21 Investor Day. We're very excited to host you today and share some of the excitement we have for our business and get into details we typically can't cover in an earnings call or a 30-minute one-on-one at a conference or an NDR. We're a bit disappointed to not be able to do this in person, but definitely did not wanna delay this event. First, on logistics. We have seven speakers with roughly two hours of prepared remarks. A 30-minute Q&A session will come at the very end once all speakers have concluded their presentations. Submit your questions in the viewer throughout the day. I will moderate the Q&A and pose your questions to our speakers for you.
Last on logistics, the full deck will be posted after the event ends. I will spare you the reading of the safe harbor and non-GAAP sections, but encourage you to read each word when you have problems sleeping. Now to the fun stuff, lineup and material. Michael will kick us off with opening remarks and share his vision for the future. Katie, our COO, will then give you some CTV market dynamics that are worth repeating, even if you generally get the concept, because they're really powerful. Adam, our Chief Product Officer, will follow Katie and tell you how we build all this cool stuff, and Allen Dove will talk about our scale and dev eng and key priorities, as well as how we plan to combine platforms. We will shift to sales and go-to-market and customer perspectives.
Sean, our CTV Chief Revenue Officer, will cover how we win in CTV, followed by Joe, our DV+ Chief Revenue Officer, who will cover this for all areas not CTV. Lastly, and with no introduction needed, David, our CFO, will show you how our TAM has expanded and how bright our future is, and how our goals and how we plan to make shareholders very happy over time. On format, you'll notice we go from section to section. Many of our speakers will touch on what appear to be the same opportunities, CTV, DV+, and audience, and they will also differentiate how we win in both the reserve and open markets. This is for good reason. Each speaker will discuss how they, in the context of their role, and how their teams contribute in making each of these areas a success.
This stretches from strategy to product to dev eng to financial performance to sales. To identify which area is being presented, we've added tags in the top right corner of each relevant slide to help you properly keep oriented. Let's get started. Michael, you're up.
Thanks, Nick, and really appreciate your interest and everyone who's dialed in today's Magnite's call. I'm really excited to introduce you to the Magnite leadership team. You've heard plenty from David, Nick, and me over the years, so now it's time to hear from the expert that actually do all the hard work. Today we'll focus upon the job at hand, how we'll integrate and run Magnite, while also sharing some multi-year predictions on where the industry is headed and how Magnite is expertly positioned to take share and extend our leadership position. I'll briefly set up today's presentation and then turn the slides over to our first presenter, Katie Evans, our COO. Let's start by reviewing where we are today. Magnite is the clear leader in the independent SSP market.
We are the only omni-channel scaled SSP with a focus on CTV programmatic ad spend and CTV ad serving. Publishers and buyers have been asking for a company that looks like Magnite since the rise of Google and Comcast FreeWheel's dominance. They want choice, don't want to sacrifice capabilities or performance. We've built that full stack alternative for the first time in ad tech history. David will talk about financial metrics more deeply, important to acknowledge our current scale as we shift to exploring future opportunities. $500 million+ in 2022 revenue, of which over $170 million will be CTV on an annualized run rate, with margins of over 30%, and importantly, over 400 talented dev and engineers, which will help us win in a new software-first approach that the team will walk you through.
Let's look at today's programmatic marketplace and how it might change in the coming years, and how Magnite will extend its leadership position. The programmatic market today is predominantly DV+, which stands for display, video, and the plus is all others. Display will be mobile, desktop, and all non-CTV video. Both mobile and desktop covers the video, and the plus covers emerging programmatic media types like audio ads, particularly podcasts, and digital out-of-home. CTV is the fastest-growing and most valuable segment, and the main reason why we have acquired Telaria, SpotX, and SpringServe. We are the only independent SSP that services both market segments at scale. Over the next several years, we expect the market to literally flip, where the majority of programmatic spend will be CTV, and DV+, while still growing, will be the smaller portion of the market.
Our omni-channel CTV-focused approach sets us up nicely in this new world. Perhaps more importantly, this spend shift brings with an even more significant change in the way these dollars will be transacted. They will all be transacted programmatically, however, the vast majority will be reserve auction versus open auction. Think of reserve as today's PMP, PG, or upfront markets, where a publisher works directly with a known buyer, pre-negotiates a rate, pre-negotiates an audience segment, a forecasted delivery schedule. In this environment, a publisher works with one main SSP to host and execute the majority of their reserve auction business. A big contrast to open auction, which is predominantly header bidding, which by definition entails multiple SSPs competing in a unified header auction. A tougher market to own for sure. Today, the vast majority of reserve auctions are conducted by two companies, Google in DV+ and FreeWheel in CTV .
This is a very large market that is poised for disruption. Magnite is the only SSP with the capabilities to be that disruptor. The reserve market is dominated by the largest omni-channel clients, all Magnite clients, and all very motivated to find an independent alternative to their current market options. We'll extend Magnite's leadership and scale in the coming years by focusing on and investing in our five top initiatives. First initiative, aggressively grow CTV. Second, grow DV+ share with scale and efficiency. Third, win the premium reserve segment. Fourth, strategic role in audience and targeting. Fifth, deliver growth and profitability. We'll dive down deeper in each of these initiatives with the executives of those respective divisions throughout today. I think it's important to double-click on that fifth initiative, delivering growth and profitability. Magnite is a long-term growth and earning story.
We believe long-term growth in the 25%+ area, coupled with long-term margins in excess of 40% are possible given our market position, increasing TAM driven by CTV programmatic spend, and leverage of the business model. In five years out, Magnite could achieve the following: up to $20 billion in annual ad spend, up from several billions today. A CTV market share of over 30%, a DV+ share of over 20%, up from high single digits, and continued growth and investment in developing programmatic markets like audio and digital out of home. Again, I thank you for joining us today. I look forward to answering questions at the end of this presentation. Now I'll turn the program over to Katie Evans, our COO, who will provide more details into the evolving CTV marketplace. Katie?
Thanks, Michael. Good morning all. Over the next 10 or so minutes, I'm gonna be walking through the CTV market dynamics and why we at Magnite are so bullish. Programmatic is and will change the future of TV buying. The growth of connected devices, fragmentation of content, sheer price of television subscriptions, and advancements in ad tech will all lead us to where we sit today. A very pivotal time where brands are reevaluating their budget mix, and their old school ways of TV buying is being challenged. Technology is in fact disrupting linear. For the first time, traditional television is at a vulnerable state. Content options are creating streaming wars. It was only eight years ago that Netflix created the first original programming with House of Cards, and only four years ago that Hulu won an Emmy for The Handmaid's Tale.
Since then, many broadcasters and other apps have been invested in and launched. Cord cutting is at its all-time high, and probably the most important, audience expectations have drastically changed. Control, choice, and consumption, convenience of consumption, and the ability to binge is what matters now. The time of appointment viewing no longer exists. The MRC just recently released updated definitions of OTT and CTV, which we felt was really important to mention today. Just as Magnite has always defined from the beginning, the MRC now declares that CTV is streaming content on the big screen. It has to be viewed on a television set. Content viewed via Hulu on a mobile device is not considered CTV. It is important that the MRC has made this clear, as people had started calling the broader umbrella of OTT as CTV.
It's key that CTV is displayed in the living room, and this is the premium environment that brands are looking for as they transition linear dollars over to digital connected TV. We have three key predictions. One, all TV will be CTV. That is that all TV content will eventually be consumed over IP address. two, the future of CTV is ad-supported. The economics of ad-supported content will prevail, and the ad experience will win more viewers. Lastly, that all CTV will be transacted programmatically. The monetization of the future is about programmatic. The detail and data that it brings is just as important/more important than the shaking hands of making deals of the past. Eight out of 10 households now have a connected TV. 6 million cut the cord in 2020. That works out to about 11 cord-cutters per minute.
Just in 2021 already, we've seen 1.6 million cut the cord in Q1, and another 1.2 million cut the cord in Q2. According to The Trade Desk, 42% of cable subscribers plan to cut the cord in the next year. Cordless TV consumers are on track to become the predominant TV consumer viewer in the next year. Cord-cutting accelerated as television programming, such as live sports, became unpredictable through the COVID-19 pandemic as consumers' hunger for on-demand grew. This also was helped by the ability to get live sports via different applications that did not require a CTV cable subscription. The shift of CTV appears to be solidifying with the majority of TV viewers 18 - 34 and 35 - 54 already cutting the cord. These are the key groups that advertisers wanna reach. Consumers don't hate ads.
Remember, this is the way we've all consumed content since the beginning of when the first TV spot aired in the 1940s. Most understand the value exchange, and that ads enable consumers to get content at a much lower price, specifically when regarding to cable subscription prices today. Research from both The Trade Desk and from Magnite found that many would rather watch ads and pay less, hence why we're seeing the continued success of many applications on connected devices. Marketers like the ease and efficiency that programmatic brings. More buyers are planning on moving dollars over to CTV via programmatic means in the next year than anything else. They are also stating that over 69% of ads will be spent in programmatic on CTV this year. The SSP is becoming more essential for publishers to monetize their content.
The value of the SSP is changing with CTV and linear merging together. There is a need for publishers to optimize across all channels, auction or not. The SSP technology supports what the new upfront, the new direct, and the new way to monetize is going to bring these premium content owners. Technology will and will only continue to advance this in the years to come. The SSP is synonymous with programmatic. The technology is needed more than ever. An SSP is necessary for our publishers to successfully monetize all their inventory, whether that's programmatic or direct. As the industry shifts towards more and more to programmatic, the SSP will only continue to become more valuable.
Bringing unified auctions, reach frequency management, the ability to create addressable audience and improve their targeting, precision, control and real-time functionality are just a few features that our platform enables for our clients to be successful. With that, I will pass over the baton to Adam Soroca, who will dive deeper into our product initiatives across all of our platforms.
Thanks, Katie. Good morning, everyone. I've been at Magnite a bit over four years now. I came over to Magnite by way of the nToggle acquisition, where I was the Founder and CEO. nToggle was a machine learning-based traffic shaping platform that's now deeply embedded into the Magnite infrastructure. I spent the first four years of my time at the company heading up the buy side of our business. Me and my teams were in market every day talking to brands, agencies, and DSPs, and of course, working with our seller teams to bring demand programs to their clients. I'm really a career product guy. In the new role that I've taken on, the pulse and voice of the market has never been closer to the Magnite product and strategy team.
In my section, I'm going to touch upon our product vision, do a quick programmatic primer and identify where value is created. We'll cover our CTV and DV+ platforms and talk about how it all ties together with a next-generation audience targeting and addressability platform. We'll summarize our key product investments over the coming years. It's talked about for decades as the Holy Grail. Deliver the right message combination to the right audience with proper reach and frequency controls across multiple formats like television, desktop display, and mobile. Fulfilling this vision is what will make marketing budgets the most impactful and the most efficient. Cross-format publishers have sought to provide such a solution to their advertiser clients and what advertisers have been after for years.
Sophisticated publishers and advertisers immediately honed in on this potential for Magnite with the incredible assets that we amassed via the Telaria merger and which got even stronger with the SpotX acquisition. Magnite serves ads across all the media formats today. The 60-inch living room glass on Monday Night Football, we gain consumers' attention surfing the news at work, and we run video interstitials in the most popular mobile apps. We possess the industry's leading audience and targeting platform with a clear vision for the potential cookieless future.
