optimize the best of both worlds.
Can you elaborate on your fixed versus variable marketing? What's included in each, specifically national media and team and league deals?
Well, I think we already answered that. You know, the variable marketing includes any external advertising. You know, whether that's national, whether that's local, it doesn't matter. If it's, you know, external, if it's advertising, if it's marketing intended to acquire players, it's included. Things that are in the fixed costs are things like headcount for, you know, people running the campaigns, the cost of producing, you know, creative third-party tools that we use in order to optimize. Those are all part of fixed costs.
Looking at TAM, why did New Jersey OSB grow so much in 2021, and why not assume growth in the future?
Well, the reason, as I noted earlier, we didn't assume growth in the future is that we felt like, you know, presenting something that really would be above question was really important. We didn't wanna get into a debate around how much growth to assume or anything like that. Our strategy is to put out there what the numbers are today. You know, what we know for certainty is where we are right now. It's anyone's guess what the future will hold. We'll let investors make up their own minds about what they think the true growth rate of New Jersey going forward is, and if you extrapolate that to the remainder of the U.S., what that means for the TAM.
Turning to M&A, are you still looking at international acquisition targets? What are you focused on from an M&A perspective?
Well, you know, we have a number of conversations. We're always interested in exploring things. I think right now, however, if you look at where we were maybe a year ago to today, we're definitely much more focused on our organic growth path. I think that while certainly we'll always be exploratory and there's things that could be a strategic fit and could financially make sense on the M&A front, we're very happy with our current setup. We feel like we have everything we need from a product perspective and technology perspective. We feel like with the Golden Nugget Online Gaming acquisition, which hopefully will close soon, we're gonna have a great dual brand strategy to go out and reach all segments of the online casino audience. We feel like we have the full plate right now.
Does that mean that there won't be something interesting on the M&A front? No, it doesn't mean that, but our focus certainly has shifted more towards organic growth.
What do you attribute your recent iGaming success to?
Product. We made huge strides in our product over the past year, launched a ton of new games. The, you know, DraftKings Rocket, as we mentioned, was our most successful new game to date. It was also very innovative. I think it was not just another sort of slot or blackjack game. It was something totally different than what you see in most iGaming apps. I think really continuing to have a great selection of differentiated games, having great customer experiences, you know, and then just data and testing that allows us to optimize all the other things we do from the operational and CRM side around it. Great data science. You know, all those are things we continue to build, and I think it just enhances the product, the customer experience, and the way that we operate.
Looking at your fixed costs growth in 2022, can you provide any color on what your priorities for investment are? Is it OSB, iGaming, something else?
You know, as we noted earlier, the bulk of the investment's going to OSB and iGaming. There certainly is some investment going into some of our new growth initiatives like media and Marketplace, but those are not things that we expect to be, you know, as significant as the investments we're making in OSB and iGaming. We also are still investing in our DFS product, but again, it's much smaller as a percentage of the total fixed cost than what we're investing in our OSB and iGaming.
Certain competitors have very large customer databases, and you have about 6 million daily fantasy users according to the deck. Do you feel that you're at a disadvantage when new states launch?
Well, I think different, you know, places present that number differently. We presented 6 million paid players, people who've actually made a deposit and played a game on DraftKings platform. If you look at our total database just of registered names, it's, you know, closer to around 20 million. I think if you talk about it in the sense of apples to apples, we're probably in a pretty similar spot to most others who have databases. I also think the fact that our database is made up entirely of people who participated online makes a big difference.
Not only are they comfortable and used to playing on mobile apps and on websites, we also have, you know, those players that are active visiting us online on a regular basis, so the ability to convert them to different products online is much more seamless.
Looking at media, are you taking a build or buy approach to content development?
Right now, we're taking more of a build and partner approach. Certainly, there's like I noted earlier on the M&A question, there's always things that we might look at, and you never know, but right now our going-in strategy is to build and partner, and we've done both of those successfully, I think. You know, both pieces are a very important part of our media strategy.
