Filed by AdTheorent, Inc. pursuant to
Rule 425 under the Securities Act of 1933
and deemed filed pursuant to Rule 14a-12
under the Securities Exchange Act of 1934
Subject Company: AdTheorent, Inc.
(File No. 333-259027)
This filing relates to the proposed merger involving MCAP Acquisition Corporation. and AdTheorent Holding Company, LLC (“AdTheorent”) pursuant to the terms of that certain Business Combination Agreement, dated as of July 27, 2021.
On November 18, 2021, AdTheorent CEO, Jim Lawson participated in a fireside chat with Jarrett Banks, Editor-at-Large of IPO Edge, and Alexandra Lane, Multimedia Editor at IPO Edge.
Jarrett Banks:
Hello, and welcome to another IPO Edge fireside chat. I'm your host, Jarrett Banks, editor-at-large, joined by my colleague, multimedia editor, Alexandra Lane. We've got an exciting event for you today. We've got the CEO of AdTheorent, and this company is merging with a SPAC called MCAP Acquisition Corp, and that's under the NASDAQ ticker MACQ.
So before we bring on Jim Lawson, the CEO of AdTheorent, let's take care of a little bit of housekeeping. One of the great things about these events is that you, the audience, get to ask our guests questions directly, and you can do that by submitting them via the Zoom portal or emailing editor@ipo-edge.com. And we will have a replay available of the event about an hour after we finish our live event here. And you can watch that at ipo-edge.com. Now, before we bring on our guest, let's watch a quick video to see exactly what AdTheorent is all about.
All right. Great. Let's welcome Jim Lawson, the CEO of AdTheorent. Jim, welcome to the program.
Jim Lawson:
Thank you, Jarett. It's great to be here. Appreciate it.
Jarrett Banks:
Now let's start from the top. Please give us a bird's eye view of what AdTheorent is all about and maybe some people who aren't familiar with the industry. Tell us what you're about and what kind of customers you work with.
Jim Lawson:
Absolutely. So AdTheorent is a programmatic media platform, but we have transformed what that means. Put simply, we use machine learning and data science to target digital ads to the right users at the right times. So we work with the most sophisticated data driven advertisers in the world, advertisers who know digital and expect their digital campaigns to drive business results for them. And I think it might be valuable to state in plain English how we fit into the programmatic digital ecosystem. Sometimes that can be a bit jargon heavy and complicated. And I think this can be quite simple actually.
If you were to access on your device, your mobile device, or any device, really any digital device content, a page would render on that device, and that screen would likely contain an opportunity to serve you an ad, which we call a digital impression. Digital publishers like The Wall Street journal, The New York Times, you name it, they want to sell that digital real estate. Advertisers want to know which of these impressions or opportunities to serve ads, which of these are most likely to yield engagement or conversions or interests by their customers. For example, credit card signups, online sales, a hotel or airline booking, you name it, whatever action is behind that ad campaign, they're trying to drive those actions and they want to buy the media that's going to do that.
So behind the scenes, there's a marketplace to buy and sell those ad opportunities. It's an auction system that takes place in microseconds. AdTheorent occupies the demand side platform silo in this ecosystem, and that we enable advertisers to purchase digital media, digital ad impressions in real time and one at a time. But we have transformed how that is done. And the way that we've done that is we use machine learning and data science as our core method of ad targeting and campaign optimization at the impression level. So this is contrasted with cookie-based or other ID-based or user profile based targeting, which really focuses more on individual identities and retargeting users and profiles across the internet.
Jarrett Banks:
Right now a lot of companies say they use machine learning and big data analysis, but we're never sure how much that is really true. Tell us how you use it and how that benefits advertisers.
Jim Lawson:
Yeah, that's a great question. I mean, many companies practice machine learning and AI mostly in their marketing rhetoric. But machine learning and data science have been core to AdTheorent's operating thesis since inception in 2012. In 2012, our business premise was conceived. And that was how do you leverage machine learning and data science to replace cookie-based targeting? Because at the time, mobile advertising was big. Mobile advertising was new. How do you target in mobile without cookies? And we realized that machine learning and data science was the answer. Quickly we realized that it worked very well and we pivoted across channels and became truly omnichannel. And that's what we've been building since 2012. But we are aware of no other programmatic media buying platform that uses machine learning and data science as the core method of ad targeting and campaign optimization. We call it predictive advertising. We believe it's transformative and better. And I can take a minute to kind of tell you how it works. It's a little bit technical, but I think I can keep it pretty high level.
So our platform ingests digital signals tied to digital ad opportunities, the impressions I mentioned a minute ago. In this auction system that I talked about, these data strings that we receive are called bid requests. And in those bid requests, there's a lot of data. We supplement the data and we have access to many, many data attributes, over 200. And we have to decide based on those bid requests with the data whether or not to bid on those impressions in microseconds. And the question is whether you bid and how much. Is this a valuable impression, or is this a wasteful impression?
Our platform in microseconds assigns predictive scores to every single impression that comes into our platform. And it's based on historic conversion activity. Conversions, meaning business actions taken by users when they respond to ads or when they engage with ads. So put very simply our machine learning system identifies the correlations between conversions that occur. For example, if someone buys a product online or they book a hotel or a flight, and then the data tied to those impressions. So if the KPI was an online credit card signup, what data attributes are most often present when there is a conversion? It could be a number things. It could be the creative type. It could be the size of the creative, the publisher, or the publisher category, keywords in the URL or on the page. If the user is on Wi-Fi or cellular, if it's cellular, what's the carrier. There's countless data points that we receive that you can use to identify correlations between successful outcomes and the data itself.
