Okay. Hello, everyone, and welcome to Viant Technology Inc.’s fourth quarter 2025 earnings conference call. My name is David, and I will be your operator today. Before I hand the call over to the Viant Technology Inc. leadership team, I would like to go over a few housekeeping notes for the program. As a reminder, this call is being recorded. After the speakers’ remarks, there will be a question-and-answer session. If you plan to ask a question, please ensure you have set your Zoom name to display your full name and firm. If you would like to ask a question during the call, please use the raise hand function located at the bottom of your screen. Thank you for your attendance today. I am now happy to turn the call over to Nicholas Todd Zangler, SVP of Investor Relations for Viant Technology Inc.

Thank you. Good afternoon, and welcome to Viant Technology Inc.’s fourth quarter 2025 earnings conference call. On the call today are Tim Vanderhook, Co-Founder and Chief Executive Officer; Chris Vanderhook, Co-Founder and Chief Operating Officer; and Lawrence J. Madden, Chief Financial Officer. I would like to remind you that we will make forward-looking statements on our call today, including, but not limited to, statements regarding our guidance for Q1 2026 and other future financial results, our strategy, our platform development initiatives, including Viant AI, our pipeline and potential partnership opportunities, growth of our total addressable market, our share repurchase program, and industry trends that are based on assumptions and subject to future events, risks, and uncertainties that could cause actual results to differ materially from those projected. These forward-looking statements speak only as of today and we undertake no obligation to update or revise these statements except as required by law. For more information about factors that may cause actual results to differ materially from forward-looking statements, and our entire Safe Harbor statement, please refer to the news release issued today as well as the risks and uncertainties described in our Annual Report on Form 10-K for the year ended December 31, 2025, under the heading “Risk Factors,” and in our other filings with the SEC. During today’s call, we will also present both GAAP and non-GAAP financial measures. Additional disclosures regarding these non-GAAP measures, including a reconciliation of non-GAAP financial measures to the most directly comparable GAAP financial measures, are included in the news release issued today and in our earnings presentation, which have been posted on the Investor Relations page of the company’s website and in our filings with the SEC. I will now turn the call over to Tim Vanderhook, Chief Executive Officer of Viant Technology Inc. Tim?

Thanks, Nick. And thank you all for joining us today. We delivered strong fourth quarter performance achieving new company records across all key metrics. Revenue increased 22% year-over-year, and contribution ex-TAC increased 19% year-over-year, both above the high point of our quarterly guidance range. When excluding political advertising, revenue and contribution ex-TAC increased 28% and 24% in the quarter, respectively, and more accurately reflects the true strength of our business. Growth was broad-based across verticals, driven by accelerating CTV demand, strong digital out-of-home and mobile demand, increased utilization and further adoption of Viant Technology Inc.’s addressability solutions, and expanded use of the Viant AI product suite. Adjusted EBITDA increased 45% year-over-year to $24,700,000 for the quarter and exceeded the high end of our guidance range. Our fourth quarter performance completes a solid year for Viant Technology Inc. In 2025, revenue increased 19% to $344,000,000. Contribution ex-TAC increased 18% to $209,000,000, and adjusted EBITDA increased 29% to $57,000,000. While these are standout results as reported, our underlying performance was far stronger than these results indicate. Our contribution ex-TAC rose nearly 20% in 2025 while absorbing the effects of tariff-related pressure, cycling a difficult political comparison, and navigating the migration of a material client off platform due to a corporate merger. We were also able to increase adjusted EBITDA nearly 30% while absorbing incremental operating expenses associated with our strategic acquisitions of IRIS.TV and Locker. As we shift our focus to 2026, we foresee a year of accelerating performance attributable to a number of catalysts worth noting. First, we see a healthy ad environment, evidenced by strengthening customer demand trends observed through this point in the quarter. Our new flagship customer Molson Coors is live and actively deploying ad spend in the first quarter, with plans to ramp throughout the year and…