Magnite will be the one in a consumer-first, privacy-centric way to serve up the BMW roaring down the road in surround sound in the living room. Provide display ads to attract users to click to the BMW site for deeper research on their laptop, and finally, to drive them into the nearest dealership for a test drive via an ad in the latest mobile app, all while connecting BMW to their target audience, highly affluent in-market auto intenders, controlling the message sequence and the number of exposures. Talk about impact and efficiency. This is a premium advertising experience which requires state-of-the-art reserved auction tools to allow BMW to execute with publishers. I'll take you through the path Magnite has embarked upon to deliver this vision.
This section covers programmatic advertising mechanics, areas that create meaningful enterprise value, and a readout about what we've heard from our sellers and buyers. What goes into serving programmatic ads? Well, when that ad serves on your CTV, laptop, tablet, or phone, most likely they were auctioned off by machines. The opportunity typically starts with a seller's ad server. This is why Google historically enjoyed success as the on-ramp for impression opportunities. Over the last few years, publishers helped to foster a new model that pulled the programmatic decisioning out of the ad server into a new layer. No longer was the ad server the dominant force. From this programmatic decisioning layer, advertising opportunities are made available to exchanges who send the request out to demand-side platforms or DSPs such as The Trade Desk.
We now have well over 150 DSPs connected into our platforms. Those DSPs house the campaigns, bid prices, and campaign objectives for the brands and agencies. The DSPs respond back to the exchange to run an auction to produce a winner back to the programmatic decisioning layer, who ultimately returns the opportunity to the ad server. All this happens in 100 ms and across our platform nearly 300 billion x per day. Where's the value created? Magnite has built products and acquired companies that allow us to participate in the more interesting and moving forward, the more strategic programmatic transaction components. We're now serving all media formats, CTV, display, audio, video, and native across any device globally. In CTV, we provide ad-serving capabilities by way of our recent SpringServe acquisition.
We power the more contemporary programmatic decisioning layer in both CTV and traditional media on DV+. We operate the world's leading independent exchanges with innovative seller and buyer capabilities. This sophisticated stack, particularly in our reserved auction business known as private marketplaces or deals, where sellers and buyers negotiate ahead of time, secured Magnite's position with few exceptions to work with all major sellers. Together, our platforms feature state-of-the-art audience and targeting capabilities that are the foundation of our next-generation cookieless audience platform. Our massive scale requires highly specialized infrastructure and traffic-shaping capabilities. This allows us to filter out non-monetizable inventory at our front door and only bring the good stuff through our entire system. Machine learning-based traffic shaping also provides our DSP partners with optimized traffic based on their particular demand needs. This is just one example of how Magnite maintains specialized connections and roadmaps with our DSPs.
We've led the buy side consolidation known as Supply Path Optimization or SPO as a leading partner to the major holding companies, agencies, and brands. What do these sellers and buyers want to achieve with programmatic advertising? Sellers look to maximize revenue from ads that run adjacent to their content. They do this with tools that intelligently manage yield, enforce business rules, and protect user experiences. They look for their programmatic partners to bring new revenue opportunities through new data and new formats. On the buy side, the primary objective is to maximize the efficiency of every dollar spent through gaining price advantages based on spend levels, deploying the best data to make the smartest buying decisions, and consolidating through the best channels. Means the most advanced SSPs that add the most value to every impression.
As an SSP, we're most focused on helping sellers win, but also helping buyers spend smarter. This means building essential tools for sellers and buyers so that they decide to steer their traffic and their dollars our way. I'm gonna start by talking about how we do this in CTV. The CTV landscape is incredibly complex. The content is highly produced. The patterns and consumer behaviors are quite different than traditional media. This, of course, allowed Magnite to build a moat around our CTV business because it's not as simple as taking a system built for desktop display and pointing it at CTV. CTV requires high-end tailored capabilities to allow sellers to monetize highly coveted inventory with an imperative to protect their consumer relationships. Buyers need a simple solution to unify real CTV inventory access with the tools to help them deliver smoothly on long-form video.
As I mentioned, CTV is fundamentally different than serving banner ads. It's a highly fragmented landscape because of the way content makes it to consumers. There are content providers, there are device manufacturers, there are MVPDs, all who touch the consumer and ad experience. Imagine an ad that gets served on the Pluto app on CNN content on a Samsung device. This high-value supply leads to higher expectations from sellers and buyers. It's 100% share of voice. It's displayed in the living room on the big glass. It's often served up during prime time, and it's often served up in live viewing. Of course, it's scarce supply. It's quite heavily supply-constrained. The sellers hold the user experience in the highest regard and deeply protect the consumer relationships.
After all, Samsung dropped serious money to acquire that 4K 60-inch crystal television consumer. As a result, we see the sellers behaving a bit like walled gardens. This premium medium deserves premium tools that simplify the complex landscape and premium prices rooted in protecting the user data. Magnite's CTV product set is underpinned by the SpotX and Telaria platforms. We now offer the best of the best inventory and capabilities, enabling us to compete vigorously with the likes of FreeWheel and Google by delivering unique value for our sellers and our buyers. As a product lead, one of my jobs is to sort through the best of the best assets in the industry that we now own to deliver leading CTV advertising experiences. That premium environment I described is powered by the basics like reserved auction management tools, competitive separation, position with the ad break, and much, much more.
Both companies tackled similar challenges, but each shined in quite different ways. SpotX features an advanced audience platform, while Telaria's real-time reporting tools prevailed as the gold standard to help sellers and buyers effectively deliver television campaigns. The comprehensive supply and tooling make Magnite the natural go-to partner for the major buyers across the globe. Now we'll focus in on the ad server and programmatic decisioning layer that SpringServe powers. The SpringServe ad server and proxy enables us to become an even more integrated partner with our sellers. This contemporary CTV ad server built over the last 18 months is state-of-the-art, with an incredibly high velocity to deliver customer features. For those clients that already have an ad server, the SpringServe proxy provides programmatic decisioning instead of the ad server. The SpringServe proxy is technology that pulls forward the programmatic decisioning outside the ad server.
It makes the calls out to the demand partners, such as our exchanges, and enables that demand to see all the inventory opportunities and compete at direct sold priority. The proxy ultimately allows us to win strategic seller positions without the ripping and replacing of the ad server. Sean will go into additional details on the proxy and ad server in the go-to-market section. Tying our product investments back to what our sellers and buyers want, these investments allow us to add more supply and win more buyers selecting Magnite as the preferred exchange. On the sell side, we provide superior yield through our purpose-built algorithms to fill pods and handle the specific nuances like live TV. Our tools provide sellers with in-depth controls to maximize their yield while also protecting their subscriber relationships.
We bring them new revenue streams through our work around audience and data, which I will cover later. Of course, deliver new spend through buyer marketplaces that we're building out for the largest buyers in the world. For buyers, we're their go-to choice to access CTV. We provide them with unique targeting opportunities and audiences that can only be accessed through the sell side. We provide curation tools to build their marketplaces and automated tools that enable negotiation of volume-based media discounts. The sellers and buyers, of course, serviced by the industry's leading operations team and toolkit with more experience than any other company. The insights this team uncovers helps the sellers to earn more and buyers to navigate the supply landscape. The sheer nature of live TV requires sophisticated real-time reporting. Helping our sellers monitor this real-time delivery is core to our tool set.
We've become the go-to interface for our partners to log into. Our seller clients log into our platform on average 5x per day for an average of 12 minutes per session. They can literally see the money accruing in real time, and if not, they troubleshoot as needed. The key CTV investments put more money and controls into the sellers' hands, and on the buy side, provide efficiency tools that lead to spend consolidation on our platform. On the sell side, in the exchange, we continue to make enhancements on how we handle live TV, especially in our calls out to the DSPs. The spiky nature of live TV is just wildly different than serving banner ads steadily throughout the day.
We're building features to maximize pod usage to ensure the space yields the strongest results, and we're investing in more sophisticated integrations into our ad server and proxy, as well as other proxy layers. On the ad server and proxy front, we're building to offer publishers an alternative to the legacy ad servers. With more and more linear money moving to CTV, we're building out capabilities around forecasting, booking, and reservations. In particular, programmatic guaranteed will get significant investment to serve as an on-ramp to programmatic benefits. On the buy side, we're investing in tools that allow buyers to achieve discounts in exchange for volume agreements. We're building out marketplace tech to allow buyers to better curate supply, and we focus on leveraging our traffic-shaping leadership on our CTV platforms.
We're working towards emerging with a single CTV platform that combines the best from SpotX and Telaria while powering the platform with our next-generation audience capabilities that I'll take you through shortly. Turning our attention to DV+, the platform serves all other programmatic media, display, online video, mobile, audio, digital-out-of-home at incredibly high velocity. The platform processes 8 trillion ad requests per month. J. Allen will cover the instrumentation on the platform to operate at high volume and low cost a bit later. The DV+ business is segmented out into two distinct lines, open auction and reserved auction. DV+'s open auction success is a result of the tooling we've built for buyers who turn off other exchanges in favor of the likes of Magnite or supply path optimize.
Our buyer-facing capabilities deliver better outcomes as a result of our direct supply integrations and machine learning platform. For sellers, we provide all the controls and tooling to set floors, select which advertisers appear on their site, and gain valuable reporting insights to run their business. To win more open auction business, we're investing in performance advantages. With 8 trillion opportunities per month, small win rate improvements yield significant revenue and benefits for all of our partners. Our buyer-facing investments allow us to win more share as SPO continues. DV+ also led the reserved auction category amongst the peer set over the years. Although much of this money currently runs in Google, the publisher tools that we've built provide one of the only viable alternatives to Google, and our Prebid investments have started to unlock those dollars.
Prebid launched in 2017 as a direct result of the seller community's distaste with the ad server and other proprietary wrappers controlling the programmatic decisioning and gaining unfair advantages. Magnite, the Rubicon Project at the time, co-founded the open source initiative, which quickly became the header wrapper industry standard, and we continue to be a leading contributor to the code base. Prebid enables publishers to operate a standardized code base that won the position as a walled garden alternative. There's no question that our open market success over the years came in part by the fair democratization of supply access that Prebid facilitated. As we look towards the next battleground, Prebid is helping to unlock reserved auction dollars for our exchange.
Not all sellers want to operate Prebid code, which led us to launch Demand Manager, software and services that make it super easy to operate the wrapper for our publishers without engineers. While Demand Manager generates software fees for Magnite, its value is far more strategic by locking in Magnite as the seller's priority exchange. As the priority exchange, we protect existing reserved auction spend and become a center of gravity for new reserved auction spend to migrate towards. Demand Manager also adds to the expansive roster of DV+'s strategic publisher integrations, which helps to accelerate SPO in our direction. Joe will talk more about Demand Manager successes and the SPO flywheel it enables in his section.
Once again, to connect our capabilities back to what sets us apart to win more sellers on our platform and be the buyer exchange of choice, for sellers, we've connected all the potential demand in market through highly optimized connections. We provide more and more software for sellers through Prebid and Demand Manager to take the programmatic decisioning layer into their own hands. This again enables us to migrate reserved auction spend away from the incumbent. Publishers are also the beneficiaries of our private label marketplaces as the world's largest buyers concentrate their spend on fewer premium publishers, clearly playing to our current product and market position. On the buy side, we continue to lead the industry with our results. This is much attributed to the algorithmic development powering our platform. As buyers take spend allocation decisions into their hands, they favor Magnite at an accelerating rate.