So our platform assigns in microseconds predictive scores to impressions where the data attributes correlate with success. And as a question of scale, we evaluate and score over a million impressions per second and more than 87 billion impressions per day. We bid on less than one tenth of 1% of these impressions. And that's really a statement to the precision and the efficiency as you really are looking for needles in a hay stack here. You're trying to find conversions among all the digital activity that's happening in the market.
Jarrett Banks:
That's super fascinating. I'm going to bring in my colleague Alexandra, and I'll be back a little bit later in the broadcast. Take it away, Alexandra.
Jim Lawson:
Thank you, Jarrett.
Alexandra Lane:
Thanks, Jarrett. And great to have you here, Jim. So many companies in this part of the industry say they're using machine learning and big data analysis, but we're not really sure to what extent that is true. So how does AdTheorent use machine learning to benefit advertisers?
Jim Lawson:
Well, at the end of the day, the predictive scoring that we utilize is designed to optimize towards impressions that are going to convert. When we talk with our customers, our very first question is what is your objective? What are you trying to achieve with your media dollars? What business outcomes in the real world are you trying to drive? And our algorithms are custom for every campaign based on that answer. So as data comes into our platform, our platform learns in real time as to what data attributes in future impressions are going to be most desirable. And then when we see those, we assign these predictive scores and we seek out those impressions and buy those impressions and prioritize those impressions and pay more money for those impressions. And we avoid the waste. We avoid the impressions that are least likely to drive the bottom line or the business outcome that our customers trying to drive.
So it's a real deployment operationalized within our bidders that drives the decisioning in terms of where we spend our money on media. The contrast to that or the kind of the prevalent status quo in the industry has always been ID-based targeting, which is very, very different. ID-based targeting is basically you have IDs, whether they're cookie IDs, because somebody visited a website or they're in an audience and you've licensed a list of IDs, and then you're just finding those IDs. But what we're doing is really trying to bring the targeting game to a more precise and accountable level and have it be less one to one.
Alexandra Lane:
Gotcha. Yeah, that just helps kind of round out I think my perspective on this whole notion of using all these data points to better target customers. So there's this term that is used called the open internet versus the walled garden. So AdTheorent, you talk about the huge opportunities for advertisers on the open internet, so to speak. So how do you compare that to advertising within the so-called walled gardens? Maybe just round out a bit of the perspective on that.
Jim Lawson:
Absolutely. That's a great question. So from our perspective, the walled gardens advertise to a captive audience based on the endless amount of profile data that they've generated from their users' activities within those platforms. Facebook is probably the largest human profiling operation in world history. They know everything about you and they monetize that with advertising. Snap made news recently about its ad revenue falling because of its dependence on the Apple device ID, the IDFA for ad targeting. AdTheorent, the way that we use data, the way that we don't use data, the way that we're not dependent on specific ad IDs for targeting, we are data agnostic. We ingest a wide variety of statistics for targeting. We are not dependent on the cookie ID, the IDFA, the Android ID, or any other ID for targeting. And because of that, we have flexibility across the open internet.
We're not dependent upon the users within a walled garden and knowing how many kids they have and where they live and those types of personal things, and then generating profiles and advertising to them. In our world, we don't need to know that. We don't care about that. We just want to know when there's a conversion activity online in digital, what are the statistics related to these 200 or more attributes that are not individualized? And how do we use those in machine learning models to target ads? It's a privacy-forward and performance-first way to do it.
Alexandra Lane:
Awesome. And could you just maybe a simple definition of kind of what a conversion activity is that you had just mentioned?
Jim Lawson:
So, a conversion activity used to be things like selling clicks and selling views of ads, which is really a vanity metric that only benefits ad tech vendors. What we have attempted to do is redefine what that means. And again, our first conversation with our customer is you're going to spend money on media -- what do you want to get from that? What is your goal? What business outcome are you trying to drive? If we're working with a hotel, they want to sell online reservations. If we're working with an airline, they want to sell bookings. Credit cards want to sell credit card signups, you name it. Whatever that business outcome is, we create a custom algorithm and a custom model that we deploy into our platform. So as those conversion activities result from ad engagement, we know what were those commonalities, what data attributes were present when those ads converted into conversions. And from that, you can learn about how you want to buy media going forward. You become a much more intelligent purchaser of media when you have that level of data.
Alexandra Lane:
So how does AdTheorent differentiate itself from others in this space, especially the larger players?
Jim Lawson:
Well, I think many of the larger players are focused more on bulk media purchasing. They have built very efficient engines that allow advertisers to very quickly and efficiently spend money on media. And they can do that in a relatively easy and elegant workflow. What we're trying to do is address a gap in the market and that gap in the market is a drastic gap in the market. And that gap in the market is savvy, smart, digital advertisers that want to know what they're getting for their digital investment. We believe that we are filling a gap here by providing a privacy-forward, performance-first method to actually drive ROI, return on ad spend. That is how we evaluate ourselves. The performance component, the digital advertising market is like $90 billion in growth at 18 or 20% every year.