Thanks, Tim. I will provide an update on our customer go-to-market strategy, particularly as it pertains to the launch of Outcomes. But first, let us take a step back and survey the broader advertising landscape. This year in the U.S., total advertising dollars are expected to reach nearly $450,000,000,000. Of this, 30% will be allocated to brand budgets, while 70% will be allocated to performance budgets. Today, performance budgets have largely been dominated by search and social walled gardens, including Google, Meta, and Amazon. With the launch of Outcomes, we intend to compete directly for performance budgets, aiming to divert spend to the open internet by leveraging a complete end-to-end view of attribution across the entire customer journey, connecting initial CTV exposure to final conversion. Our go-to-market approach starts with our existing customers, where we see an opportunity to drive significant organic growth as we increasingly service their performance budgets, in addition to their existing brand budgets. Virtually all of our customers allocate spend to search and social walled gardens as part of a well-rounded holistic marketing strategy. We intend to leverage our existing relationships to showcase the effectiveness of Outcomes and win new performance budgets from our existing customers. This initiative is well underway, and I would like to highlight a few examples of our Outcomes product at work. Over 20 existing customers have extensively tested Outcomes, with a number of them implementing Outcomes on an ongoing basis, one of which is MacKenzie-Childs, a luxury home décor brand and prominent seller of tableware, kitchenware, and decorative home furnishings. In our initial tests, we ran two separate ad campaigns for MacKenzie-Childs, each identical in scope, with both campaigns seeking to maximize sales conversions over the same time period with the same budget. The only difference was that one campaign was…

Thanks, Chris. Before I begin, I would like to remind everyone that we have posted a presentation on our Investor Relations website that includes supplemental financial information to accompany today’s call. We concluded 2025 with a strong fourth quarter, executing against the key strategic priorities Tim outlined—CTV, addressability, and AI—and translating that momentum into record financial performance. Before diving into our detailed fourth quarter results, I will provide a high-level summary of our full-year performance. For the full year of 2025, we achieved record results across all key metrics. Revenue totaled $344,200,000, increasing 19% year-over-year. Contribution ex-TAC totaled $208,700,000, increasing 18% year-over-year. Adjusted EBITDA totaled $57,400,000, increasing 29% year-over-year, and adjusted EBITDA margin expanded by approximately 250 basis points year-over-year to reach 28%. Finally, non-GAAP net income totaled $41,100,000 in 2025, increasing 19% year-over-year. I will now move on to our results for the fourth quarter. Revenue for Q4 was $110,100,000, up 22% year-over-year and 5% above the high end of our guidance range. On a sequential basis, revenue increased 29% from Q3. Contribution ex-TAC for Q4 totaled $64,600,000, up 19% year-over-year and 1% above the high end of our guidance range. On a sequential basis, contribution ex-TAC increased 22% from Q3. Both revenue and contribution ex-TAC represent record results for a quarterly period. It is important to note, as Tim mentioned, our underlying business is performing stronger than our reported results indicate, primarily attributable to the difficult comparison brought about by last year’s high political ad spend contribution. When excluding political ad spend contribution from the prior-year election cycle, which weighed on revenue growth by approximately 600 basis points and contribution ex-TAC growth by approximately 500 basis points in the quarter, revenue increased 28% year-over-year and contribution ex-TAC increased 24% year-over-year on a pro forma basis. New customer momentum…

Thank you, Lawrence. We will now proceed to the Q&A session. If you would like to ask a question, please use the raise hand feature in your controls located at the bottom of your Zoom window. Our first question comes from Jason Michael Kreyer with Craig-Hallum. Jason?

