Okay, we're going to get started. I'm Laura Martin. I'm the senior media and internet analyst at Needham & Company. I'm going to introduce Mark Douglas, our next speaker. Mark is the founder and CEO of Mountain, where he has spent over a decade transforming connected TV into measurable performance marketing channel. In addition, Mark recently led Mountain's 2025 IPO, a seasoned engineering leader and entrepreneur. He previously served as the VP of engineering at Magnite Rubicom and VP of technology at eHarmony, where he developed the platform's foundational matching algorithm. So I want to start with a question about leadership. And I think most people would agree that generative AI is transformative. And so what kind of demands does that put on you as a leader that are different from when maybe times are a little more incremental?

Yeah, well, when something is transformative, I think everyone has to decide, is this going to become an enabler? Or is this meaning, is this something I'm excited about there's this something I'm afraid of and so for me personally I think one is I like change the as a matter of fact I love change and so I see AI as completely enablers something that me personally and the team in mountain have kind of jumped on very very early we're the first introduced AI targeting which we call mountain match we I think when it first introduced create AI creative tools because the vast majority of our customers have never advertised on TV before so we had to enable a way for them to build TV ads we're putting AI throughout our platform we're adding MCP or we just added MCP support meaning you can MCP yeah oh that's how AI platforms talk to each other so that you can have like open chat gpt or have plaid basically you can use them to tell our software what to do and that's something that's also important so the it's all about just basically seeing these opportunities seeing these changes as opportunities rather than seeing them as a threat and we definitely view it that way okay does it does it put any demands on you as a leader to handhold more or be more visible with the humans in the enterprise because of the fear level? You mean within the company? I think a lot of people are viewing AI currently at this moment as something that reduces cost. We're seeing it as something that increases speed. And there are cost benefits, but the technology doesn't come for free. And so when we articulate, when I articulate to the company and to our customers as we want to enable more faster and they are a part of that, then, you know, I don't think that drives fear. I think that drives excitement that enables us to be a market leader in adoption of new tech like general AI and other forms of AI.

Okay. Does it make a difference? You guys don't have any physical plant, right? You guys are all distributed. Yeah, we have no office. We're fully remote. So does that make your organization more easily adaptable to technological change?

I think it does, mainly because we had to, so post-COVID, we never went back to office. We've been fully remote. We've hired many, many people who are fully remote. And so it drives a discipline in the business in terms of setting clear goals, measuring the effectiveness of that. we don't manage people based on time we manage based on result on goals and results and so I think just the nature of the company DNA is one of doing things differently again enables I'm not I don't have conversations in the company at all about people like viewing as fear I've had it as like if anything like what they want to I want to and they want us to see around corners like Like, how do we enable, again, rather than kind of be threatened by that? So I'm completely excited about all the technological change that's occurring before and unrolling it.

Right. So you and Reddit are both 100%, I'm going to call it virtual, where you don't have actually a head office. And I'm just wondering if that affects the adaptability to generative AI. And your point, which is well taken, is if you're already 100% distributed, you must have clear goals and deliverables. Otherwise, how do you tell if somebody's pulling their weight? So that discipline from not having an office and not just being in office means people have to be more substantive.

We also, like I was talking to our head of engineering literally last week, and we were talking about adoption of AI coding and the various ways it's being used and what are the right ways and what ways don't make sense. And so I don't think that being fully remote is playing much of a role. It's much more about is the leadership of the company leaned in? Are they tech savvy? Even if it's a tech company, it doesn't mean you always have tech savvy leadership. Are they tech savvy and are they leaned in? And are they curious and are they actually looking not just for adoption to say it's adopted, but to produce material benefits? So in our engineering organization, we are definitely like, what is the right way to do AI coding, to use AI tools, to make sure we have maintainable code and things like that? Same thing happening in our sales organization. We leverage AI as part of our go-to-market motion, but we obviously don't want people to feel like they're talking to a computer if that's not the conversation they want to have. So you have to leverage it very smartly. And those are conversations me, my COO, my head of engineering, my head of sales are having regularly.

One of the unique things about your strategic position is that increasingly in CTV, advertisers want to tie their connected television purchase to performance.