The performance component of that is a very, very big 15 billion or so component of that market. And not a lot of companies in our space have the wherewithal or the capability to actually drive performance. And that's what we were built to do. We believe that we have a seat at the table among the bigger companies because we've built something that's different. We've built something that addresses a very clear need right now in programmatic advertising. It's not just about bulk consumption of media. What are you getting for your bulk consumption of media? You need to be targeting your ads to the right people, and you need to know you're getting value for it. And that's what we do.
Alexandra Lane:
Yes. And you've said the total digital media spend is forecasted to be 171 billion this year. So why is AdTheorent going public now? And why did you choose a spec deal over a more traditional IPO route?
Jim Lawson:
eah, it's a great question. I mean, one of the reasons why we wanted to go public now is because we realize that a number of factors, both from our business and just from the broader market and the macro conditions are really much in our favor. The market is growing at an incredible clip. The US programmatic digital media spending will probably exceed 90 billion in 2021, growing at 18%, 13 billion of that is hyper-focused on performance-driving executions, which we are uniquely able to drive.
We believe we have an advantage there. We believe that we've never seen more demand for what we're doing. We're being told that there's an opening for performance first privacy solutions to drive digital ads and the louder the microphone we have, the bigger stage, the more lights we have, we can tell our story more effectively. We had a lot of strategic options for this next chapter. We have been partnered with Monroe who is obviously behind the MCAP merger, and we've been partnered with them for years as our lender since 2016. They have a track record of success with other SPACs. They share our vision for AdTheorent and the unique opportunity that we have. And we wanted to go into this next chapter with partners that we knew and frankly that we had confidence in, in getting us to that public market.
Alexandra Lane:
Yeah, that's great. It's, correct me if I'm wrong, it's Monroe's third SPAC deal that they'll be doing. And that's amazing that you've had this longstanding relationship. Can you talk a bit about your 2021 growth numbers?
Jim Lawson:
Yeah. In 2021, we've had a great year. We expect to generate about 161 million in revenue. That's our updated guidance up from 157.7 and 106.2 in adjusted gross profit and north of 30 and a half in EBITDA, which represents for both metrics 30% year over year growth. Today actually we published our third quarter results. We're very proud of that performance. And we believe that our unique track record of financial performance and demand for what we're doing is accelerating rapidly. And that's why we're able to produce consistent, positive financial results.
Alexandra Lane:
And then how will being publicly traded help give you that competitive edge and how will it keep helping your growth strategy?
Jim Lawson:
Yeah, little things like the access to incredibly smart and gifted industry experts joining our board, the ability to meet all the people that we've met during this process, the bankers, the analysts. Being able to educate the best of the best in our industry and get to know these other businesses and these other people and have them get to know us and understand our thesis and understand what we're doing and tell our story in front of a larger audience is critical. We think that we have an opportunity to expand into new verticals. We have an opportunity to grow our brand direct business because of the end-to-end level of service that we provide. CTV is such a huge opportunity. Our video revenue grew 67% in the third quarter alone year over year.
We think that there are so many different things that we can do with the additional capital that we would have as a public company. There are a number of examples I could go through, but we just think that now is the time for us. We've always been private equity owned. We've always been very much focused on driving profitability and that's not going to change, but we think that now is the time to go capture more of that market share, especially because the privacy winds are at our back given the way that we operate and the way that we target.
Alexandra Lane:
Yeah. And then you also, you've also mentioned M&A as part of your strategy as well. And do you think that this really puts you in a position to be market leader in this space?
Jim Lawson:
A hundred percent. Our view is that we don't need to be the largest DSP in the market for us to be extremely successful. I mean, there are some very, very good, long tenured public company programmatic advertising businesses that are performing quite well in the public markets. Our goal is not to disrupt or unseat them. We believe that we bring something new to the table. We believe that performance first, privacy forward, machine learning driven, ad targeting and optimizations are very much in demand. We believe that the regulatory landscape, the privacy landscape, the sentiment around user profiling for advertising all work in our favor. We have a way to drive results without going backwards on privacy. We think all of those things taken together give us an incredible advantage.
And we're excited to talk about all the different ways that we can leverage those tailwinds and really continue our growth story. We've grown for 10 years. We've been profitable. We've had an incredible track record of profitability over that time. We've been a rule of 50 company with our EBITDA margins and revenue growth. So we're really excited about what we've done. You mentioned SPACs. We're not like the typical SPAC. I mean, we have a financial performance that sets us apart.
Many SPACs are going to market basically with just forward looking, one day we're going to do this great stuff. For 10 years, we've been operationalizing a business and we've been driving and growing consistently and generating incredible profits. And we believe that sets us apart. So every company should be reviewed based on its business metrics and not on the means by which it goes public. We're going to be a public company in short order, and we believe that we're going to perform very well, because frankly, we're going to keep our heads down. We're going to continue to innovate. We're going to make new things. We're going to make things better. Our data science team and our product team and our tech team, our roadmap is longer than we have people to execute on it right now. We're very much looking forward to hiring more people into those groups and preparing for success. And we've already started doing that, frankly.
Alexandra Lane:
Yeah. That's awesome. And you kind of mentioned the tailwinds, I think, correct me if I'm wrong, but the privacy concerns and having just increased consumer interest in privacy, so how will you be able to keep growing in light of potential regulation and parameters put on consumer privacy?