Alright. Thanks, guys. I appreciate that. Maybe I will start off just where you ended that, Larry. You talked several times about the opportunity for accelerating growth through 2026. Maybe frame expectations for the year relative to the guide for Q1. Kind of what drives that upward swing as 2026 progresses. I want to step back and maybe talk about the late-stage deal pipeline that you guys had talked about a couple of quarters ago. Just want to get an update on how you feel that has progressed the last few months. How you feel about win rates in deals that have closed, and then just maybe what has been your ability to add or to replenish that pipeline of additional late-stage opportunities?

Yeah, certainly. Thanks, Jason. Well, if you break down—we have talked a lot about the tailwinds we are having right now—and if you break them down relative to our Q1 guide, which can kind of speak to what we see in the future quarters, first of all, in Q1, we had limited contribution from Molson Coors and WHOOP. Both of those advertisers onboarded during the quarter and only have spent modestly. We expect that for both of them to significantly ramp beginning in Q2. Similarly, with Outcomes, really a lot of early stages of testing, very little contribution in Q1, and as we move through the quarters, we expect that to obviously contribute nicely. We have talked about other customer wins that we have not announced—we talked about them a bit generically—many of those are also ramping up in the second and third quarters, so we expect to get a lift from that. And relative to Q1, you know Q1 is historically our lowest quarter, and I think this year, based on our mix of clients, maybe we are over-indexed a little bit towards customers that have the most negative seasonality in Q1, which impacted Q1 guide a little bit, certainly relative to Q4. But we see a nice ramp up as we move through based on the new clients we are winning, Outcomes coming through and starting to build up, that it will build nicely. We are very confident that it will build nicely as the quarters progress in terms of growth.

I would just—I will start with that. One of the big areas of investment around operating expenses is building out the enterprise sales team, and so I think we have done a great job of putting leaders in place that are from high-quality places. We have pulled leaders in vertical categories from Yahoo and many other very high-quality companies. So that investment is going to continue to replenish that sales pipeline. It is not just the win rate, I would say. We are beating much larger competitors at late stage in the game. We are always up against a very large competitor, and typically you are seeing the advertisers select Viant Technology Inc. for the innovation that we are pumping out. So that pipeline continues to grow. We talked to most investors—we mentioned last year around $250,000,000 of pipeline. We have closed very big wins in that. Some of those have been delayed to this year. A lot of times when we do not win a customer, they have chosen not to make a change until a future period because it is a fairly big lift to change platform providers, and so I think some of those will get kicked into the back half of this year as that determination of when to switch. You want to add anything? I would just say, though, in these pitches, it is becoming very apparent of our advantage around our proprietary data—both around Household ID, the continued scaling of IRIS ID, our supply quality models that do not get enough attention but clients find incredibly valuable, our Direct Access program of being able to be directly connected to the largest content owners in the world. All of that is really opening a lot of eyes with a lot of these large brands. And really these large brands, I think they all have a commonality: they have to drive higher growth. And they have—many of them are looking at their playbook that they have run for the last four or five years, giving the largest platforms in the world most of their money. While those companies have outsized growth, the largest platforms in the world, their business suffers. And I think we are a great counterpunch to that, and a lot of marketers are really taking a look at the proprietary data that we have and saying that is a way for them to deliver more growth in the future years.

Perfect. Thanks, guys. Appreciate it.

Thanks, Jason.

Our next question comes from Laura Anne Martin with Needham. Laura?

Somebody just has to tell you that because—and I am going to ask about your growth in the quarter. So when we look at DV360, their third-party was down 2%, The Trade Desk up 13% in net revenue, you guys up 19% in net revenue. Really big size difference—like, you guys are tiny compared to those two. My question is, are you taking share from these—you just said you were taking share from these bigger companies. How much of this is sustainable over time? Because I sort of feel like Wall Street thinks globally scaled large footprints have competitive advantage over smaller companies. Right now, you are disproving that, but convince me that small can win at these much higher revenue growth numbers than Google and Trade Desk, who are your sort of globally scaled competitors in your direct business, because these numbers are amazing. And then I wanted to drill down on IRIS ID. I remember at CES a year ago, we were talking—you had just bought IRIS. Maybe it was two years ago, I am sort of forgetting. You said it was up five times year-over-year, and you are now at percent of the incoming CTV bidstream had IRIS IDs. What is the gating factor there? Would you expect that to get to 75%, 100%? What is stopping more usage, or do you expect that kind of up 5x to continue over in 2026 and 2027?