You guys really embedded the market in this in performance CTV. but what's changing in 2026 and is that good for you or bad for your economics yeah so the the literal concept of performance we the term performance tv was created by by mountain and we deliberately did not trade market because we wanted we the whole idea when we launched mountain performance tv i said to the team the easiest way to win a market is to create that market so we're to create the market for performance television and that's what we did we anticipated that if we're successful in that it will eventually it's not just mountain it's bigger than mountain so now that we see other companies using that term number one that's a validator that the of the scale of the opportunity here the tam um and number two it you know puts more customers in the market It kind of starts to move the market from early adopter phase where we're evangelizing to more mainstream where companies, especially S&B companies, believe they can be there. For Mountain, we are purpose-built for the S&B portion of this opportunity. So I think, you know, so other companies may potentially be looking at performance should be a part of the enterprise brands. That's not an area we play in. We feel that SMB has unique needs, unique go-to-market motion, unique knowledge and talent needed to win there. That's where we're the market leader, where we intend to play, and we feel very secure in all the capabilities we have in our products and also our go-to-market motion around that.

Okay, so performance TV done by other people isn't really necessarily SMB-targeted. They're doing something different.

They're more folks. What's happening, I think, when you first brought up the point, is larger brands are also saying, hey, we should be measuring the impact of the spend. There's tens of billions of dollars a year, and we agree with that. It's just that segment of the market is not a segment we're pursuing, and I think some of the bigger enterprise companies in the TV space find it very difficult to come into the market segment that we're in.

Yeah, that makes sense. So one of the things that happened in the most recent quarter is your revenue slowed, and then your guidance is for slowing revenue. Are we running up against the TAM in SMBs, do you think?

No, I think we basically have a lot. As the opportunity moves from an early adopter to mainstream, we are basically enhancing a lot of things in the business. We added Garland Hill, who's the former head of growth at TikTok, now Mountain CRO. He built TikTok's SMB revenue literally from zero. He was the first person to billions in revenue. We have a lot of product initiatives rolling out, and we just thought it was prudent with so many things happening at the same time. We just wanted to be very prudent in our guidance, and that's what we did. But we're not seeing any long-term revenue slowing, certainly not any capping of the TAM or anything like that.

Okay, because you wouldn't think hiring a big, huge executive from TikTok would slow your revenue growth. So to me, that's the opposite. If you hire a big guy, you'd think he'd accelerate revenue growth, but I'm getting...

Right, which is why we guided hire in the second half of the year.

All right, so we think he has impact in the second half of the year. Okay, cool. All right. It feels like... I do have a lot of competitors like Buyant saying they're going to target the SMB. They're doing self-service. They're adding it to their core DSP, and they're trying to come down into your market. Do you just think they won't be successful because they weren't purpose-built for SMBs, and it's a harder hurdle than they're exposing?

I think there's a couple things. One, easier said than done. You actually have to build the technology and succeed in doing so. The go-to-market motion is very difficult. And then most of all, we actually are not just the market leader in terms of the idea of performance TV. We are the market leader in terms of the actual performance. and so all and that is many many years of investment in performance technology plus now many two plus years of investment in AI technology around performance and we're not standing still so more than half a mountain is engineering so we have a big investment in technology and we I honestly think not everyone is going to succeed in that effort and the ones that do there's more than just having a self-serve platform you actually have to deliver performance on the level with paid search and paid social but do so for television and that's what mountain has done for many years

now and we're continuing to grow in terms of and one of the things we're seeing is where these people decide to go into your business they're hiring separate sales forces because they can't get their normal sales force that calls on big clients to call on smbs yeah and so how do you do that? Do you guys have sales or is it mostly, like you're mostly using programmatic advertising to get your clients?