Jim Lawson:
I mean, from our perspective, it's all a benefit to us. I mean, we're not against cookies. We're not against other individual IDs. We're not against IDFAs. We're not against using individual identifiers for advertising. We think that those things can be used responsibly. But we're not dependent on any of that. Because of the fact that we ingest data the way that we do, we ingest a very wide array of data, and then we optimize based on conversions and we connect the disparate dots from our ML models and our optimizers automatically bid on the right impressions and drive efficiency and drive conversions. We think that at the end of the day, that's the future. So the changes, the fact that there is a backlash or there's an awakening, if you will, that profiling users based on their behaviors online and that creating pools of user IDs and targeting those user IDs in some cases, outrageous things like health information.
There have been some media reports about some of these walled garden organizations that have leveraged health information and very sensitive information about people's private lives for advertising. That's not the business we're in. We believe that we are the solution to that. And there might be a perception that most digital activity is taking place within the social walled gardens. But there are actually a lot of good reports, I think e-Marketer put out a report that most internet activity occurs outside of the social media four walls. And we want to engage on behalf of our advertisers. We want engage with users in the open internet and we want to help them activate on ads that matter to them and keep the internet free and keep the content that users come in contact with in ads, elegant and beautiful.
We have an incredible in-house creative team that -- you got to deliver the science, you got to reach the right people, but when you reach them, it's got to make sense. It's got to work. It's got to be beautiful and engaging. And we're working with the top digital advertisers in the world and they come to us for one-stop service. We can do it from beginning to end. You don't need to be a data science expert. You don't need to be a digital marketing expert to kind of navigate through a lot of the challenges that customers face in this area. We're experts on this and we make it easy for our customers.
Alexandra Lane:
Yeah, that's a great way of explaining it. And I think that that explanation was really great because I think when a lot of consumers think of data and analytics, they think they kind of get this whole big brother notion that they're just being increasingly spied upon and watched. But I really appreciate that explanation that no, AdTheorent is really a solution to these issues and it's not dealing in this notion of just consumers just being watched all the time. Let's talk a little bit about the SPAC market a little bit. It experienced a bit of a reckoning in the first half of this year. Did that worry your decision at all to join forces with Monroe?
Jim Lawson:
Not really. I mean, you can kind of get distracted by other companies that are in completely different scenarios or different financial realities than we are. I mean, there are companies that are going public through SPACs that have no revenue, or they have no profitability, or they have no product. They have a thesis. We've been doing this for 10 years. We have a track record of making commitments to our stakeholders, exceeding those commitments, building products that work, satisfying customers, growing a customer base that's loyal, keeping a team in place that's loyal because we treat them well and value our team - we're a team organization. And we've done the hard part. There are many companies that are going public through SPACs that have yet to do the hard part, which is proving that you have a product, that it matters, that customers care about it, that it will resonate and last the test of time.
We're scaling our business right now. We're not testing out whether we have a business right now. So if we're going to get lumped into the SPAC market, I think that's shortsighted from the perspective of the investor. We bring a real business, real profitability. I can't wait to talk more about our Q3 results. I can't wait to talk more about our other results that are coming in the future because we have a real business that's generating real financial outcomes, and that should drive the decision making from the investor, not some nebulous statistics about SPACs that have nothing to do with AdTheorent.
Alexandra Lane:
And when will investors be able to learn about your Q3 results?
Jim Lawson:
We today actually published our updated S4. So I'm actually really happy to be able to talk about some of those results. In the third quarter, we increased our revenue year over year by 36% from 29 million in the third quarter of ‘20 to 39 and a half million in the third quarter of 2021. Our adjusted gross profit increased 36% as well year over year to 25 million, up from Q3 2020 by 6.7 million. Our net income for the quarter increased 87% to 3 million, up from 1.6 million in the third quarter of 2020. Our adjusted EBITDA increased 49% year over year to 8.9 million. And our adjusted EBITDA margins are now 35% for the third quarter, which is an increase from 32% in the third quarter of 2020. We're very happy with 2021. Our revenue and our EBITDA beat our expectations in the first three quarters and we feel really good about where we are, and we see a lot of meaningful opportunity ahead.
Alexandra Lane:
Yes, and we've already got some really great questions coming in from the audience. We just encourage them to keep coming. You had mentioned Facebook earlier. Just wanted to get your take on Facebook's conversion into Meta.
Jim Lawson:
Yeah. That's an interesting one. I mean at the end of the day, they can call themselves whatever they want. They're still Facebook. It is, again, a human profiling operation where they learn a lot about you in exchange for which they send you ads. That's not the business that we are in. I think that there are a lot of advertisers that are looking for ways to advertise their products and services that don't involve that level of engagement with one-to-one and potentially privacy-backward type executions. And we think that as more and more information becomes available about how those platforms operate, I think that there is more and more demand, frankly, from the customers. We talk to them every single day. They're very interested in driving performance.
Many of them previously believed that the only way to drive performance was through cookie-based retargeting or advertising within the walled gardens where you have all this personal information about users. What we're bringing to the table is disruptive in the sense that our thesis was, and we believe that we have proven, you don't need to know all of the personal information about the people that you send ads to. You can have statistics, you can have a broad array of inputs, that suggest browser settings, location, if you're in a Starbucks, if you are in San Diego in a Starbucks, if you're in a public library in Denver while on a Nokia phone, and the ad size is this.