Why do you not take IRIS?

Yeah. So with IRIS, we have seen incredible adoption of the IRIS ID. What that means is that we need content owners to carry it, and we have made announcements with some of the largest content owners, the largest television networks. We have had the IRIS technology in and of itself to be able to run computer vision, to contextualize video, pull out emotional sentiments, check for brand suitability. This is checking a lot of boxes for marketers. Most marketers, I will say, are completely unaware of the fact that when they buy CTV, they are only able to buy at the app level through other platforms. So you can only buy the app. A lot of these large apps have 20,000 to 30,000 titles of content, so a brand may not be the right fit for half of those content titles. And most brands are unaware of that, and when you give them that problem statement, IRIS completely answers that. That is number one. Number two, marketers have been testing it. As we said, revenue grew 90% from Q3 to Q4. So huge growth. Why? Because they see the performance improvement. When they see the performance improvement, they bid higher for those IRIS IDs that are relevant for that brand. It is driving, on average, we have said, a 466% increase in conversion rate. Forget upper-funnel metrics—brand awareness and consideration—just straight up conversion rate and sales. So marketers bid up for it, content owners get more money for it, that increases the amount of content owners that will then carry the IRIS ID. So we are at approximately 50% now. We believe we are going to continue to scale that. We think it is reasonable that we would get to 70% penetration this year, and so the future looks really bright for IRIS, and it is also really bright for our customers because they are taking advantage of that and they are driving greater returns. Again, being a buy-side player, we only care about what drives our customers’ business. If we drive their business and their growth, they are going to spend more money on our platform.

So, your first question—can the smaller company beat the larger companies? I think we are proving it now, and we really believe it is sustainable over the long run due to proprietary data. When you can only target at the app level and not deliver the performance, that is really a big gating item. The second concept here is that the large platforms have been self-reporting their own success back to these brands. Meanwhile, the total sales of the brand are actually down. So these things are not correlating, and they have had now half a decade or a decade of working with these larger companies where this is just a continual output year after year. So they have really lost faith in the reported metrics that the platforms are using. I think they are looking for an independent buy-side platform to help them understand what is driving success for their business. So what drives our success? It is proprietary data and Viant AI, which is the automation and the autonomy where they can get way more productive with their media dollars at work. And I would add to that Direct Access. By pulling out all the middlemen, the same dollar has more working media. They are getting more ad impressions per dollar than they were prior to Direct Access being there. It is really the combination of all of this that is driving better efficiency when you work with Viant Technology Inc. relative to—you mentioned Google, The Trade Desk, and Amazon. Amazon has a different type of perspective where they are really good at subsidizing businesses in the near term—you are seeing that with their 1% fees that they have been out in the marketplace—or bundling of the products of AWS plus subsidization. But in the end, Google and…

Thank you very much. Great numbers.

Thank you. Thank you.

Our next question comes from Tom White with D.A. Davidson. Tom?

Great. Thanks for taking my questions. Nice end to the year, guys. Maybe just with regards to your commentary about the expansion of your addressable market, you know, if I think back, it seems like a lot of the recent product innovation that you have launched were initially conceived around going towards the smaller end of the market, right—those search and social advertisers. But over the last several quarters and thus far this year, it seems like you are getting traction with the bigger boys with some of this innovation or at least getting their attention. When you look out to 2026, 2027, 2028, what is the bigger opportunity, do you think, for you? And then just if you could quickly comment on IRIS ID as a competitive moat. Obviously, you guys have a head start there, and the numbers look great, but what is stopping any of the other big platforms from going out and trying to convince content owners to start embedding this ID or coming up with something similar? Thanks.