Yeah, it's a combination of we have lead gen using our own product as well as using social and other channels, LinkedIn and other channels as well as we have, yes, a very efficient go-to-market motion on the sales side. For the mid-sized customers there's a bit more selling involved because there are more decision makers involved. for the very small for the much smaller businesses it's much more marketing driven but again that didn't happen overnight we had to build that out we had to learn how to do it we had to grow it and then once you do all that now you have someone who's like yeah this sounds exciting like to try it and then and then you actually have to deliver the goods which is you help them grow their business not just you grow your own yeah they don't renew i assume if you don't help sales lift and renew is spend tomorrow there are no contracts in performance marketing like there's no one has signed a contract to run ads on google adwords right so it's not like next week it's like i spent money today do i see the sale tomorrow yeah i mean they generally want to see results within the length of their sales cycle so they like um like one of our customers is hex lab and you know their sales cycle is people look at the pans and then they look go shopping someplace sales and they get interrupted by their kids it takes more than sometimes a few days to commit but within a reasonable sales cycle most sales cycles we see for e-commerce for multi-hundred dollar price points are in the range of like you know one to three weeks yeah so they're they're judging within those kind of time horizons not not literally a day overnight yeah but they're not like I'll wait till next year when the contract green is because there is no contract at all in performance marketing. Meta doesn't sign contracts. Google does. Neither does not.

Yeah, okay. And are they having an always-on all year? Are you bringing in an ad campaign, or do you have an always-on kind of idea?

Yeah, it's always-on. We call that evergreen, and that's a pretty common term in performance marketing. These are evergreen campaigns. You see some... If they have a very seasonal business, they might...

Like Christmas or Mother's Day.

Exactly. Like we have a customer that does like pool stuff so they don't sell that. Of course. Pool supplies. So that's that stuff. So they shut down in the middle of the winter. They're not doing much marketing. But for the most part, most of our customers are evergreen. Budgets change throughout the year based on seasonality. If they're doing a new product launch, there might be extra. But for the most part, performance marketing is evergreen marketing.

And I remember a big client of yours early on was Purple Mattress, wasn't it? Because that was a performative. And then what happened to that?

I think they haven't, over the long term, done well. But we have a number of other mattress brands. We somehow, in the early days, became big in the mattress sector.

And you've kept a lot of those guys.

We've kept a lot of them. Yeah, we're pretty happy with that.

Well, I just bought a new mattress. It wasn't Purple. But it comes all bundled up. It's like the size of those stack of papers. and then it opens up into a queen-size mattress over 72 hours. I know, it's insane.

It's pretty insane. We actually just picked up a pretty well-known mattress brand. So it continues to be a strength of the business.

Yeah, very interesting. And I guess, like YouTube, you don't really care which mattress brand works as long as they're doing lead gen through you guys.

Well, I mean, ultimately. But, I mean, I do, I tell our team all the time, because I feel like as a company, we don't sell ads. What we do is we connect consumers with brands and products they're going to potentially love.

Like a lead gen kind of commerce?

Yeah, I mean, when you take it down the marketing terms, it's lead gen. But it's like for every brand, there's someone who's excited about the products they've built. And what we do is we connect them with consumers through 30 seconds of uninterrupted video on the biggest screen in the home. And that hopefully leads to a lot of leads and a lot of purchases. And then everyone's happy. The brand is happy. The consumer's happy. You know, you got a new mattress. You hopefully, yeah, you're happy with it. Everyone's happy. And I think sometimes people forget that television is such an amazing storytelling medium. And so for us to connect brands, to that medium, we're literally telling this story into nearly every household in America and helping those businesses grow. I say this as just like, it's kind of an emotional underpinning of what we do as a business, is connecting brands with consumers that are potentially gonna hopefully love those products. So that's kind of how I describe what we do.

And one of your competitive advantages has been this sort of unique access you have to connected television ad units. Why don't you walk the audience through how you get this unique CTV ad inventory?

Yeah, so consumers, meaning are watching TV everywhere. And I think one thing that sometimes people forget, the number one consumed entertainment medium is not social media, it's television. 5.1 billion people a day watch television worldwide on average three hours a day. 4.1 billion use social media. and then the content that people talk about there are funny memes out there on social but the big conversation piece over dinner is often like what shows are you watching what season finale did you see or did you see this so it's just a very powerful medium that small business were cut off from so we connected that we went to every streaming network in America all of them Apple TV doesn't have ads and when we first launched this platform and we built direct relationships to access their inventory programmatically. Meaning we have a programmatic technology stack that targets the consumer, bids for each campaign that's running, bids on that inventory. So if that person in that audience is watching Disney or they're watching Landman on Paramount or they're watching Warner Brothers, any network or Fast Channel, we are able to access that inventory bid on it for specific campaign when we win that auction then we stream a 30-second tv commercial into the home during a commercial break and then we measure the results of that and all of that's provided in our platform including the ability to create to create the ads through ai technology now and so you you don't see the end of any of those contracts coming because that is a big competitive advantage to see new content no no not at all i mean i think that we we are considered by most of our part our our media network partners as a growth channel so i think the key stat is more than 95 percent of mountain customers have never advertised on tv before so this is net new customers revenue it bring brought into the industry, so we're a growth channel. We're not competing with them at all. We're not competing with their sales teams. We are literally just bringing new companies into the television ecosystem. And we were talking earlier, the SMB opportunity. They're recognizing how important SMB is, and they see us as we kind of created that segment of the market.