All of these aggregate data points can drive very, very, very effective and high-performing digital outcomes from an advertiser's perspective. I can sell more hotel bookings without knowing how many kids you have, or how many cars you have, or who you are, where you live exactly, and what job you have, and what you say in your emails to your friends. It's just a different method. And I think that they can call themselves whatever they want. I think the real world is a great place to live. I don't think we need a digital version of that. But if you are trying to engage with users on these devices that we all use all the time, I think it's a great way to do it. The open internet is subsidized by advertising, and we think that we offer a very, very effective and privacy-forward way to do that.
Alexandra Lane:
Yeah. Excellent, excellent description of that. I think I share a lot of those same sentiments. I'm going to throw it back to my colleague, Jarett, who will kick off the next portion with our audience questions.
Jarrett Banks:
Thanks Alexandria.
Alexandra Lane:
Thanks so much Jim.
Jarrett Banks:
Jim, I didn't know people still use Nokia phones, but ...
Jim Lawson:
I don't know where that came from actually. That was just somewhere in the back of my mind.
Jarrett Banks:
Blast from the past. Well, speaking of which, we've got a question here, "Stepping back to old-fashioned advertising models, is there still going to be room for static banner ads that mirror what happens in print, or is targeting going to completely take over?"
Jim Lawson:
Well, the ad type doesn't necessarily mean there's no target. I mean you can have a static banner ad, you can have a standard display ad, but you can target it very effectively. So I mean we have both very elegant, we have rich media, video, all forms of media that we can advertise in. Display, we can have the most engaging game units. Our creative team, Studio A\T, makes some of the most elegant digital ads that you'll see. So, there's really two pieces to it. There's the generation of elegant ad units, and we can help our customers do that from beginning to end. But then it's where do you deliver those? I think you can have both. You can have a standard display ad that you send to only people who care about it in the way that I mentioned. I think that the evolution of ad tech is going in that direction. Then there's also a way to customize and make more dynamic the elements within the ad unit, and we can use machine learning, we call it advanced predictive creative. We can use machine learning to decide if you receive an ad for, for example, a car, which car? And if you get an ad from a restaurant, are you going to get the ad focusing on the beer and the bar, or are you going to get more of the family ad? There are ways that you can have dynamic customization in real time that takes the game to the next level as well. But I think the key is being able to be agnostic to the screens, being able to handle all media types, make them beautiful and engaging, and then use machine learning data science to get those ads in front of the right people.
Jarrett Banks:
Now, I know why I keep seeing those beer ads.
Jim Lawson:
It sounds like you might be being retargeted to.
Jarrett Banks:
We got another question here. “Could you explain targeting at impression level without using identifiers and cookies? In particular, Apple deprecated IDFA. What digital signals do you use to target audiences?”
Jim Lawson:
Right. The premise to that question is, without that IDFA how do you know where to send the ad? And as I mentioned earlier, when we get the bid request from a given publisher it has a data string of attributes associated with that opportunity. And so AdTheorent is deciding, as is every other DSP, AdTheorent is deciding, "Do we buy this ad impression for one of our advertisers? How much if we do?" So we have all these predictive scores from all of our campaigns. They're scoring these ad impressions.
So the decision that we're making is based on whether the algorithms give a predictive score that is high enough that suggests to us with a greater probability that by serving an ad to that user at that moment in time that a conversion, an online sale, a credit card signup, a pharmaceutical prescription questionnaire complete, whatever that action is, when you have enough actions that occur you can then identify commonalities of data that are present when those actions occur. It's kind of like when you think about genomics in the health space. Where if you have certain DNA or you have certain genes, they know enough now about this to say, if you have these seven genes combined with these other attributes, then you might be more likely to have asthma, or you might be likely to have some other condition.
It's more about enough data about outcomes that you can then begin to predict future outcomes, and that's what we do. So we don't need to know the IDFA of a user. We don't need to know a cookie ID. We don't need to know any of these individual IDs, because that's not how we target. We target based on predictive score, which is an output generated by analyzing a ton of data, and it's not just what is your user ID, oh you have three kids according to Facebook, and all these other things. That's not what we're doing.
Jarrett Banks:
Got it. No, that's really interesting. Speaking of interesting, here's an interesting question. "Wouldn't this platform be a valuable tool capability for Google to use with their advertisers, and if so, wouldn't that make AdTheorent an acquisition target of theirs?"
Jim Lawson:
Yeah, I mean that's a great question. I mean we believe that what we're doing is very valuable, and we think that there are a lot of companies that would benefit by having the types of capabilities that we have. We've been very focused on customer first, performance first, understanding how to drive outcomes for customers, and not just media buying. And I think that what we're doing is unique, and that's why there's so much demand for what we're doing. So I think there's a lot of interest in what we're doing right now, and I think it's because we're doing something different and we're doing something better.
We're approaching the problem from a different perspective. I mean today most ad targeting from DSPs is done one of two ways. You're either going to ingest cookie IDs, which is basically retargeting. So if you go to buy a pair of shoes or a tennis racket on a sports site, they're going to drop a cookie on your device, and because of that you're going to get tennis racket or sneaker ads for the next month. Or audience targeting. Audience targeting is DSPs license lists of IDs that are widely available in ad tech.
You can say, I want a list of stock market wizards, or car enthusiasts, or moms. There's countless examples like that. And what that is, beyond all the noise, they're really lists of IDs. And then what the DSPs do, or the demand-side platforms, what they do is, now I have a campaign for a customer and I'm selling that I'm going to reach car enthusiasts. I licensed a list of IDs. Now, all I'm doing when I get those bid requests is, I'm looking for those IDs that correspond with the data that I licensed. It's not very complicated. It's the easiest way to target. And what we're doing is just different, fundamentally and foundationally different.