Yes. I will answer the first one on the expanding TAM. You have to have what we see as a commonality—you think in big brands, they have brand-based budgets where they are looking to raise awareness and consideration for their products and services that might pay in a future period, but they also have performance-based budgets where they have to get sales now. And really what you have to do is create ad products that address both of those, and that is really what we have aimed to do. And if I look at these DTC brands, many of them start only in performance. But they quickly realize that they tap out in Meta or Google’s Performance Max or Demand Gen or Search—they tap out there because they cannot drive growth after a certain period. So then they realize, oh, we have to invest in our brand to raise our baseline sales. And so what we are seeing is that when we go down market to these direct-to-consumer e-commerce companies, we are seeing that what they need is they need to tap into CTV. That is really going to drive growth for them. But they need tools; they need a level of workflow that they are used to in some of these platforms like Meta’s Advantage Plus or Google’s PMAX. So we deliver that with Outcomes. But they are tapping into a really high-growth channel that not only drives brand awareness but also is capable, as I said, with solutions like IRIS ID, they can drive the lower-funnel performance for sales now as well. So, what is bigger? Look, the down-market DTC and e-commerce companies are small businesses. Meta and Google—they have an audience of 10,000,000 businesses that buy advertising from them. If you look at the open internet, I do not know, 10,000 to 20,000 companies buy advertising in the open internet. And how many companies buy television? Maybe 1,000 to 2,000, something like that. So we are looking at addressing—we want to, again, all marketers of all sizes have similar challenges. You have to create products for both. But we see them both equally as appealing.

On the IRIS ID question on why someone could not just copy it, it really is the network effect of IRIS ID, and it is why we hit it so hard last year in scaling that ID. Network effects of tying the ecosystem around this identifier, and that is why getting to critical mass was so important for us in 2025 in scaling that. How do we do it? We have done over 1,400 integrations with content management platforms—all various content management platforms. Even a big content owner, they will have many content platforms underneath it. So we have done all these integrations—again, over 1,400. That takes time, resource, and effort to actually get done. Or you could just adopt the IRIS ID. And every big platform would have to go do the similar types of integrations to replicate what the IRIS ID brings. The second area of network effect is just the OpenRTB protocol that we operate in. There is only one spot for a content object in the OpenRTB protocol, and IRIS ID is implemented nearly 50% of the time in that spot. So the content owner is really not incentivized to bring a second one in because you cannot even get it through the RTB protocol with IRIS ID installed. So there are a number of factors there. I would just ladder it up in total to network effects of this content identifier that we captured in 2025.

Thank you. Very interesting, helpful color. I appreciate it.

Thanks, Tom.

Great. Thanks so much, and congrats on the quarter. First, I wanted to ask about Outcomes. You mentioned that you ran pilots in Q4 with a number of clients implementing Outcomes. Anything you can share maybe on the initial conversion rate and where you see that over time? And then how should we think about incremental uplift to monetization as you roll out this functionality across your broad advertiser base? And then would love to hear your thoughts on The Trade Desk–OpenAI partnership and what that means for the programmatic space more broadly and then for your platform more specifically. And what are your thoughts on being involved in some of those emerging AI services?

I do not have the exact numbers on conversion rate, but there are just a number of factors. Obviously, the cost efficiency relative to the sales that Chris touched on in our prepared remarks—really, really good compared to what anybody has seen from an autonomous platform. So I think overall conversion rate is hard to give you an exact answer to that other than we are beating what the current status quo is, which is manual optimization or some level of automated optimization that is out there today.