We power our buyer partners with those private label marketplaces to aid in curation and achieve price advantages on our exchange through these marketplaces and automated media discount tools. The DV+ business also provides both sellers and buyers with robust tools and dashboards to log into. Sellers are setting floors using our powerful reserved auction tools, including loading in their audience data and gaining valuable insights to optimize revenue. The tooling we provide to help buyers curate inventory that complements their DSP targeting has become invaluable. For example, buyers can compile packages to reach certain viewability thresholds, are leading legal drinking age approved, and basically any other grouping of inventory that they can imagine. Once skeptical of another UI, buyers now lean on our tools and UIs quite heavily. Lastly, these tools are foundational elements for the private label marketplaces we've been building out for buyers across the globe.
On the DV+ platform, we also invest in seller tools to maximize revenue with sophisticated controls and buyer tools that make Magnite the optimal choice to consolidate spend upon to gain efficiencies. For sellers, we've been revamping our reserved auction platform with more contemporary workflows and diagnostic tools. Demand Manager investments include more yield management automation and connections into the ad server, making it even easier for reserved auction money to flow to our platform. We're also evaluating new formats to bolt into the platform, like streaming podcasts, outstream, and going deeper in native. For buyers, we're investing in software to provide greater curation capabilities that allow even more granular targeting. We're upgrading our APIs for our demand integrations and refining our private label marketplace tools for our largest buyers.
We're investing in bringing our world-class machine learning capabilities, hardcore speed enhancements, and infrastructure tuning to give our exchange advantages in header auctions. Of course, spanning the platform is enabling audience and preparing for a cookieless future. The two-year third-party cookie reprieve affords us the time to properly prepare for this potential future. Keep in mind, the cookie remains alive today, and though likely to go away, may not go away entirely, albeit it will play a diminished role. There is a real opportunity to build audience and targeting foundationally on CTV today. This tectonic shift and the sheer nature of CTV seller fundamentals, going back to protecting the consumer data, will create new value on the sell side. For a glimpse into a cookieless future, this identity matrix describes a likely outcome and highlights our investments.
Moving across the axis, you'll notice the trade-off between accuracy and scale. Starting with the lower right-hand corner with the industry IDs, today we work with all the major identity providers, making those IDs available in the bid stream, including UID 2.0. These are just four examples, though we're instrumented to pass along virtually any IDs available. These IDs rely upon user logins, high accuracy, but we only expect that to cover 15%-20% of the web opportunities, low coverage. With CTV potentially being even lower, given the CTV seller operating principles, which opens up a massive opening for Magnite. Moving to the opposite top left corner to the Privacy Sandbox, we're of course influencing and helping to make it a workable solution for buyers and sellers. When the cement gets a bit more solidified, we'll of course support it with our platform.
The areas Magnite is investing in are publisher first- and third-party segments and data clean rooms. Audience targeting for CTV is the cornerstone of our strategy as the sellers lean heavily on exchange partners to protect their consumer data. Today, our Audience Lock product obfuscates user IDs, yet provides buyers with signal to reach targeted audiences. Procuring audiences on the sell side in CTV highlight two very important markers. First, sell-side audience onboarding and activation is a seller requirement, which represents a new revenue stream and value creation. Second, our CTV audience work today allows us to move seamlessly into the cookieless web at scale and sophistication that no other platform can match. Given that we're required to obfuscate users, we've had to reimagine how we provide buyers with targeting signal to reach their intended audience.
Magnite has onboarded the leading DMPs, third-party data sources, publisher and advertiser data into our audience management platform. Providing this service aids sellers to close larger, more interesting deals for their clients and buyers to turn to Magnite for their targeting needs. Sean will talk about the growth and success we've enjoyed here shortly. Adding more data partners with more precise onboarding signal from this multi-billion dollar industry is a key investment area. Our aim is to make sell-side data enablement a clear revenue driver for our sellers and for Magnite. Honing this model for CTV lends itself quite well to the cookieless future, where sell-side data activation will be a requisite. One of the technologies we're building to onboard data is our light clean room, more of a secure matching service.
Built to ensure seller and buyer data are protected, the service allows for encrypted email addresses or other string matching without the core data ever leaving the seller or buyer's walls. Using multi-party computing , our software creates a matched stable ID. This means that sellers with registered users can match to a buyer CRM file and virtually any other data source with corresponding user information, thus protecting against data leakage or any identifier reconstruction. This secure service helps us to get inside the CTV walled gardens as the seller's trusted partner to onboard data and generate new revenue. We'll leverage this platform to securely deliver targeted ads to users across all screens in a consumer-first, privacy-respectful way. Turning our attention back to the web world, with the waning of the third-party cookie, buyer's ability to create audience segments will fade.
Publishers' direct consumer relationships and observed behaviors have never been more important. We're investing to help publishers group users into common segments, such that a sports enthusiast means something similar across The New York Times, The Wall Street Journal, and Daily Mail. That we can federate those segments to DSPs such that buyers can reach a standardized segment across our publishers at scale. Our work here is part heavy machine learning to create segments, part mixing data from our other datasets along with modeling, and part helping the industry standardize on a common taxonomy. This latter work is heavily driven through Prebid. This will result in high coverage and high accuracy audience targeting necessary to make the ecosystem effective once the cookie fades.
To recap our key audience investments, we're building out audience as a shared component that all formats and supply touch points from ad server to proxy to the exchanges will leverage. We're working with the web-based world to help them turn publishers' consumer relationships into valuable targeting opportunities. The audience platform will house robust targeting and datasets, as well as enable frequency management and sequential messaging to span all our formats. This platform underpins the BMW example I illustrated in the opening remarks. Our role in CTV puts us in a position to drive the industry in this new direction of onboarding the most interesting datasets onto the sell side while creating new value. How do we deliver that vision for BMW and our other amazing marketers to run targeted campaigns across our sellers?
On the sell side, we're building out more software capabilities to take on programmatic controls through Prebid, Demand Manager, and our proxy in CTV, which locks in and opens up the reserve auction business. To win on the buy side, we provide the leading capabilities such as private label marketplaces, tools to realize automated media discounts, and access to the most strategic seller integrations, all tied together by our cookieless audience targeting that centers around CTV and unlocks value for the sell side. We're focused on accelerating efficiencies and infrastructure advantages to gain a competitive edge. Lastly, emerging with a next-generation CTV platform to realize synergies, which Allen will cover off on now. With that, I will turn things over to Allen Dove, our CTO.
Thanks, Adam. Great to be here talking to you guys. I'm Allen Dove. I am the former CTO of SpotX and now current CTO at Magnite. I've been in the ad tech industry, if you can believe that, for going on 20 years now, from the days of image-enhanced pay-per-click all the way into, of course, video and CTV. Been a part of many of the standards bodies over the years, from the formation of VAST 1.0 all the way through. Couldn't be more excited at the ability to lead Magnite from a technical perspective and work with the rest of the executives. Today, what I want to be able to do is take you through our technology platform and the footprint, and really give an understanding to how we do the vision of the company, the product, and the business have combined.
Our platform is a globally distributed and fault-tolerant footprint. We take millions and millions of transactions a second, and those are all, of course, IP-based transactions globally. The concept behind the platform is to provide a very dependable platform that enforces everything from availability and integrity and reliability, and security, et cetera, to be able to take these ad opportunities across any device, any inventory type, any place around the world. We do leverage a hybrid on-prem plus cloud infrastructure. The idea there is that we get the cost benefits of on-prem, as well as the full performance of on-prem with the burstability and dynamic capabilities of cloud, and I'll talk more about that as we go.
We have great expertise in understanding cloud scaling and cloud burstability, and how we actually leverage the cloud for the things that it's really great at for us, and yet be able to control those costs. We've, we have a lot of experience in running on-prem infrastructures as a, you know, number of the companies that have formed Magnite started long before cloud was a thing, and the cost benefits and our expertise are great there for us at this kind of scale. Probably one of the most important parts of Magnite as a whole is that we have all these great market-leading engineering resources and product resources and business resources, but certainly on the engineering side that are no longer competitive.
We actually get to focus ourselves as a unified company in people, product, business vision, and focus on the same goal, and more importantly, expand that vision now to be able to include, you know, tackling an omni-channel world. You know, for us, Magnite, it is a privilege to be able to have such an arsenal of technology and capabilities and vision to be able to lead. Our people are among the best in their specialties across the board, and Magnite as a whole and a company is definitely on par with the biggest of folks in this market, and we consider ourselves, you know, a premium platform in the ad tech world. So scale everywhere, scale wherever you are. We take billions and billions of ad requests a day.
We do about 8 trillion-plus ad requests globally a day, 37 trillion bid requests or outbound programmatic requests to buying platforms a day that spread all across from North America to APAC and Australia and LATAM. You know, for us, when you talk about this type of scale, it is literally millions of transactions a second. Our global footprint, our ability and the technologies we have to control and control costs and maintain the speed and performance is really a part of what sets our platform apart. The one thing that's really important to think about is the fact that in computer speak, you know, a millisecond is an eternity, right?
We do all of this in, you know, milliseconds of time, where we're getting bids out the door and ad requests back. The one thing that to point out, though, is that 37 trillion outbound ad requests, those would be significantly higher if we didn't have a lot of the filtering technologies, both on inbound ad requests and outbound bid requests that Adam had mentioned earlier. For those technologies, our filtering capability, what it allows us to do is really be able to say to any publisher, "Send it, send it to us all.
We'll help you find the value within those ad requests. We wanna be able to turn every ad request we have into something valuable. Then the things that maybe buyers aren't interested in or buyers may not quite have the same type of bidding for. We actually are able to take and filter that down and really only run the inventory or run the things that we know are extremely valuable. This filtering technology is part of what brings us and helps us achieve scale and protect our partners in the ecosystem as well from burstability. One of the things to think about in burstability is that much of what we do, certainly in the CTV space, is highly concurrent, meaning millions and millions of requests in an instant.
Our filtering helps us control that. Our points of presence across the world, this is places where we have, both, Magnite-owned hardware, footprint, technology, compute infrastructure, as well as where we leverage cloud. U.S., we, you know, we have multiple data centers across the United States, certainly in APAC and EMEA. Really what this does is it gives us, again, we talk about low latency, when it comes to buying and programmatic exchange and getting bid requests to and from buyers, every millisecond matters. Every millisecond that you take is more time that a CPU is burning, et cetera, more time that someone may be waiting, more time it can certainly impact the user experience. Well, we've put our data centers as close as we possibly can be to bring that latency down.
You know, a la when New York and Chicago spent billions of dollars to reduce latency 3 ms for trading. You know, we have the same types of advantages here in ad tech, and certainly where Magnite has placed our facilities. Engineering resources across the board and global. You know, as I mentioned, we have experts across their fields, data science, software and engineering, architecture, systems operations at scale. You name the discipline, Magnite, we have that and available for engineering resources, and we've been able to take market-leading companies, you know, SpotX, Telaria, legacy Rubicon, now all within Magnite, SpringServe, and bringing all of this expertise together.
We have engineering, 400+ engineering and products spread across the globe, of course, in most of the major cities here in the U.S. We also have offices in Belfast in Northern Ireland, London. We have a Kyiv footprint as well. We have folks in Tokyo, and Singapore, and Australia as well. You know, again, the people that we have and the expertise that we have and the, you know, the knowledge that the engineers have at this kind of scale and within ad tech is you know, it's formidable. I definitely put these folks up against any in-market. We're very fortunate and, as a company bring a lot to bear from the engineering resource side.