Jarrett Banks:
Well, you said something that sort of triggered an idea in my mind which is, how are OTT applications kind of revolutionizing the way that you advertise and the way that you get data about customers?
Jim Lawson:
With respect to like connected TV and those types of ... That's a great question, and it's very exciting for us. The connected TV opportunity is relatively early. It is early for us. But we are investing a great deal in building out our connected TV capabilities so that we can do a number of interesting things, and that we can bring this same level of machine learning post-view conversion optimization to CTV. I'm excited in the future to talk about some of the things that we're doing in CTV. We have an incredible Partnerships and Product team that's very, very focused on CTV.
Our third quarter CTV, or video generally actually, growth, a lot of which was CTV, was like 67%. We've published previously our CTV growth, which we have very much I think in the year we're going to triple what we did last year. And again, we haven't made historically a huge investment. But we've started to make a big investment in CTV because the opportunity is huge, and we think that we are uniquely situated to capitalize on a lot of the different data that's available in the CTV programmatic world that we can use for machine learning models that's not currently being used by any other DSP.
Jarrett Banks:
Got it.
Alexandra Lane:
We've got a question that's saying he's guessing you must be able to use some latent factorization models coupled with causality. I'm not quite sure what that means really.
Jim Lawson:
Well, I mean we use a number of ... Our special sauce is not one algorithm. We have countless algorithms, our Data Science team. What's special about us is the flexibility to use a number of different algorithms and a number of different data science models and inputs, and then deploy them into our platform for automated execution. And the models speak to each other. For example, we have competing campaigns. And maybe one campaign needs, because of scale, maybe it needs a conversion more than another. So the models communicate with each other, and the models learn over time. The more data that we have, they learn over time. In the beginning of the campaign, every campaign, there's a learning period. Before you have enough conversions, you can rely upon some of the more traditional targeting methods. For example, contextual or geo targeting.
Then you get enough conversions that the ML models learn. Then from there, they might initially learn from a click-through rate. So in the beginning of a campaign, where are the clicks coming from? Oh, most of the clicks are coming from these attributes. Heavy-up on delivery there. And then, again we're not in the business of selling clicks, eventually clicks yield conversions, hotel bookings, you name it. Then as we have more data there, we stop caring about clicks and we start optimizing more towards the actual ultimate KPI, which in that case is the hotel booking. I hope that answers the question. I wasn't entirely clear on exactly where that was going. It's probably better suited for our data science professionals. But yeah, I hope that answered the question.
Alexandra Lane:
Yeah. Then there's a question that came in after that saying that, you mentioned tons of data that you use for your algorithm several times. Can you give some examples what those data attributes are?
Jim Lawson:
Oh yeah. I mean, so the data that we're referring to is not data about a person, not data about a user, it's data about an ad impression opportunity. In other words, I go down this thing, I go to the New York Times or Wall Street Journal. I go to that page, the DSP / AdTheorent would receive a string of data. They would know the publishers provide this data to the DSPs. It would include things like the size of the ad unit or the slot, the ad unit slot, keywords on the page, keywords in the URL. What type of device am I on? Where am I? Am I in a city? Am I at home or am I at work? We have a point of interest capability, which allows us to provide geo contextual information on top of raw geo data. So we could know that you are in a university library. We could know that... and not you, I misspoke. We can know that that impression at that moment in time is an Android that is five years old, that has a X inch screen, that is on a publisher X publisher with Y IAB category, on a Verizon connection and any other number of data points. We also enrich the bid request with a number of demographic signals. As I mentioned, geo contextual signals. And so when you take all that data in, it's hundreds of data points, and that's what the model has to work with in creating its predictive scores. So we're not scoring users, we're scoring individual impressions at a moment in time. And then the next impression comes and it's like, what are the data attributes associated with each impression?
Jarrett Banks:
That's great. Speaking of mobile, we have a question here. What about the impact of increased time on mobile? It wasn't long ago, even Facebook had to work hard to monetize mobile. Some people don't even seem to have laptops or desktops anymore, they basically live their lives on mobile phones.
Jim Lawson:
Oh, we love that. I mean, that's where we came from. We were founded on the mobile opportunity. I mean mobile is a fantastic place to reach people in different ways, different geographic places. It's a great way to engage users when they're interested in different content. You can use that geo contextual information as part of models.
So, I mean, mobile, again, you don't have cookies in mobile. So what we do is very valuable in that we can use these signals to target in mobile. And I think the more time people spend on mobile, it's not surprising. We saw this in 2012 and again, that's kind of why AdTheorent came together. That was the genesis for AdTheorent being formed. But then we realized very quickly that machine learning driven ad targeting and optimizations was very omnichannel. It works on desktop. It works across mobile. It works across CTV. It works on all different ad types. Rich media, display, other types of video. Everybody's talking about CTV, but video in general is a very, very strong and growing channel for AdTheorent. And we have a number of ways to use video to drive, not just awareness but performance.
Jarrett Banks:
All right. Another question here. “Do you work with AppNexus?”
Jim Lawson:
Well, I'm not going to get into which exchanges and which publishers we work with and don't work with, but I will say we work with all of the big ones. So we work with the top supply. We have access to the top SSPs. There is no supply that we don't feel confident that we have a scale and access to. So there is an incredible amount of SSP partners that we are integrated with and other inventory that we're integrated with.