A lot of those providers that have very high-quality video are very sort of crazy about the quality of the ad.

And some of these SMB ads you see now, like on Fox News or something, are pretty sort of like low quality is what i would call it and especially if they're performance driven right so how do you keep the quality standards up so that you don't get kicked off one of these very high-end content suppliers well ironically we actually use ai technology to review the ads yeah there is a little bit of a human it used to be entirely human now it's predominantly ai and ai will technology reviews and flag it for a human review um so they do get reviewed their technology standards both um automatically enforced meaning um it has to be a 30 second tv commercial down to the frame like like you can't even have a single frame missing that's automatically maintained but the also it's just bringing down the cost of creative so they can build better quality creative but do so at a lower cost that's the key thing the brands want high quality professional level creative and that's why we've invested in building quick frame AI because we consider it part of our mission is that they can produce a lot each brand can produce a lot of creative at very reasonable cost and as creative they would be proud to show their customers their friends their family and so forth on a 65 inch screen on the 65 inch screen and the the mob the AI models were or quick frame which is the technology that provides that is orchestrating multiple models across C dance cling the VO and and orchestrating these in the scenes to produce really good video our generation times are 15 minutes to generate 30 seconds so this is a lot of steps that are happening to generate higher quality video yeah we think there's so some of those so so quick frame start out as a marketplace of independent creators and we were kind of like uber for that marketplace like we would connect the creators with with the brands so now for the even when you bring down the costs people are very busy and even though it's easy they're very busy so some some of the brands now use those creators to do AI video for them and the our marketing team does that ironically like we've been putting out quite a bit of video and then we Some of that is created in-house, and some of that we use some of the creators to create, and they do it at a very reasonable cost. We increasingly also have been connecting with social media creators and connecting them with brands using AI tools.

To create a 30-second spot, not to talk about it on their internet.

Exactly, to create a 30-second spot. So what you're seeing is that this is an arena where AI is actually creating more work. So instead of people saying, oh, well, it costs a lot less, so I'll do a lot less, they're saying, no, it costs a lot less, so I actually create a lot more ads, and I'm very busy, so maybe I can find someone to help me with that. So we're creating a new marketplace of creators for quick frame AI, as well as some of the existing ones. And it's actually creating a lot of work for people.

So they make five ads, let's say. So they make five ads. Then does the software, they try to put it in, does it A-B test them? Like once you make more ads, how does that get used in the mountain ad text?

Yeah, so we have full A-B, you've said it exactly. So we have full A-B testing capabilities, meaning testing one creative versus many. And so the more creatives you create, the more you test. And this is something that's pretty common. Like, it's been done for many years, like in Instagram's ad platform. And then Google, there's, you know, the creative is tech. So it's been done for many, many years on their platform. And it's done on our platform, too. But bringing down the cost of creative, it allows a lot more testing. More testing tends to lead towards even better outcome, even better performance. And so, again, if you're going to do, you know, new creative, the average user in quick frame is creating seven ads seven projects a month and so like there was no way they were doing that you know like picking up a camera and hand editing video creating seven separate projects a month so people are definitely seeing it from that kind of data as a way to enable more and more creative and more and more testing and then those independent creators come in by being like they'll do that they'll help them with that for a very reasonable price right right right okay um makes sense i think it's funny they still offload it they just don't want to do the creator you know from the moment we started creating quick frame i'll say this is everyone is like okay so is this going to replace the creator and i said i the response from the market is going to be oh so now i have to create the videos too everyone's oh if they can delegate they're going to delegate especially the people you even with AI you still want someone has a sense of aesthetics right like specialized right when you said like sometimes you see ads that are cringy on CNN like they probably were not created by right exactly exactly yeah yeah okay okay so where you're using AI

today. Internally, do you actually, like, how are you using it and where today is new forms of AI, general AI, showing results, either on the revenue or the cost side?