And I would just say, you know, the real through line here about the performance improvements is the fact that we have proprietary data signals that are extremely valuable. But when you couple that with an autonomous workflow, the speed of that is what is driving the improved campaign results or the performance improvements. And it is doing it at a level of reliability for marketers that is way greater than that of human-based, stressed-out workflows. When I think of the jobs of traders—the gun that they are under, so to speak—they have to come up with the optimization strategies and the tactics that they are going to pull, and they do that on a day-to-day basis. And it is an absolute grind, and that leads to an instability in the reliability of the metrics. And really the autonomous workflows that we have put out here are really driving tremendous value, and it is what we think marketers over time are going to continue to adopt.

Yeah. Obviously, the number of users using bots is really exciting when you look at it as a brand new channel. It is kind of like social when it started to originally emerge and users flocked to it. So there is a ton of real estate available there for advertising. I think OpenAI’s strategy in partnering with third-party DSPs is kind of like Facebook’s early strategy. They were a part of RTB; everyone was involved, and then they pulled it all back and went with their current go-to-market. So we are always a little slower in going to work with organizations like this because we are mindful that they may change strategy overnight, like you saw with agentic checkout with commerce transactions—that has already been abandoned. So I would caution investors about putting any level of excitement until you really see what is the ad format, how does it actually work, and what level of data would be shared. I think the announcement around The Trade Desk—it does not really fit with the RTB protocol as we do. You would have to have sensitive user-level data be sent across. Usually, big platforms do not pass that level of granular information due to consumer privacy reasons. So the truth is we do not know what OpenAI is thinking here at Viant Technology Inc. We are watching, but there is a huge amount of users, a huge amount of real estate and time spent, and certainly a whole bunch of interesting insights that OpenAI knows no one else knows—kind of like search data is unique. But when I chat with an application, I am very rarely ready to buy right now. I am usually middle of the funnel doing some research and information. So although very interesting and exciting—so exciting around future opportunity—I do not think that is a 2026 revenue generator in a large scaled way. Of course, these guys are innovating at incredible rates, and so I may eat my words in the back half of the year, but we are watching, we are paying attention to it, and it appears to be chatbots appear to be a brand-new channel that is opening up tons of available inventory for advertisers. So as we learn more and go throughout the year, we will certainly participate where it makes sense. Relative to The Trade Desk, I think Criteo—who has actually been announced; I do want to say The Trade Desk was a rumor, and interesting timing there. But Criteo has been announced—that makes more sense around product listing or shopping-based ads that they could provide given Criteo’s customer base. So Criteo feels more like a natural fit. I would say The Trade Desk and Viant Technology Inc. seems a little bit different.

That is very helpful. Thank you.

Thanks, Maria.

Our next question comes from Matthew Dorrian Condon with Citizens. Matt?

Thank you so much for taking my questions. My first one, just to follow up on some earlier comments about Amazon. They have announced new partnerships with Netflix and their integration with Roku. It seems that they are obviously pushing more and more into the ability to service third-party inventory. Can you just talk about how you would expect—I mean, obviously, today, a lot of spend is going through Prime Video and into Amazon’s own platforms—but they are clearly more aggressively going after that third-party inventory. Just how do you see that shaping up in 2026 and 2027? Why are they not as big a threat as maybe the media seems to portray them? And then just a follow up on—I believe when you landed Molson, they talked about findability being the key metric that was the reason why they went with you guys. For WHOOP, was there a similar metric that they found? What was the product that really got them over the goal line to go with Viant Technology Inc.? Thanks, guys.