And I'll just say that we're very, very prudent about how we manage our QPS or our queries per second, that come into our system. We have a great partnerships and inventory team that makes sure that we don't have redundant inventory and redundant opportunities where there's waste or we're bidding against ourselves, or there's inefficiencies there. We limit our QPS as needed based on the types of media datatypes that we're looking for, the geographic needs that we have so that we don't kind of have the air conditioner and the heat on at the same time and create inefficiencies. We can very prudently and intelligently monitor the inventory QPS from the different publisher partners that we have so that we have what we need to deliver on the campaigns that are in our queue.
Jarrett Banks:
Can you look at inventory globally, from publishers in any language?
Jim Lawson:
Yeah. We're connected with a number of international publisher opportunities. Historically AdTheorent has focused on marketing in the US market only, and Canada. I'm sorry. We do believe that the EU and APAC represent incremental net new opportunity for us. We think it's exciting to think about, especially the EU because of the GDPR restrictions and the perspectives that they have about using personal data. So when we operate campaigns in the EU and we've had a few of those, again, it has not been our focus, but we can de-identify personal data from our EU campaigns. So we can operate there in a turnkey way without getting involved in relying on consent under the GDPR and there are lawsuits saying that even the consent framework is invalid, and there's just a lot of noise there.
And what we've said is in the EU, when we handle campaigns there, we will de-identify the personal data that comes into our system, so that there's no chance that we could possibly violate GDPR. But the punchline is the EU is a big net new opportunity for AdTheorent because of the fact that we're not personal, one to one in nature when we target our ads.
Jarrett Banks:
Got it. Who do you compete against most directly? And can you provide any kind of win rate?
Jim Lawson:
Yeah, I can't provide any kind of win rates, but I mean, I can say that we go up against the main players in ad tech. But I would say that also in some respect, we're kind of in very limited company. We work obviously in a market where the Trade Desk is doing extremely well. They're working in the open internet targeting ads and in the open internet, outside the walled gardens. I think we have a different approach. I mean, we've approached a bit of a different customer base and different customer makeup. We work very much with brands. We work with independent agencies and we're very performance-driven and we create custom solutions based on vertical. So we'll have an automotive solution that's different from our pharma solution, that's different from our travel and hospitality solution, different from our dining solution. We have different measurement capabilities for them. We have different data that we use.
So we've really tried to create a much more custom, vertical-based, hands-on, full-service offering. We have a full in-house creative suite. So I think what we're doing is really complimentary to what the Trade Desk is doing very successfully. The open internet's a great place to advertise as opposed to just the walled gardens. And so, I mean, we compete and go up against a number of different players, but our customer retention is strong. Our business has grown year over year, really because we execute, we do what we say we're going to do with our big customers. We'll get a couple hundred-thousand dollar test, we'll convert that into a million, into a multimillion dollar customer over the years. That's kind of our recipe. That's kind of our game plan. And we're going to just continue to do that. But we come up against the big players, we come up against the big platforms, but we think we bring something new and something that the market really wants.
Jarrett Banks:
Speaking of which, in your investor presentation, which I encourage everyone to view, you have a slide showing one of your biggest differentiators is keyword optimization. Could you elaborate on that?
Jim Lawson:
Well yeah, that's an interesting one to bring up. So, I mean, I think when we think about AdTheorent machine learning driven targeting, that is kind of in the context of a lot of other kind of legacy methods of targeting, so contextual targeting, many companies target contextually, which basically means that you assume interest based on the content on the page. And you, therefore, if you're trying to reach, if you're selling athletic apparel, you send all of your ads to a sports blog or a sports website. That would be contextual targeting.
What we have done both with contextual targeting, geo-targeting and other standard targeting methods is elevate that by saying, not only are we going to contextually target, but we're going to do it in a different way. We are going to ingest into our models data about presence of keywords in the URL that may or may not drive conversion activity. So rather than just assuming that this sports site is where you should serve ads for a sporting goods product, we're going to actually utilize machine learning to say, “machine, tell us which keywords are driving tennis racket sales.” And so we can ingest those keywords and run them through the models and not just rely on assumptions. Assumptions being, this is a sports site so these people are going to want to buy tennis rackets. And that's not always true. It's not always the intuitive assumption that drives the best business outcomes for advertisers. That's something that we've been looking at - our data science team has been researching and looking at it for 10 years.
Jarrett Banks:
That's an endlessly fascinating topic, but unfortunately we've run out of time. My thanks to Jim Lawson, the CEO of AdTheorent. Thanks to my colleague Alexandra Lane. And thanks to you, the audience. A reminder that this will be up on our website. The replay will be available at ipo-edge.com in about 30 minutes to an hour from now. So thank you everyone, and see you next time.
Jim Lawson:
Thanks for the opportunity. Good to be with you guys.
Cautionary Language Regarding Forward-Looking Statements
This communication contains “forward-looking statements” within the meaning of the Private Securities Litigation Reform Act of 1995. In general, forward-looking statements usually may be identified by through the use of words such as “will likely result,” “are expected to,” “will continue,” “is anticipated,” “estimated,” “may,” “believe,” “intend,” “plan,” “projection,” “outlook” or the negative of these terms or other comparable terminology and in this communication include, but are not limited to, future opportunities for AdTheorent and MCAP, AdTheorent’s financial guidance for the full year 2021 and, the proposed business combination between MCAP and AdTheorent, including the expected listing on Nasdaq. Such forward-looking statements are based upon the current beliefs and expectations of our management and are inherently subject to significant business, economic and competitive uncertainties and contingencies, many of which are difficult to predict and generally beyond our control. Actual results and the timing of events may differ materially from the results anticipated in these forward-looking statements.