Talking about, you heard Adam Soroca talk a lot about the things that our platform does, the form and function, what we bring to the various markets from display all the way to CTV and OTT and audience, et cetera. The tactical advantages, you know, being an omni-channel company is one of the most exciting things for me to be able to be a part of helping to lead the technology for. We have proven global and extreme scale. We have unique technologies to help us maximize that scale for both sellers and buyers, help the sellers maximize their inventory, and have the buyers find exactly the audience and the types of folks that they want, the lookalikes.
Our inventory reach is, you know, it's very broad from, you know, as Adam told the story, from the display world all the way down into your watching on your big glass. Audience is everywhere. With our omni-channel and footprint and the amount of data that we're able to bring in and be able to leverage and help buyers and sellers, it very unique within the market. Our unified insights, all the inventory types, yes, they're different in how they're delivered, and they're different in how you can interact with audience.
All of that is still data, and all of that can be tied back in real time to find those pockets of extreme value, to leverage scarcity and behaviors within a publisher's inventory and gives us an ability to really truly bring real-time actionable intelligence to bear. You know, we have a great deal invested in our machine learning capability, which I'll talk a little bit more in a bit. This concept of holistic inventory modeling, how do we actually look at and help publishers look at their direct and reserved campaigns and how it blends with programmatic and how it blends from, you know, you name the thing, time of day, you know, the old joke of weather parting, whatever it might be. We have that capability in real time across an omni-channel footprint, an omni-channel interactivity.
CTV and OTT, of course, that's, you know, legacy SpotX, legacy Telaria, SpringServe now coming in. You know, it's not just a scale game in the CTV and OTT space, and I think that that can get lost. You know, even the display world and the OLV world, the online video world, the native world, et cetera, a great deal of scale. CTV brings the scale as well as the complex ecosystems behind it from trying to understand audience modeling, keeping audience privacy compliant, helping publishers and even buyers bring their first party and very, very, you know, secret data, if you will, into play and maximize, monetize. SSAI environments, how you actually get things and, you know, very high-quality video ads, blending all that in real time.
This concept of abrupt concurrency when you have live events, very complex ecosystem. You know, for Magnite, we have a great deal of experience, of course, in the CTV and OTT space. Live events at high concurrency, you know, a millisecond or a second is an eternity in computer speak. When you have these live events and concept of all of the folks watching a sport event, goes to break, bang, it all hits an infrastructure at 1 time. We've got experience in that. We've done that for college football, a number of live sporting events. We've built infrastructure and ecosystem to help us protect our buying partners and other folks as well in these instances.
The ad server and programmatic decisioning layer that SpringServe brings into us, it really is a key component to helping us model that direct and programmatic world together. The advanced insights, our ability to help publishers find pockets of value for market making and offering to advertisers, to helping advertisers find their just exactly the audience they want. That concept and that programmatic decisioning layer really brings us and puts us in a pole position with publishers as well. Our audience capabilities, as Adam had mentioned, you know, we've spent as independent companies, each company spent a lot of time thinking about audience and how we're gonna maximize audience and leverage audience. And for Magnite as a whole, bringing all those together now, we have just this amazing intrinsic capability to help publishers leverage audience. And the capability is audience context and content.
We all know that as the cookie disappears, whenever that might be, that's really what buyers care about, is how do we get there? How do we actually? You know, for Magnite, we've built our platform and are integrating our capabilities to be agnostic to source. It doesn't matter what the identity garden might be, whether it's, you know, from a cookie base, whether it's from first-party or publisher data, whatever it might be, we have a capability that's core to our markets intelligence. It's identity agnostic.
You know, our audience capabilities enforce provider control of that data, that allows publishers to feel very confident, and we end up in a trusted position of bringing their data that they can actually leverage and help drive value from, or buyers bring their data, help us drive value from it, et cetera, in our ecosystem, in our omni-channel footprint. Magnite becoming that authorized source, it's great value for us as a company, and it's also great value for publishers as well. Privacy and security. We talk about audience, but Magnite and Adam mentioned some of this as well. For us, our ability to leverage audience everywhere, we of course do that compliant to all privacy signals.
Our clean room gives an ability to further obfuscate data and keep you know, things in provider control of that data, and we're able to control access, you know, to who can access what sets of data based on publisher requirements. You know, our ability then from this omni-channel footprint to actually leverage all this data and build segments that buyers can trust is pretty critical and intrinsic to what we do as a company as well. You know, data intelligence, this is the area that's one of the most exciting for me. Yes, scale. We're expertise at scale. We know how to run this every day. We've done it for, you know, for a decade plus.
This concept of data intelligence and building actionable intelligence in real-time and near real-time, it's something that Magnite excels at, and we have from our filtering to our unified forecasting, to our ability to have real-time deal and programmatic debugging or direct debugging, understanding what's happening in real-time. We have this telemetry across all of our inventory that our real-time machine learning layer can leverage. One model rules them all for us, and it's a great way for us to bring all this data together and find the insights across it and let our, you know, data science expertise bring their thoughts to bear on this. We have forecasting potential across all inventory types. This omni-channel play is very unique for Magnite.
The holistic intelligence of market dynamics, audience correlation, the demand drive when things are interesting, you know, as Adam talked about, telling the story for that BMW buyer, whomever it might be, we have that capability in near real-time. Our revenue insights, you know, we wanna be the trusted tool for our publishers to help drive and know that the machine learning and intelligence we have are driving revenue insights for them, just-in-time monetization, and really helping them understand the true programmatic potential that they have.
You know, because that's a really important one of, you know, you have reserve and you have guaranteed deals and direct deals, but it's sometimes, you know, being able to understand how programmatic could play in and still let you as a publisher fulfill it all, a capability that Magnite brings to bear. We talked a bit about, you know, being an authorized source and a trusted partner for data for our publishers, but, you know, for us it is, it's about inventory assurance. At Magnite, we take very seriously, you know, traffic quality and inventory quality, and we've been a part of, you know, founding the ads.cert movements inside the industry, ads.txt, ads.cert, those types of things. It's a constant focus for Magnite.
We have people and we have tech that we dedicate to that, custom and third-party tech for us to ensure that we have the cleanest inventory possible. We really do see ourselves as a trusted omni-channel marketplace. You know, from Demand Manager to Prebid Server to our, you know, Prebid wrapper, the programmatic SpringServe proxy, it gives us an integral position with our publishers and our partners. The idea is that we're able to leverage any ad server. The SpringServe proxy can leverage any SSP a publisher wants to bring to bear. Prebid the same. You know, it helps us become part of the top of the publisher's monetization stack and really gives us a trusted position with our publishers.
Thought leadership, you know, Magnite chairs and leads a number of industry groups from Prebid to the Tech Lab, the IAB Tech Lab, et cetera. We've been a part of the foundational work of VAST and AdCOM and OpenRTB and ads.cert Prebid. You know, this all becomes very, very important as we start talking about broadcast stacks and converging with IP-based and programmatic demand, et cetera. You know, as Katie talked about, we all believe and we know that what's gonna happen is that programmatic and the SSP become even more important as time goes on. You look at the education of looking at the legacy broadcast stacks and the playout systems and those things.
As it converges and blends, Magnite has the capability of bringing all of that and all the same types of technologies to bear for the IP and connected side as well. The platform integration opportunity, and one of the things that we have been focused on is the unification of knowledge, effort, and technology. You know, the unification of people is amazing, and now it's time for us to turn all of our folks to bear on bringing forth our unified vision as a company. You know, Magnite has two industry-leading CTV platforms to integrate form and functionality, and, you know, we are going to be blending and are in the process of blending the best of both.
You know, for us, it's about providing our publishers an easy way, short and long-term, to leverage all the technologies you've heard about and all the features you've heard about across the Magnite platform, the easiest way to bring them to bear and market. As we go forward, over the next year and we bring these pieces together, audience, you know, our audience piece is underway and the data integration's happening already. As we bring this to bear, you will start to see us put together easier and easier ways for our partners to leverage all of Magnite and leverage everything we have. It really is about bringing this omni-channel data and this omni-channel opportunity together to create powerful and actionable intelligence.
You know, rapidly merging the core of our data flows into data science, pretty critical for us as a company, and really it is a part that is a next generation type of advantage that I believe Magnite has. Our year two synergies, you know, you've heard about where we were during the acquisition of SpotX, and our year two synergies and our savings are on track. We're making a lot of efforts there in both cost savings and then bringing people and now unifying vision to make sure that we're working on things that Magnite as a whole sees as very, very valuable. With that, I will now turn it over to Sean Buckley.
Thank you, Allen. Hi, everyone. I'm Sean Buckley, Chief Revenue Officer for CTV. I joined Magnite through the SpotX acquisition earlier this year, where I served as Chief Operating Officer, I'm very excited to speak to you all today about the opportunities in front of us. I do wanna take a minute to reiterate a few of the key points Michael made in his section as it pertains to the CTV market. While there is a level of open market activity, this is a business where most transactions take place via private marketplaces, different from other formats in digital media. As a result, deep and long-term strategic relationships with a consolidated set of trusted technology partners is a common path. As you'll see in the overview to come, this dynamic has a heavy influence on the approach and strategy for CTV specifically.
I thought it would be helpful to start by laying out a visual of our positioning in the industry. Media owners sit at the core of our business, and with the acquisition of SpringServe, Magnite now offers a modern feature-packed ad-serving platform alongside and tightly integrated with our market-leading programmatic capabilities. On the buy side, we've developed a robust services organization delivering meaningful value across the spectrum, from brands to agencies on both a national and regional level while working in close collaboration with our demand-side platform partners. Audience enablement presents a unique point of differentiation for us with both the buy and the sell side, given the breadth and depth of our presence in the CTV space, which I'll touch on in more detail later in the presentation. We break the CTV market down into four distinct groups in terms of understanding and analyzing our customer base.
The first group, which has become increasingly important in this space over the last few years, are the device manufacturers. Through sticks, boxes or smart TVs, these companies play an important role with consumers as distributors of third-party streaming apps, for which they often receive an inventory share in return for that distribution. Most have also come to market with their own free ad-supported streaming apps, also known as FAST services, and their valuable first-party data assets present an additional value proposition to the market. The next cohort of partners are the virtual MVPDs, who bring the multi-channel viewing experience well-established in traditional TV into the streaming space. Through carriage agreements, these services are typically monetizing portions of inventory across the programmers who are made available through their apps. Digital-first services built from the ground up alongside the development of the streaming ecosystem represent the third cohort.
As AVOD viewership has taken off, many of these services have reached impressive scale, reinforced by some of the M&A we've seen in this category. Finally, the programmers and broadcast names we're all familiar with. While these companies represent some of the largest inventory sources in the space, we're still in the early innings of programmatic monetization for many in this group. As more of the business transitions in that direction, this presents both a significant shift in how TV advertising is transacted and a huge opportunity for growth. I'm proud to say all of these companies featured here are active Magnite platform customers, highlighting both the breadth of the ad-supported CTV market and our opportunity across the full spectrum of constituents, especially on a global basis. Magnite offers a very flexible set of pricing terms for our customers based on the service level and transaction type.
It's important to note that platform customers often operate across most of these pricing models, which can vary campaign by campaign. To touch on each category, ad serving represents the most fundamental type of transaction in our business, where the publisher is given an ad tag or a creative asset directly from their buyer. These are non-programmatic, traditional direct sold campaigns. Here, we'll charge a CPM-based ad serving fee. One-to-one private programmatic transactions that the publisher manages themselves end to end with their buyer is the next step. An example could be publishers managing their upfront deals as that business moves to include programmatic transactions. Next are programmatic deals leveraging the support of our demand facilitation team through sourcing the demand or providing hands-on support during the execution phase.