Well, I do not want to say they are not a threat. They are a threat. They are subsidizing their products. They are doing the bundling strategy that Google executed. So it definitely is a threat, and at Viant Technology Inc., we are paying a lot of attention to Amazon. So I do not want to discount Amazon as a competitor in the space like some others have. I think we do focus on Amazon. But what Amazon knows—they know a lot about customers of Amazon. They know very little in all the other retailers like Walmart, CVS, all these other products. And so if you are a QSR, is Amazon DSP a good fit for you? Likely not based on the proprietary data that they have. If you are a product that is sold through Amazon, Amazon DSP makes a lot of sense to actually partner with to track the sales and reach consumers on Netflix or some of the other platforms. I caution the other big content owners out there because if Amazon runs the same playbook as Google, what they do is they say Prime Video outperforms every other app on the buy. And it is all about—by doing that integration—we bought the ad, but hey, the Roku app was not as good as Prime Video. I can basically write what the report is going to say, as long as they follow that same strategy—and they have been. So I think they have a major trust issue when it comes to what they are reporting in the platform if the transaction does not happen on Amazon.com. That being said, a lot of products and services are sold on Amazon, and I think it makes sense for those customers to use the Amazon DSP in that way. But if your product is also sold in 50 other retailers, the Amazon DSP really is not a good fit for you.

Yeah. So they had a huge focus on CTV. This is a growth brand. They are very focused—they are a fast-growing company. They are very focused on continuing to grow their brand. And in CTV, what do you want? You want addressability—both in terms of “I want to reach the right households, the right people,” and then “I also want to know what type of content they are consuming so I can make my ad relevant to that content.” That drives performance. So, yeah, heavily looked at our addressability solutions. A lot of clients kick the tires hard on that right now, and they see our scale relative to other players is—ours is dramatically larger. And one of the big reasons is we have been at this—this is not a two- or three-year effort. We have been at this for over ten years. We are a leader in this space. So I expect many more brands to continue to focus around addressability. A lot of people think addressability is just for targeting. The largest advantage of addressability is measurement—for you to truly see, “I showed a CTV ad, did I get a sale?” But it is not just that. It is what else did they do in the journey? “Oh, we showed a CTV ad. They then went to Google, searched, later were exposed to a social ad, and then purchased.” That level of visibility that you can give to a marketer and help them properly allocate their money accordingly is incredible. Without an addressability solution for measurement that we offer, they will continually just put money to whoever showed the last ad—which marketers are increasingly not falling for anymore.

I just want to add to that too, not about WHOOP, but about Molson around the addressability solutions of IRIS. If you are a regulated industry like alcohol, you cannot show ads in children’s content, and so IRIS becomes a critical content identifier as well for you to actually deploy money with confidence that you are not going to get a fine for showing ads in children’s content. And it is the combination of these two proprietary data signals that we have that is really pulling the large enterprise customers our way.

Okay. Our final question will come from Barton Evans Crockett with Rosenblatt. Barton?

Okay. Thanks for squeezing me in. So I was wanting to ask two questions, really. First is, just looking at the growth rates that you are talking about, contribution ex-TAC ex-political up 24% in the fourth quarter, but slowing to the high teens in the first quarter. Do you see the ex-TAC revenue number at some point returning to what you were doing in the fourth quarter? And I know there were some seasonal factors in the first quarter, but that is kind of a notable slowdown. So I was wondering if you could address that first. And then the things I was just wondering about in terms of the LLM debate—you mentioned that it might be easy for an LLM to code an interface, but the value is really elsewhere, kind of the data and presumably the execution. Would it make sense for someone like a Viant Technology Inc. to perhaps use an LLM, though, as an interface, as a way to perhaps penetrate clients that are now wedded to The Trade Desk interface at the agency level—essentially to be an MCP where the execution is through you, but maybe the front end is Claude? Is that conceivable? Could that be an opportunity over time? Or is that something that is just not on the table because of the risk of them getting too much leverage or scraping your data?