The following factors, among others, could cause actual results and the timing of events to differ materially from the anticipated results or other expectations expressed in the forward-looking statements: inability to meet the closing conditions to the business combination, including the occurrence of any event, change or other circumstances that could give rise to the termination of the definitive agreement; the inability to complete the transactions contemplated by the definitive agreement due to the failure to obtain approval of MCAP’s stockholders; the failure to achieve the minimum amount of cash available following any redemptions by MCAP stockholders; redemptions exceeding a maximum threshold or the failure to meet The Nasdaq Stock Market’s initial listing standards in connection with the consummation of the contemplated transactions; costs related to the transactions contemplated by the definitive agreement; a delay or failure to realize the expected benefits from the proposed transaction; risks related to disruption of management’s time from ongoing business operations due to the proposed transaction; changes in the digital advertising markets in which AdTheorent competes, including with respect to its competitive landscape, technology evolution or regulatory changes; changes in domestic and global general economic conditions; risk that AdTheorent may not be able to execute its growth strategies, including identifying and executing acquisitions; risks related to the ongoing COVID-19 pandemic and response; and risk that AdTheorent may not be able to develop and maintain effective internal controls.
Actual results, performance or achievements may differ materially, and potentially adversely, from any projections and forward-looking statements and the assumptions on which those forward-looking statements are based. There can be no assurance that the data contained herein is reflective of future performance to any degree. You are cautioned not to place undue reliance on forward-looking statements as a predictor of future performance as projected financial information and other information are based on estimates and assumptions that are inherently subject to various significant risks, uncertainties and other factors, many of which are beyond our control. All information set forth herein speaks only as of the date hereof in the case of information about MCAP and AdTheorent or the date of such information in the case of information from persons other than MCAP or AdTheorent, and we disclaim any intention or obligation to update any forward-looking statements as a result of developments occurring after the date of this communication. Forecasts and estimates regarding AdTheorent’s industry and markets are based on sources we believe to be reliable, however there can be no assurance these forecasts and estimates will prove accurate in whole or in part. Annualized, pro forma, projected and estimated numbers are used for illustrative purpose only, are not forecasts and may not reflect actual results.
No Offer or Solicitation
This communication shall not constitute a solicitation of a proxy, consent, or authorization with respect to any securities or in respect of the proposed business combination. This communication shall also not constitute an offer to sell or the solicitation of an offer to buy any securities, nor shall there be any sale of securities in any states or jurisdictions in which such offer, solicitation, or sale would be unlawful prior to registration or qualification under the securities laws of any such jurisdiction.
Additional Information About the Proposed Business Combination and Where to Find It
In connection with the proposed transaction, MCAP filed with the U.S. Securities and Exchange Commission (the “SEC”), a registration statement on Form S-4, which includes a proxy statement/prospectus, and will file other documents regarding the proposed transaction with the SEC. MCAP’s stockholders and other interested persons are advised to read the preliminary proxy statement/prospectus and the amendments thereto and the definitive proxy statement and documents incorporated by reference therein filed in connection with the proposed business combination, as these materials will contain important information about AdTheorent, MCAP and the proposed business combination. Promptly after the Form S-4 is declared effective by the SEC, MCAP will mail the definitive proxy statement/prospectus and a proxy card to each stockholder entitled to vote at the meeting relating to the approval of the business combination and other proposals set forth in the proxy statement/prospectus. Before making any voting or investment decision, investors and stockholders of MCAP are urged to carefully read the entire registration statement and proxy statement/prospectus, when they become available, and any other relevant documents filed with the SEC, as well as any amendments or supplements to these documents, because they will contain important information about the proposed transaction. The documents filed by MCAP with the SEC may be obtained free of charge at the SEC’s website at www.sec.gov, or by directing a request to MCAP Acquisition Corporation, 311 South Wacker Drive, Suite 6400, Chicago, Illinois 60606.
Participants in the Solicitation
MCAP, AdTheorent and certain of their respective directors and executive officers may be deemed participants in the solicitation of proxies from MCAP’s stockholders with respect to the business combination. A list of the names of those directors and executive officers and a description of their interests in MCAP will be included in the proxy statement/prospectus for the proposed business combination when available at www.sec.gov. Information about MCAP’s directors and executive officers and their ownership of MCAP common stock is set forth in MCAP’s prospectus, dated February 25, 2021, as modified or supplemented by any Form 3 or Form 4 filed with the SEC since the date of such filing. Other information regarding the interests of the participants in the proxy solicitation (including AdTheorent and its members and executive officers) will be included in the proxy statement/prospectus pertaining to the proposed business combination when it becomes available. These documents can be obtained free of charge as indicated above.
Contact Information
For AdTheorent:
Investor Relations:
April Scee
AdTheorentIR@icrinc.com
Media Relations:
AdTheorentPR@icrinc.com
For Monroe Capital:
Investor Relations:
Theodore L. Koenig
Monroe Capital LLC
312-523-2360
tkoenig@monroecap.com
Media Relations:
Caroline Collins
BackBay Communications
617-963-0065
caroline.collins@backbaycommunications.com
# #