These two transaction models leverage deal IDs, and the latter can also include multi-publisher deals we structure with buyers to improve workflow efficiencies. The open market is an aggregated marketplace where publishers can choose to make some or all of their inventory available for monetization to a wide variety of buyers. Managed service offers the highest touch transaction model on our side, where we're able to provide buyers with a full service offering, deploying their campaigns across the broader CTV landscape through our platform with tailored performance and delivery goals. This is a popular model for onboarding buyers just getting involved in programmatic CTV or those looking to be provided with a very high level of service and support. We've also structured tiers for fixed monthly recurring seat fees, which are built around the specific bundles of technology and services we're providing to media owners.
These are incremental to the transactional fees we've covered here. In terms of our end market strategy, I wanted to touch on three of the key areas that we feel differentiate us from others and create a durable advantage for our business: demand facilitation, SpringServe and advanced integrations, as well as our audience management strategy. Let's start with a deeper dive into our demand facilitation business or our DF function. The demand facilitation team is structured into four distinct groups focused on different buyer cohorts. It's important to note that we now have over 50 team members across these groups in the U.S. alone, and another 30+ outside of the U.S., while continuing to bring new talent on board. Years of building partnerships and refining our approach has created a team with established relationships and strong institutional knowledge, which has proven quite difficult to replicate.
Team members are also equipped to work with buyers across all formats and transaction types, creating strong economies of scale for us. I'll provide a high-level overview of each group, starting with brand partnerships. The trend of brands leaning into programmatic and bringing more responsibility in-house has been developing for years, and the pandemic further accelerated things in many cases. As brands take on these additional responsibilities, interest in consolidating technology partners has grown, and the services our team offers can provide meaningful value, from inventory discovery to sharing best practices and marketplace trends. Magnite is deeply engaged in working with leading brands like T-Mobile and Bayer.
A lot of attention is focused on ad dollars shifting from traditional TV into streaming, and rightfully so, what may still be underappreciated is the opportunity for thousands of brands where TV is not an ideal or economical fit. For example, brands focused on niche audiences, where the inherent addressability and targeting capabilities of CTV is a game-changer. This opportunity provides an additional tailwind and also favors an aggregated inventory partner, like Magnite, to help more efficiently find those audiences and help deploy advertising programs across the broader ecosystem. Our mid-market team addresses the hugely important set of regional and local buyers throughout the country, as well as smaller independent agencies for verticals like tier 2 automotive and travel tourism. Granular geographic targeting is a frequent need, which again favors an aggregated inventory partner like us in order to effectively scale campaigns.
Local TV represents a significant portion of U.S. television ad spend, and we're well-positioned in helping transition those dollars into the streaming ecosystem. Many media owners see especially strong value from working with our mid-market team, where it often doesn't make sense for them to staff sales personnel to sell only their own inventory for this type of buyer. They're able to lean on our established resources and partnerships to tap into these budgets economically. Our performance pod has experienced strong success tailoring our services for buyers with more of a direct response focus. This was showcased through our recent announcement with Quigley-Simpson, highlighting our ability to secure sizable spend commitments across formats given the service and results we're delivering for clients. We'll have more partnerships like this to announce in the coming months.
Agency development presents a slightly different approach in terms of the services we can provide to the market. Our team has established strategic relationships with each of the major holding companies, structured based on their strategy, preferences, and operational approach. While many of our platform customers lean on us to help open doors with the large holding companies and evangelize their offerings, larger media owners often have deep, long-standing relationships with these entities. While large media owners may negotiate business terms directly with the agencies, it can often be a challenge to get those budgets delivered and executed, especially with the complexity of programmatic. For agency buyers, we provide a unified resource and streamlined workflow to deliver, optimize, and troubleshoot campaigns, working together to develop product and operational solutions for challenges they may encounter. For media owners, we help ensure business commitments are met and delivered in full.
Our partnership with GroupM, which SpotX announced last year, highlights this type of relationship, and it's been a big success. Finally, our partnerships with demand-side platforms, or DSPs, are critical for the business, and we have a dedicated team to support them as well. Alignment is key, so staying in lockstep on product roadmaps to bring new innovations to market and cooperating on key industry initiatives are big focal points. While we've spent years building the deepest and broadest set of DSP integrations in the industry, our DSP team also integrates and onboards new partners when those opportunities arise. Now I'd like to cover SpringServe, along with our broader advanced integration initiatives, which present another key differentiator for us in market.
The SpringServe platform provides customers with modern ad-serving technology packed with the features to help them succeed in the rapidly changing media ecosystem today and long into the future. The ability to manage both traditional direct sales and programmatic transactions effectively in tight coordination and at a granular level is increasingly important and one of SpringServe's strongest suits. In conjunction with Magnite's programmatic capabilities, delivering a high-quality user experience during commercial breaks by eliminating duplicate ads and competitive advertisers from serving in the same pod is absolutely paramount. SpringServe provides customers with advanced optimization capabilities such as revenue per second, factoring in the optimal ad length of each ad delivered within the specified time allotted for a commercial break. Distribution rights and inventory splits between distributors and programmers is a critical need in CTV specifically, requiring a unique set of functionality.
Routing inventory based on rights agreements in an intelligent and streamlined manner is another area where the SpringServe team has invested heavily and built out a differentiated set of features. As you heard from Adam earlier, there are two main implementations for the SpringServe platform with customers, including the ad server and ad proxy. The SpringServe ad server has already become a popular choice in market, particularly among new media companies and those where tight coordination between traditional direct sales and programmatic deals is a priority. The advanced integration created between the SpringServe and Magnite CTV platforms, known as Total Connect Plus, provides platform customers with unparalleled capabilities and control across both transaction types. The ad proxy implementation is a newer structure and caters to customers looking for more advanced programmatic capabilities, more deeply integrated into their tech stack, but without making a switch from their existing ad server.
This is a stepping stone solution, particularly relevant for larger media organizations with pre-existing technology investments. The ad proxy setup brings a meaningful step forward in capabilities for our customers, and allows Magnite to scale these relationships from a much more advanced point. It also builds familiarity for more of our products, which is hugely helpful when full ad serving opportunities do arise. In specific circumstances, some publishers have opted to build out their own private label proxy layers, where we have also built advanced integrations for Magnite CTV, creating similar gains in capabilities and improved inventory access for programmatic monetization. Either through SpringServe or proprietary proxies, this new model represents a meaningful step forward for programmatic and CTV. Magnite is well-positioned and ready in either circumstance.
Advanced integrations as a whole have been a huge focus for the team, and the number of active accounts has nearly tripled from January through Q2, with most of those in place through the SpringServe implementations we discussed. There is much more opportunity in front of us as programmatically increasingly takes center stage within the CTV business. Audience activation and data enablement is the final area of focus I'd like to touch on. Our investments in this area started 4+ years ago as publishers with scaled consumer relationships expressed interest in activating their first-party data assets in a controlled and secure way through our supply-side technology. Preventing data leakage and managing this process in a privacy-compliant way are critical issues, and this is proving to be a preferred workflow for data activation in CTV specifically.
Interest in data activation over the past 12 - 18 months has grown to include agencies and brands on the buy side as well, who are integrating their data assets with us. Given the fragmented deal-driven nature of most CTV transactions, data integration through our platform provides buyers with material enhancements for planning and forecasting against specific audiences, given the broader inventory visibility we're able to provide. This helps give buyers a better understanding of where their audiences are across the ecosystem and better inform their investment decisions based on that. Refining data preemptively on the supply side can also produce more streamlined programmatic guaranteed transactions with the added benefit of audience targeting included. Along with our investments in infrastructure to support these capabilities, many of the partnerships we've established on both sides have taken years to build and develop.
These partnerships require a high level of trust and are unlikely to be replicated in many other places, given the focus on preventing data leakage. Data activation through SpotX's Audience Management Engine has been growing consistently for years, and activity has really exploded here in 2021. We expect this to be a continued trend in CTV specifically, and have a strong pipeline of additional buy and sell-side partners eager to integrate and activate their data with us. As you can see, the combination of our advanced technology and unique demand services, paired with the breadth and depth of our well-established footprint across the market, put Magnite in a great position for continued success. That concludes my portion of the presentation today. Thanks again for your time. From here, I'll pass it to Joe Prusz to discuss our DV+ business.
Hey, thanks, Sean. Appreciate it. Hey, gang. Joe Prusz here. I am the CRO of DV+, and I've been at the company for 13 years. The cool thing is I feel like we're just getting started, which is why I am very excited about running DV+ because it truly is the other growth story for Magnite, and it's very much a complement to accelerate our CTV business as well. Bottom line, we're working really hard to grow both. Okay, what is DV+? As Michael said, it's not just a display shop. It powers video, mobile, audio, digital out-of-home, out-stream, native, and many new emerging formats to come in the years to come, and it's massive, 8 trillion requests a month. Few, including Google, compete at this scale across so many valuable formats.
All right, back to the growth story that Michael Barrett mentioned earlier. We have high expectations from DV+. Our goal is to increase our market share from kind of sub 10% now to 20% in a market that is growing to $63 billion in the next few years. These are kind of the two core pillars that we're going to focus on, open auction and reserve auction. All right, starting with reserve auction. How do we win in this extremely valuable segment? In one sentence, we are obsessed with building the easy button for the open internet. Before I explain that, let me give you a quick history lesson. Google is pretty much the ad server of default for most publishers globally, especially in the U.S. and in most key international markets.
In the past, it was easy to use Google's ad server combined with their SSP, Google really is the easy button for these most premium, you know, reserve auction deals. No one got fired for using Google. However, publishers started to yearn for two very critical things. They wanted choice and they wanted control. They didn't wanna be locked into walled garden black boxes anymore, and they wanted more control of how they ran their advertising operations. Really, they needed a transparent and open source solution. That's how Prebid became mainstream. Publishers stopped relying on Google for everything, they massively adopted this new open source revolution, giving them the choice and the control they craved. I like to say in effect that, you know, Prebid really cracked open Google's ad server dominance.
Publishers no longer needed to swap out their ad server. They just needed Prebid to help open it up. Now Prebid technology, it's vast, a little complicated to maintain, which is why publishers started to reach out to Magnite for help. Their Demand Manager. Demand Manager is our software layer on top of Prebid, making it much easier to use, and that is why it really has become the easy button for publishers. I'll also explain in the next few slides why this is such a strategic initiative for Magnite, driving revenue growth above and beyond just tech fees, but more importantly, helping drive more overall spend into our platform as Adam outlined earlier. Before I explain that, let me explain why we keep talking about Prebid so much.
Prebid has become the dominant open source header bidder system used by the vast majority of premium publishers. As a reminder, a header bidder is what actually makes the decision of whose programmatic bid wins in the downstream ad server, it really is the most critical part of the programmatic ad stack. We co-founded Prebid with AppNexus a few years back to ensure that this revolution would continue to grow and be supported. As you can see from the companies on this slide, the entire industry has adopted Prebid, from the world's biggest publishers on one side to the world's biggest buyers. In effect, the open internet. The good news is that Prebid won. You know, 95% of publishers today have adopted Prebid. The not so good news, it's a little bit hard to use.