Yeah. First is it is just the mix shift of our advertisers in the way that they spend their money. It is better to look at it on an annualized basis. So if you look at our contribution ex-TAC on an annualized basis of 2025, we are going to, hopefully, outperform that in 2026 given the customer wins. And as we mentioned, WHOOP, Molson Coors—some of the large advertisers were, I do not want to say de minimis in Q1, but slowly coming on, learning, understanding how things go. So those ad spends will kick in in the later quarters. And so I think the biggest thing to take away from our call is we are going to grow these numbers sequentially throughout the year. And as you look at the second half, political really kicks in—about half the money is spent in Q3 and about half the money is spent in Q4, roughly, is the way that it goes all the way up until that election cycle. So I would really caution everyone to look rather than quarter-to-quarter—when you look at advertising-based businesses, there is some seasonality. There is a mix shift of the various advertisers on our platform that is really driving the Q1 number that you are seeing. But I can tell you the pipeline is strong, the growth of our business is very strong, and we think we will deliver just like we did last year. No. Look, the Viant AI interface is an LLM interface today. The users of that interface are not with the traditional self-service user interface. So I think we have already delivered on that. It is getting users comfortable with that. I hate to say old habits die hard, but people like to click buttons. It is just a lot of work, and they want to know, is there a hallucination in the data? So a lot of this is test and learn and users getting comfortable with this new interface. That is a big one. As far as an MCP, it is certainly going to be a big factor in the future. We are thinking it each and every day. But again, the market is moving so fast quarter to quarter. You kind of have to project out, and so I think if the interface is your moat, you are in serious trouble in 2026 this year and in the outer years. So I think we have a lead there. We delivered it, I think, eighteen months ago or so. We delivered that interface to customers for them to initially test and learn on, and people love it because there is no training, there is no certification. You do not need to go to Trade Desk Academy for two weeks, and still make mistakes after that. So I think overall, interfaces are dead. Dashboards are dead. You really want an LLM to deliver the info. And I would just say,

what we are doing today—if you look at digital advertising and programmatic—these are human workflows. There are entire organizations that are built around these human workflows. When Tim is saying old habits die hard, although there is an incredible amount of innovation, and I believe that we are leading in that. We are the ones who are bringing from human-based workflows to autonomous workflows—autonomous being the key word. Where is all of tech going? It is going autonomous. So we are leading there. However, if I rolled out and said all brands today, “If you want to interact with me, you must build your own agent that then plugs into my MCP,” we are so far away from that today. I know that there are early—I would say—kind of the green shoots that are out there that we are looking at, and I think those are going to be in the DTC e-commerce space that we are first going to see that. We are very focused on that market. So as we think about our own go-to-market as going after these DTC and e-commerce brands, we think that is a really good solution there. But over time, it does make a lot of sense that an agent will come to the infrastructure that provides all of this and will want to go to infrastructure with incredible proprietary data, which we also have. So, we do look to enable that in the future. And let me just add, we believe

that the LLM is the commodity. It is the proprietary data on top of that LLM which is unique, or the application layer tied to the RTB infrastructure that we actually have. And I think, getting to Chris’s point here around humans—right now you have humans in a five-step process. The human is in the middle of it all. We have taken the human out of the middle and put it at the very front of the line. You set the guardrails, you set the goals, and then you let autonomy actually happen. And so that is really the big difference that I see. The LLM—most of their moat is with consumers. ChatGPT has 650,000,000 consumers spending tons of time per day on that. That is not commoditized—that is very valuable as a media seller. But for us, from an enterprise perspective, the LLMs are commoditized. I could swap out Gemini with OpenAI. I could swap out OpenAI with DeepSeek. It does not really provide noticeable levels of difference in the output sets there or noticeable levels of difference in quality. So to me, from an enterprise perspective, the LLM is the commodity. Proprietary data is the moat. And I think the commodity piece—the reason why many of them give you back the same answer—is that they are all trained off of the same data. They are all trained off of scraping the web, all of them. Certainly, you could argue some of them have proprietary data assets—certainly. That is very—I think that is becoming understood that that is the piece that is commoditized. But that said, to your core question, will we enable third-party agents to be able to interface through an MCP into our infrastructure? Yes.

Okay. That is interesting. Thank you.

Thanks, Barton.

—concludes the Q&A portion of the call. Thank you.

Thank you, everybody.

Have a good evening.