Is there risk to your gross margin percentage over time from tax rates?
Well, I think it could go either way. At least in the fantasy space, we've actually seen several reductions in taxes and fees. We haven't seen any states increase taxes. You know, I think right now we're very focused on trying to make the case to states that launch that there's a sensible place to set the tax rate that allows us to effectively compete with the illegal market. I think most states and most policymakers understand that if you set the tax rate too high, it makes it impossible for legal regulated operators to compete with the illegal market, which sort of defeats the purpose of doing it in the first place.
You know, based on what I'm seeing today, I think there's a broad understanding of, you know, that the tax rates need to be in a reasonable range, and I have no reason to believe they're going to increase over time. You know, hard to predict, as I noted earlier. It could go either way.
Turning back to New York, do you think the state legalizes iGaming in the near term or the long term?
You know, very similar to what I've said, it's very hard to sit here and predict what a state's going to do. There's so many factors that go into it. You know, I think it's encouraging to see that more states are considering iGaming than ever before. I think it would be a great win if New York were to do it, and we'll just have to see how that plays out.
How does DraftKings think about the cost of its loyalty program?
Very simple. You know, when we launched the loyalty program, there was a clear mandate to the team that we would take from existing budgets and more make this a systematizable way versus ad hoc promotions of getting value back to consumers. That's really how we thought about it. We said, "Listen, you know, out of everything that we have, all the bucket of sort of generosity that we're spending, how much of that do we wanna try to automate?
How can we increase that over time, by testing into it and increasing the benefits where it's truly impactful and not wasting money in places where it's not impactful?" Really, it's all part of the same pool, and the mandate to the team was to not increase the overall rate of promotion, but instead just to shift some of it from more ad hoc promotions to more automated promotions to the loyalty program.
How confident are you in your unchanged estimate for $1.1 billion of long-term fixed costs, considering your 2022 guidance implies you'll probably spend around $900 million this year?
Well, we feel very confident in that. As we noted, we have a multiyear plan, and we've been executing against that plan, and so far, we've been able to hit that plan. So, I have no reason to believe the team will not be able to continue executing against it. The plan that we've laid out calls for $1.1 billion in fixed costs when we reach that $6.7 billion in revenue, and I think we're gonna be able to do that very successfully.
Somewhat related question: Is it fair that some of your fixed cost growth is due to Marketplace and media?
It is, but it's not a very large amount. You are right that when we show the $1.1 billion, that specifically excludes Marketplace and media. Right now, that's not a huge part of our fixed cost spend, but it is something. You are right that that number is a little bit of apples and oranges, albeit very small differences.
Does your long-term EBITDA guidance from the deck today exclude synergies from GNOG or include them?
It excludes synergies. We noted that there's about $300 million of synergies we've said we expect from GNOG. Those are not included in the EBITDA number.
Last question today, Jason: Why are you raising your iGaming share goal to 20%-25% compared to what you had in your deck last March?
Well, that was really just taking our existing share at the end of last year and adding GNOG's existing share at the end of last year. As we mentioned, we do think there could be some revenue synergies that might lead to higher share combined. You know, we wanted to be cautious. We didn't wanna take credit for something that we hadn't achieved, and that they hadn't achieved. We thought the most sensible thing to do at this point of time was just to add their share to ours, which would put us a little bit higher than that midpoint we used at 23% versus 22.5%. I guess we like round numbers, so we went with 20%-25% instead of 20%-26%.
It was really just adding the two existing shares together and assuming no benefit from the combination.
That's our last question for today, Jason.
Well, thank you all for joining us for today's presentation. Hopefully, it was informative. Hopefully, some of the new information we shared, you know, was able to convey things that we've been talking about at a high level in more detail. If you have any questions, please feel free to follow up with our IR team, and we look forward to speaking with you again soon.