After pubs installed Prebid, they came to us asking for help optimizing it, and that's how Demand Manager came to life. Again, Demand Manager, our software layer on top of Prebid, making it much more efficient, easier to use and maintain, but it's more than just kind of advanced plumbing. It's a full suite of UIs and analytics so revenue teams can go back to doing what they do best, easily test, measure, and optimize their programmatic businesses without having to rely on engineers for code changes every other day. This, of course, drives more yield for themselves and more value for advertisers. All right, how's it going? Well, it's growing quite nicely, and it's winning against self-hosted Prebid, competitive wrappers, and a multitude of free options.
To date, we've signed over 200 customers with many of the biggest media owners in the world. What's really exciting is the amount of greenfield we still have left. We estimate that Demand Manager sees about 20% market share of all Prebid requests, so there's a lot more room to go grab in the years to come. As you can see from the light purple graph, our Demand Manager requests over the last 18 months have been going up and to the right. The bottom chart shows a proprietary wrapper fading away quickly. All right, what happens for publishers when they adopt Demand Manager? First, they take control back of their business, and second, they drive more spin into Magnite. These two charts represent two Comscore top 50 publishers, two of the world's biggest and most advanced media content owners.
As you can see in these charts, post-launching Demand Manager, Magnite's share of reserve auction spend skyrocketed, 8.5x on the left and 12x on the right. Really, here's the point that I wanna underscore. When our software and our services put Magnite in a position to become a trusted advisor for the publisher, driving insights and incremental yield, our share of ad spend grows along with them. Here is a final data point, looking across our platform. Demand Manager clients are spending 4.5x more in reserve auction than we see from other integration types. The great news about it is we win with both. Whether it's Prebid or Demand Manager, we win with both.
You can imagine why we spend so much time talking about this and putting so many resources into educating our clients on both. Let's switch gears for a minute, and let's talk about how we win in the open auction side of our business. This is where we have a lot of low-hanging fruit to grab. As you know, the open auction part of our business, this is where there's very little human interaction. It's all computers, software decisioning, algorithmic, and a speed play. We've been diving much deeper into DV+ recently, more than ever before, and I'm excited to share how we're gonna win here as well. These are the 6 initiatives that will help us win in open auction.
It's funny because at the end of the day, it's all about focus for us. You know, we've been on a tear acquiring, you know, many companies to become the biggest omni-channel SSP in the industry, which means now it's about focus, going much deeper into our core businesses, that's one of the reasons why we created two revenue orgs between Sean and I. We're putting a lot of attention of extracting more value from DV+, driving increased operating leverage for our company. Focus, focus. It's simple, it's true. Second leads me to advanced bidding algorithms. Think about the scale that we have, 8 trillion requests a month and growing. Machine learning, data science wins at this scale. This is about leveraging new technology to bid faster, bid more often, bid with more precision.
That leads us now to increasing our supply aperture. A buyer can buy a single impression through multiple paths, 5, 10, 20, sometimes 30 or more. We had this belief for a long time that we should only access a few select paths, we've learned that buyers wanna access via all, thus opening up our supply aperture. How do we do this efficiently? Well, when we bought nToggle a few years ago, as Adam Soroca outlined earlier, it was about creating a tighter funnel of bid requests for buyers, making it more concentrated, more value, increasing their win rates. Well, we're gonna flip that to the other side as well. Advanced filtering now allows us to open up our funnel of supply without always having to buy more machines. This is critical. Well, speaking of machines, as J. Allen Dove outlined, Rubicon and SpotX were largely on-prem.
We would have loved to have leveraged the cloud for some parts of our business for certain algos, but it's different to capture those economies of scale doing both. With the acquisitions, Telaria and SpringServe were largely in the cloud, we now have this massive hybrid data center allowing us to leverage the best of both worlds, and in the end, allowing Magnite to win more auctions away from our competitors. Finally, SPO. This is getting exciting. Combining the best supply across CTV and DV+, the best audiences across all formats, and with the best tech underpinning that, chiefly Prebid and Demand Manager, we have now created a flywheel for buyers who are consolidating spend on Magnite as we speak. Let me dig a little bit deeper here.
As I showed you, we plan to win with both reserve auction and open auction on DV+. For both, it's important to win by also becoming the favorite choice of brands, holdcos, and DSPs. We are well on our way to becoming this preferred partner. Over the last year, we've been asking holdcos, "What will make you switch your business from walled gardens?" Which they know are already a massive threat to their business, and instead to the open internet. It's actually more about how do we accelerate this. Their answer, help them create more value for their clients. We built our offering, our buyer offerings to deliver on that, allowing agencies to curate their own supply partnerships, go much deeper with supply partners. SSPs have become critical to facilitate that.
We're helping buyers create custom bidding algorithms, custom controls, onboarding multiple data partners, so they can fine-tune the secret sauce they provide to their clients. That's really the goal, build technology differentiation for the buy side versus their peer set, so they're not always competing on price. Finally, when you bring all this together, Magnite helps manage buyers' reach and frequency across all audiences. That's really hard to do when you're fragmenting your spend across multiple platforms or inside of black boxes. In other words, we're building an offering to give holdcos the ability to spend on their own exchanges outside of the walled gardens. In effect, the easy button for buyers. In the last few months, we've seen this offering gain a lot of traction. We're seeing significant momentum of buyers across the board.
As Sean outlined earlier, everything from brands to holdcos to DSPs providing more value for them. We're both really excited to see these preferred deals growing in the coming weeks across both CTV and DV+. All right, wrapping up, last slide here. Here's what I wanna leave you with. While it's great to have the largest omni-channel SSP, the technology to connect all of these valuable cross-screen audiences together, the part I wanna underscore is that we are obsessed with building the easy button for the open internet. This will drive more auction revenue, sorry, more reserve auction revenue, more open auction revenue, and more SPO consolidation overall. With that, I'll hand it over to my colleague, Mr. David Day, who will talk about our financial machine.
Great. Thanks, Joe. My name is David Day, and I'm Magnite's CFO. I've been involved in digital advertising for 20 years, and I've been with Magnite for eight years. Today, I'll first speak to our addressable market. Second, I'll highlight revenue, including revenue growth for the overall business and for CTV. We'll also touch on channel concentration and our revenue growth baseline. Third, I'll review profitability, margin, and efficiency trends and our overall financial leverage. Fourth, I'll discuss our capital structure and in particular, our debt leverage. Prior to 2020, we had access to a roughly $50 billion addressable market through our focus on online video and display and audio, which ran across mobile and desktop channels. Over the last 18 months, we merged with Telaria on April first, 2020, which created our initial entry into CTV.
We acquired SpotX on April 30, 2021, which increased our CTV and online video scale, expanded our CTV features and functionality, and expanded our unique demand through SpotX's demand facilitation team, which Sean discussed. Last but not least, we acquired SpringServe on July 1, 2021, which added a purpose-built programmatic CTV ad server and further expands our CTV tool set. As a result, Magnite has an expanded potential long-term addressable market of over $150 billion in just the U.S. alone. This market includes linear TV ad dollars, which will increasingly shift to CTV and which, as we've discussed today, will increasingly be transacted through programmatic channels. In 2018, we generated revenue of just $125 million after having weathered a challenging disruption in the several years prior to that time.
Since then, through organic growth and through M&A activity, we've gained significant scale, and we expect to exceed $400 million in Revenue ex-TAC in 2021, an increase of over 250% in three years. In 2022, we expect to be well over half a billion dollars in Revenue ex-TAC. As a result, we've reached a scale that allows us to develop product, leverage synergies, and provide solutions for our clients and compete at a level that we've never had before. Prior to our recent M&A activity, we had no CTV revenue. This slide shows our quarterly CTV Revenue ex-TAC growth over the last six quarters, growing from $9 million in Q2 of 2020, subsequent to our Telaria merger, to over $40 million expected in Q3 of 2021.
Based on the midpoint of our guidance for Q3 2021, CTV would represent over 35% of our revenue ex TAC for the quarter and is growing faster than the rest of the business. Moving to our revenue ex TAC channel mix in our actual results. In Q2 2021, CTV represented 34%, mobile 39%, and desktop 27%. The three-year expected industry CAGR for these channels is 44% for CTV, 20% for mobile, and flat for desktop. If you run those industry growth rates against our current mix on a weighted average basis, that would imply an overall revenue ex TAC growth for Magnite of 23% annually, and that is prior to any share gain in any channel. We are very well positioned for sustained revenue ex TAC annual growth of over 25%.
Our international revenue represents roughly a fifth of our business, and we see opportunity for growth in both CTV and DV+ internationally. Our revenue growth is driven by our strong customer relationships. On the buy side, we serve over 150,000 marketers running through over more than 150 DSPs. On the publisher side, we have over 2,100 clients with effectively zero attrition, and our net dollar retention over the last 12 months was 134%. I'll now discuss some key operating efficiency drivers in the business. One of the key efficiency measures that we focus on is our infrastructure cost as a % of revenue ex-TAC. In our infrastructure cost, we include CapEx, data center, bandwidth, cloud, and other related costs.
In 2019, those costs represented 22% of Revenue ex-TAC, decreasing to 21% in 2020, and expected at 19% in 2021 based on consensus Revenue ex-TAC. This progress has resulted from organic improvements, as J. Allen Dove discussed earlier, from the profiles of our acquired companies and from our growing scale and the related synergies that we are achieving. As J. Allen Dove mentioned, we are on track to achieve our targeted platform synergy savings. We expect that this percentage will continue to decrease over time. Our gross margin has grown from 63% in 2019 to 68% through the first half of 2021. Our gross margin or these gross margin numbers are based on our GAAP costs, including depreciation and amortization.
If we were to strip out non-cash M&A-related intangible amortization and other non-cash depreciation and amortization, our adjusted gross margin would be roughly 80%. We also expect continued improvement in our gross margin over time. We have significant leverage in our business model. While I suppose that all costs are variable at some level, we estimate that only about 15% of our costs are what we would consider directly variable. Those costs consists of our technology and data center and related costs. Our remaining costs are primarily fixed or what we'd consider step function fixed, with over 45% of our costs being personnel-related, including facilities costs, and 35% of our costs resulting from depreciation and amortization, either data center machine depreciation or intangible asset amortization resulting from our acquisitions, and from our internally developed software efforts.
As a result of this dynamic, now that we have reached some level of scale, we expect to achieve 45% flow-through of revenue to adjusted EBITDA in 2021. That's based on current consensus, we expect future flow-through of greater than 50%. In addition to this operational cash flow, we also have the ability to generate additional cash flow from our working capital dynamics. We have a beneficial cash conversion cycle, meaning that we collect receivables in advance of payment to our customers. Our DSO is currently on average 20 days shorter than our DPO. As a result, as our business grows, we actually add cash to the balance sheet. As a result of our acquisitions and the leverage in our financial model, we've been able to accelerate growth in our adjusted EBITDA and our adjusted EBITDA margins.
Our adjusted EBITDA loss of $11 million in 2018 has rapidly turned around and grown to expected adjusted EBITDA of $130 million based on consensus for 2021. This represents an adjusted EBITDA margin of 32% for the full-year. Based on our increasing confidence in our revenue growth opportunity, our cost synergy achievements, and our continued scale efficiency, today we're raising our adjusted EBITDA margin target to 35%-40%, up from our prior target of 30%-35%. With the improving flow-through dynamics we've discussed, we are now also starting to generate more significant positive operating cash flow. Note that we define operating cash flow as our adjusted EBITDA less CapEx.
In 2018, we had negative operating cash flow of $31 million and have steadily grown with a rapid expansion in 2021 to an expected operational cash flow of almost $100 million based on consensus revenue ex-TAC. As it relates to capital allocation, our primary use of excess cash generation in the near and medium term will be to reduce our debt leverage ratio and to reduce interest costs. We've already made rapid progress on this front, and we expect our net leverage ratio to be less than 4x by the end of this year, which is a step function reduction from the ratio we had at the time of the SpotX closing.
We believe that our improving leverage ratio will allow us to reduce the interest rate that we pay on our Term Loan B, and we will look at appropriate strategies to pay down portions of that debt at the right time. As we move through 2022, we believe that the ratio will continue to reduce. We're ultimately targeting a 2 - 3x net leverage ratio over time. I'd also like to highlight that from an M&A perspective, we believe that we have all the assets that we need, and we do not foresee any significant M&A activity in the future. We could, of course, opportunistically make acquisitions to accelerate our product set, but if we did so, we'd expect those to be small in size. In conclusion, we feel that we've built a financial powerhouse with significant opportunity for growth and for value creation.
We've exercised prudent financial and strategic discipline in our acquisitions, in our integrations, and throughout our business recovery over the last three years. We have assets in rapidly growing market segments that give us a solid foundation to achieve greater than 25% revenue growth over time, and we couldn't be more excited about the CTV opportunity that lies ahead of us. We have expanding margins now targeting 35%-40% adjusted EBITDA margins that will generate significant and growing cash flow over time. We're improving our debt position and building a strong balance sheet for the future, and we love where we sit as an independent and scaled omni-channel SSP with the opportunity to work with and provide solutions to all the marketers and publishers that matter in the world. With that, I will turn it over to Nick for Q&A.
Excellent. Thank you all, and we'll begin the Q&A session. If you have not submitted questions already, please do so in your viewer. You can submit those, and then I will be reading them to our team to have those answered. To begin with, our first question comes from Laura Martin. Question will be directed Michael, starting with you. I think Sean covered some of this in his presentation. The question is, can you talk about take rates on CTV when exclusive?
The market thinks you guys are getting 3% take rates down from 8% at Telaria versus 14% at old Rubicon Michael, if you can address maybe it looks like there's some CTV in there and differences within kind of the rate card that Sean addressed to start with.
Yeah. I think directionally, you know, the step-down is accurate, in the sense that, you know, we don't talk about take rates and disclose them. You know, three is probably on the low side. The simple fact is, when a publisher, and I imagine when you say exclusive, you're pretty much talking about the bigger boys that have, you know, linear sales teams. These are direct sold deals by the publisher. Because of the amount of work that we do in that scenario, it comes with a lower take rate. It's a really, still a very strong net revenue story because this is the most valuable inventory.
Even at a decreased take rate, it's an incredibly attractive segment, and that's one of the things that we look forward to fueling our growth in the years to come, is winning that business now that we have the technology solution in place to be able to do that. We'll trade off those take rates any day for the value of that inventory and the net revenue story that comes from it.
Excellent, Michael. Our next question comes from Matthew Swanson from RBC. You mentioned the survey results showing that consumers would rather watch ads than pay more. What would you think of long-term mix of SVOD versus AVOD?
You kind of broke up there, Nick.
I'll just change that.
I got it. Yeah. Yeah, I'll grab this, and then I'll kick it to Katie. You know, I think what we've seen in the market and what everyone thinks in the kind of ecosystem is that the Hulu model is one, right? You have a tier that's ad-free, that's expensive. You have a mid-tier that is less expensive and less ads, and you have a lower priced tier that has more ads. You're starting to see that folks are voting for the mid-tier and the low tier, just given the exploding cost of all these streaming services. If they were gonna cut the cord to save money, if you start to add it up, it gets pretty darn expensive if all you do is subscribe to SVOD.
I think Disney and Netflix are the outliers, and they're not a model for anyone else to try to use. I think that, there's a lot of people that think Netflix will probably get into the hybrid model too, notwithstanding their stance that they'll never do ads. Yeah, I think the Hulu model has won the day. I don't know, Katie, if you have any more color on that.
Completely agree. Just from like a stat perspective, AVOD has almost completely caught up to SVOD, so it's 51% SVOD and 49% AVOD now. Almost every household that has connected devices has both, as Michael mentioned. We do believe that the model will go, that it will give consumers choice, to pay, you know, what they wanna pay and obviously have the ads support the rest of that.
Excellent. Next question, Michael, comes from Shweta Khajuria over at Evercore ISI. Michael, over the next five years, in your slides, you said you expect to process $15 billion- 20 billion in ad spend annually. Could you speak to what this means in terms of growth for the company? Number two, the 30% share that you talked about in CTV, just to clarify, does that mean Magnite expects to have 30% share of a $50 billion market five years from today?
Sure. So yeah. On the ad spend side, Shweta asked a great question, how does that jive with our revenue growth. You know, Sean walked you through all the product lines that we now have, so it's much more expanded than just a take rate on a auction. That's gonna impact ad spend dramatically because if you become the ad server, you're gonna see all the ad spend, even if it's direct sold. Part of that inflation or the hyper-growth of our ad spend, you know, and again, these are all aspirational five-year out, kind of forward-looking statements. Part of that is that we're in other lines of business. We have a different business line with Hulu.
We have seat fees, and we have ad serving fees. That's gonna directly impact the ad spend. It's not necessarily gonna correlate directly with the projected growth that David has talked about in the long-term aspirational growth of 25%+ in terms of net revenue. There's a little bit of a dislocation with the traditional ad spend. Times take rates gets you to your revenue, right? As far as the second piece of it, you know, again, it's aspirational. Yes, we feel as though in a marketplace where we are now the leading independent SSP and a, the top-tier publishers looking for that solution, we think we're in the early stages of winning that business away from FreeWheel, who has most of it, and we think achievable 30%.
We think that the industry feels as though $50 billion in spend in CTV programmatically five years from now isn't out of the question. I don't know, David, if you have anything to add to that.
No, I think you covered it. It all it does is provide upside from our, you know, 25% + annual revenue growth rates.
Excellent. Next question comes from Nicholas Zangler at Stephens Inc. You have placed large publishers within the reserve market bucket. If reserved, how are these publishers fully taking advantage of premium inventory that they possess? Wouldn't they want some form of bidding on their inventory so that the highest bidder goes for that inventory? This way, publishers are awarded the highest price. Long as the advertiser is pre-approved, are you saying these deals will remain primarily direct deal? Any thoughts here would be helpful.
Yeah, Nick, and I'll kick it over to Sean to help provide some color. First off, you know, there's other than just a auction where the highest priced advertiser wins, there's other, as you well know, benefits to doing programmatic if it's not like an open bidding scenario. You know, there's workflow, there's data enhancement. Generally speaking, we're talking about the largest broadcasters, as you pointed out, the largest media companies, and they have a history of 50+ years in TV of trading ultimate price for volume. That's what the upfronts are all about. We'll negotiate rate, but in turn we're negotiating it for a fixed volume where we feel more comfortable about knowing we have $10 million-$15 million, as opposed to putting all of our inventory up for bid.
We may yield higher per unit, but we may be giving up our volume guarantees, which we've relied upon forever. I don't know, Sean, if you have more color to add to that.
I think that's 100% right, and we see that theme frequently in CTV. We do also provide the flexibility to do a private market deal that is biddable. If a media owner wants to set it up and make a private marketplace deal biddable, they can do that. Michael, I think everything you said in terms of the dynamics of CTV is 100%.
Excellent. Next question comes from Matthew Thornton and Anthony from Truist. Why won't the largest CTV content providers, a la Disney, go vertical or take a walled garden approach? Would Disney be building, buying a DSP? Would that be neutral, positive, negative?
I think some of them will try to do that, not in a fixed way that we've seen walled gardens in the past, but certainly you're seeing that with kind of a Roku where they are keeping their device IDs in their data, and the only way you can get ahold of it is by going through their DSP, right? You can't take that out with you. That's kind of the walled garden approach. You gotta keep in mind that I just think at the highest level, the reason why walled gardens came into existence is because it was a monopoly on the type of inventory. Google had a monopoly on search, and ergo, you were able to create a walled garden where it really benefited Google. Facebook, a monopoly on social.
No one's gonna have a monopoly on premium video, and I assure you, buyers do not want more walled gardens. They wanna be able to find users across multiple services and overlay their data onto it as well. Buyers are not leaning into this, and I think that these instances of mini-walled gardens are gonna have a short shelf life because of the buyers being able to control the way a publisher sells because they just won't be able to force someone into their buying world the way they wanna do it because no one will have enough inventory to satisfy a buyer's need, and they can go elsewhere because there's more premium inventory somewhere else.
Excellent. We've got another question. I'm gonna try to combine two questions, one from, one from Jason Kreyer from Craig-Hallum, the other one from Georgia Ortega. It has to do with clean rooms. Is the data clean room solution already available in market? What do you, what do you view as the key Magnite differentiators? The other key element of this is, in that clean room, what is the publisher, you know, appetite to share their own first-party data, especially on the CTV side?
Adam, you wanna jump on that one?
Yeah. The clean room is just coming to market today, so it's a brand-new offering for us. In terms of the advantages that it provides, we see this. Clean room is probably an overstated term. It's, as I said in my remarks, it's more of a matching service. What we see in the market is that it's a very simple and elegant solution for matching lists that exist on two sides. A publisher who has registration data and an advertiser or a data partner, anyone who has a list on the other side, we can match those together without the data ever leaving the walls of those publishers and the marketers or data sources.
It's a much more secure and much more lightweight solution that really resonates with many of the publishers that are out there who don't want their data to leave. That's what we, what we really responded to, is publishers saying, "You know, I don't want to give my IDs to a third party. I don't want to ship my data out to another entity, out into a bunker or something like that." This is an alternative to those sorts of solutions.
Excellent. Next question comes from Tim Nollen from Macquarie. Question on SpringServe. How do you win business from FreeWheel? Do you have relationships already? What do you have to do to convince TV networks to win or switch?
Yeah. Why don't we kick that to Sean? He can walk you through that. It largely, we have relationships already in place through legacy Magnite and through SpringServe. It's not a sales cycle where we have to gain trust or awareness because they don't know who Magnite is. I'll let Sean opine.
Yeah. Thanks, Michael. You saw, I think on the customer base, it's a pretty broad market with different types of customers. To date, the SpringServe ad serving platform has been a great fit for a lot of these newer media companies, Some of them have hit really meaningful scale, where they're very focused on programmatic monetization and audience-based monetization. Forward-looking, I think the SpringServe platform has proven to be a really ideal fit for that cohort of customers. Of course, we work in conjunction with the larger media companies who may use FreeWheel as the primary ad server, but us for the programmatic plumbing.
We just had an announcement yesterday with Fubo, which there's this FreeWheel ad serving customer that's using us as the programmatic plumbing, and we're working in conjunction with them on a joint customer. Of course, I touched on the proxy layer, right? If that is a stepping stone solution to give customers those advanced programmatic capabilities they're looking for, getting more deeply integrated without having to change the ad server in the near- term. That builds familiarity with our platform, we'll see where things go down the road. Hopefully, that gives you an idea sort of how we view the market.
Yeah, I think that's excellent, Sean. I think it's really important to make this point clear. Our strategies are not a rip and replace. There's a lot of things that FreeWheel ad server does that SpringServe will never do. The idea is to be that brains of the programmatic to sit in front of the ad server. Where have we seen that work? Well, Joe walked you through Demand Manager, right? Demand Manager has now become the new Google Ad Manager. It sits in front of it, and it does all the programmatic. We think that there's a, you know, success that's been in the market with that strategy, and now we feel with the full stack with SpringServe as part of the Magnite family, puts us in a really good position.
Excellent. Next, I have a follow-up question from Laura Martin, from Needham. Does your investment in buy-side tools for CTV compete with The Trade Desk, or are you moving away from being a pure play SSP?
You know, I think one of the least understood or least talked about aspect of being an SSP is we've always had buy side assets. You have to make sure the marketplace works. We have always had a team that called on the DSPs and the agencies to make sure, A, the plumbing works, and B, we were in a position of preference. What we're seeing, the trend we're seeing is a monumental increase in attention from buyers to what supply they're accessing. Those buyers are coming to us. We're not exactly accelerating our outreach. The interest is, okay, I've made my DSP partner choice, and Trade Desk, they're agnostic to supply, right? They're a software company, and they allow the agencies and the buyers that use it to choose where they wanna source the supply.
They want to make sure that the supply is premium, well-lit, curated, high quality, priced right. It's all just doing the job of an SSP. Just like a publisher has buy-side salespeople, we need to do that on behalf of our publishers to make sure we get preference from the buyers. We are in no way, shape, or form trying to be competitive with The Trade Desk, cut them out of the picture. I mean, they just hosted me at Pebble Beach last week. I'd like to go back, Laura, please don't tell Jeff that we're trying to compete with him. That is not our intention at all. It's just something we've always done, and the buy-side is so much more interested in the supply-side than they have been traditionally.
Excellent. Sounds like an invite's coming back for next year.
Yeah.
Question from Shyam Patil from SIG. Reserve auctions, how common are these now, and how are you thinking about the industry transition towards reserve auctions?
As Sean showed in the CTV side and Joe talked about on DV+, you know, reserve auctions aren't a new thing because CTV came along. Yes, CTV is predominantly reserve, and we believe that the top tier of CTV will be reserve for the next foreseeable future. Reserve has always played a huge role in DV+. The highest value inventory has always been held back by big publishers like New York Times, Disney, et cetera, and put into reserve and negotiated directly with agencies and/or direct buyers. We think that, you're gonna see that play out in CTV where the top tier guys, the vast majority of it will be reserve.
There'll be digital natives, you know, FAST services that don't have any linear sales capabilities, where they're gonna need all the monetization they can get, and they'll be a lot less fussy about whether or not they sold the deal and they controlled the price. That's not going to be the majority of the market. We're well-positioned for the open piece of CTV and obviously DV+ are extraordinarily well-positioned for the reserve, which as I said before, still is a huge chunk of value in DV+ and will be the majority of the value in CTV.
Great. Thanks, Michael. We have a follow-up question from Nicholas Zangler from Stephens. It's really on Sean's section and comment that Magnite can set up bidding for CTV publisher inventory among advertisers. For those that are enabling bidding, are the CPMs much higher than a similar direct deal CPM? I'm wondering if materially higher CPMs might lure additional publishers to implement bidding on CTV inventory.
I think it's really situational. It depends on the publisher's preference and how they're engaging with buyers. For some media owners, they may like that type of transaction model where it's a private marketplace, but it's biddable, and maybe they see exactly the upside that's being presented here. In other cases, the media owners may be more focused in terms of like the comments Michael made, where there are other things more important than the lift in CPM that the biddable environment can provide. I think the answer is just really situational based on the media owner that we're working with.
Excellent. So next question from Shyam Patil. DV+ growth, how should we think about the growth of the DV+ business in medium, intermediate term?
David, do you wanna handle that one and then perhaps Joe could add some color to that?
Yeah, I'll take that. I think if we anchor on some of the channel growth rates that I talked about, if you look at DV+, mobile is more than half of our DV+ business, and we've got industry growth rates at, you know, 20% over the next handful of years. You've got desktop that's roughly breakeven. Just foundationally maintaining share, you've got growth over time, you know, in the teens, in the low to mid-teens. As Joe highlighted, I think we have significant opportunity to take share. We would expect those growth rates to, you know, be in the low to high teens over time for sure.
Excellent. Next question comes from Daniel Prather. What would the market environment look like if there was no longer a large monopoly at the center of ad tech? The monopoly isn't stated, but safe to assume the public starts with G.
Well, it would create a nice opening for Magnite to become the new monopoly. No, I can't say that. You know, it just would free up, listen, publishers are looking for trusted, scaled folks that don't compete with them, right? That alternative has never existed. As much as we liked at Legacy Rubicon to say we were that guy or Telaria or SpotX, there wasn't just enough scale or full stack capabilities. Now we have that full stack capability. It truly, it literally is the first time a publisher has a true alternative in market. That doesn't mean you flip the light switch and it happens overnight. I don't think we need the monopoly to go away, through some, you know, third-party intervention.
I think we're in a great position right now to take a lot of money away. I think that oftentimes people think of our competition as like OpenX or Xandr or PubMatic or, and all these guys are fiercely competitive, don't get me wrong, but it's always been Google and FreeWheel, always. That's where the vast majority of the dollars are held, and that's why we built this company to be able to go after those dollars in a way that doesn't sacrifice in terms of monetization and/or feature capabilities.
Excellent. Next question comes from a follow-up from Shweta Khajuria at Evercore ISI. Sean, it looks like your CTV rate card slide's getting quite a bit of likes out there from our audience. Referring to the CTV rate card, this is really for you and David. What percentage of Magnite's business is private marketplace publisher-managed, private programmatic Magnite-managed, and managed services? Could you give us a flavor or an indication?
Why don't, David, you take that first and help Sean as you, as you hand it over to him? I don't, not because you don't have the capability, Sean, it's just that we don't probably wanna go on record as segment percentages.
Yeah. What I can share, I'll start with the managed services, and so that's publicly available. In Q2, for example, 7% of our revenue, Revenue ex-TAC, resulted from our managed service business, and about 20% of our CTV business results from managed services. Yeah, as Michael mentioned for the, you know, we've talked in broader scopes.
CTV is, you know, gonna be some type of reserve, much smaller component of, or tiny component of open auction. In our DV+ business, a majority is certainly open auction, but a not insignificant minority certainly on the PMP or reserve side.
Sean, you can walk them through, like, when we bucketed our clients by bucket, like broadcasters, you know, virtual MVPDs, why don't you can walk through what the typical need of those publishers are as it relates to our demand facilitation.
Yeah. One of the important points that I tried to deliver in that slide is that this can really vary campaign by campaign, right? A publisher could have a certain type of working relationship with one buyer or one cohort of buyers, like for example, the large holding companies. When you look at totally different cohort of buyers like mid-market, they may not have their own personnel out talking to that component of the market, that would fall under a totally different bucket, like managed service. Certain of these types of customers skew toward one or two transaction models at times.
I think it's important to note that in many cases, our customers are leveraging all or many of those different transaction models across their inventory based on specific campaigns, specific relationships with buyers, and their, the way they choose to address that cohort of the buy side.
Excellent. Next one's a follow-up from Laura Martin, from Needham. Can you talk about SSP exclusivity trends inside of CTV, and how big of an advantage is that to Magnite? Is CTV exclusivity an important upside value driver for Magnite? If so, why?
Yeah, for all the reasons we've kind of touched on, we think that in CTV there'll be far fewer SSPs. It'll be a winner take most, because the top tier guys are gonna be doing a lot of reserve stuff. You know, the way reserve auction works is that, you put your deals in, you know, what we refer to as like a deal library. The big, big agencies like the WPPs and the, you know, Omnicoms of the world get familiar with where they can find Disney's inventory, where they can find, you know, ViacomCBS's inventory. There's switching costs in terms of moving it out of that deal library. It becomes pretty much you don't want a bunch of SSPs out there with a bunch of different deals.
Generally speaking, it's the winner take most. We feel as though that trend is not going to change, in that our upside opportunity is immense because the vast majority of those deals are with FreeWheel right now. Prior to the SpotX and SpringServe acquisition, I would say to, you know, for all intent and purpose, they were kind of out of our reach. Now with the combined, you know, expertise with the Telaria folks and the assets from SpotX and SpringServe, they're very much within our reach, as, you know, Sean pointed out with the release that we did yesterday. You know, we expect to win more of those, and we think the trend line of reserve and winner take most is a positive impact for Magnite.
Excellent. I've got a follow-up here from Nick Zangler from Stephens. Can you talk about advertisers' ability to target specific users on publisher platforms? Specifically, how is this capability enabled in a reserve auction?
Adam, you wanna handle that one?
Yeah, sure. This is exactly the intersection of the future of audience along with reserve auction. Publishers, particularly on CTV, wanna protect those individuals. They don't wanna release any identifiable information out to the buy side. What we see at an increasing rate is that the publishers will group their users together, and they'll put them into a PMP, which can be targeted in a reserved auction. That allows the buyer, in the case that I used, BMW, to target highly affluent in-market auto intenders against a deal ID, but protect those user relationships and not leak those to the buy side.
That really is the culmination of how those two things are coming together in the protection of consumer data, yet still providing the buyers with signal that they need, and have those be able to be transacted against the deal ID at high value and higher CPMs.
Adam, as I've got you, I've got a follow-up different question, but since your, your line is active, can you comment on the approval of a patent you received yesterday, specifically related to, I think, some of the traffic shaping and any impact or benefit?
Yeah. I'm guessing that our attorneys would not want us to comment on what the patent covers or doesn't cover. I think we would just say that we're thrilled that the patent office has recognized the technological innovation that we created around traffic shaping and we now have a patent that protects that intellectual property.
Excellent. Another quick one, either, most likely for Allen or Adam. Does the Magnite clean room compete with, this comes from Tom Barr, compete with or work with LiveRamp data Safe Haven offering?
Allen, do you want me to grab it?
Yeah, go ahead, Allen.
Yeah. We work with LiveRamp, we work with the different clean room solutions out there. Our solution, again, is more of a matching service. It's really for those publishers who don't wanna participate with third parties, whether it be someone who supplies an ID graph, or whether it's someone who has a bunker or a clean room solution. It's gonna live alongside those solutions in the marketplace and thrive in partnership with Magnite.
Excellent. We've got one last question before we wrap up, and Michael, you can give some closing remarks. It wouldn't be, you know, an investor day or an ad tech company presentation without talking about cookie deprecation. Does a potential shift to more premium publishers, with better first-party data help Magnite?
Yeah. I think we've talked about that, several calls now in succession, and we really do believe that over time there is quite a shift in value from buy side to sell side, and we would benefit from that. It's been a third-party cookie-dominated world where that has led to value creation for data companies and for DSPs who organize that audience and resell it to their buyers. Now if you have publishers creating audience segments that are quite valuable because they have the direct relationship with the consumer and its first-party data, and we're helping organize it, keep it safe, and you know federate it and unify it across thousands of publishers, there's value creation there, and we think we'll participate in those economics.
Excellent. I think that brings us to the end of our investor day. Michael, any closing thoughts?
No, just again, couldn't be more thankful for all the participation, the questions, the support. Elated that you got to meet the team. Hopefully you feel like I do, that they're some of the most talented people in the space, and they're really what makes Magnite different, and the folks that work for them are out of this world as well. A huge thank you to the executive team, and we look forward to participating in several non-deal roadshows coming up and, you know, obviously earnings in a couple of weeks. Thank you again for your time today, and thanks Nick for organizing.