Yeah, so we have a site license. Of course people can use whatever tools they want. I won't say who, I don't want to answer that debate of OpenAI versus Anthropic or anything like that. And so the whole company has access to AI tools, and then we're doing things where people are sharing how they're using them um no more like um group meetings over zoom and things like that facilitated by some of the leadership and the um and so and for different areas so this is across the company our engineering team um our sales team our marketing team and And they're learning where it's offering benefits that maybe some people didn't think about. Also, like any new technology, on an individual level, there are early adopters and there are late adopters. So the late adopters seeing the benefits the early adopters get are another way to kind of amplify the use of technology. And we're viewing it as more of a, like, it slows the growth on headcount rather than it so much it replaces. When we do see roles in a company that don't make sense anymore, obviously we're going to make changes there. But what we want to do is move quickly. We want to grow. And, you know, it ultimately, we don't want to cut our ability to execute by just viewing it as a cost-cutting. It's more about moving faster. But your revenue is decelerating. so wouldn't this imply that your revenue should be accelerating the yeah if you're not cutting costs and you want this to be a revenue accelerator but i have revenue falling well our sequential quarter over quarter revenue growth this but so meaning q2 over q1 for this year is more than twice the size it was last year so our our in terms of actual dollars added into the business our revenue is actually growing faster the we added we're adding more dollars q2 over q1 and And then, again, as we go into the second half of the year, all of this is contributing in order to, you know, why we got it higher in the second half of the year.

Well, and I guess presumably if it doesn't accelerate, you'll have to start cutting costs. Because your first choice is to accelerate revenue growth.

Well, as our revenue grows, if we hold our costs, we're cutting costs. Yeah, yeah. That I agree with.

That I agree with. We keep track, as you know, of FTE. So we're, you know, in the big, the guys that are really using the big hyperscalers. We're all cutting people as they're adopting these Gen AI tools. Yeah.

Well, I don't think we overhired as much as maybe some of them. Maybe they did.

Yeah, that's possible. Yeah, that's possible.

And we've seen, like, we took our sales operations team, and that team had a lot of what that team does has been fully automated. And so those are areas where we see, like, there's a lot of humans hands and human steps in that process now fully automated so you get two benefit benefits you cut the cost and it's actually more accurate meaning it's less mistakes less mistakes more accurate and you cut costs but on the engineering side we build a lot of technology we absolutely don't want to cut engineers but we want to amplify their contribution yes so that's That's where we view it more as let's move faster rather than, you know, like cutting people. So one in two people at Mount is on the engineering team. I don't see that changing at all. But I think as a company, we've never been producing product at a faster rate than we are right now.

But if you're producing product, we would expect to see that in the revenue line, assuming you can charge for the products and they're not free.

Everything needs product market fit. So you've got to release it, see if people use it, adjust it. So I don't want to pre-announce, but we have a lot of exciting stuff.

And are you charging for it? You're not just including it in what you have?

Well, some of it is enhancements on the targeting side. It's just included. And some of it is new things that we're working on that are being tested with customers. There's quite a bit of AI technology that we have in kind of new uses of AI models within our product. And so we'll be announcing these as we feel like they're coming out of beta and they're fully production.

But I'm trying to get at the money. So where you have new tech, like where you said you're going to be announcing, does it just upgrade your product that you have already, or is it an upsell?

Yeah, it's not an upsell, but we are investing to basically continue to grow at the rates we are or faster.

Okay, because this is one of my big takeaways from Possible in Miami was that a lot of people are doing AI work and just including it in their product, which I don't like. Like, I want money. Like, I don't want you just to keep – because what it says to me is you're discounting price, that if you have to add new stuff to keep your market share, that to me is like a pricing discount if you can't upsell.

Yeah, I mean, I anticipate our revenue is going to grow. I absolutely, you know, like, we need to see all the technology we've built in market, people paying for it before, you know, I want to talk about the financial profile of that in the future. Okay, great.

So one of the things that's really true in ad tech right now is we're getting this supply path optimizations or DSPs or disintermediating SSPs and vice versa. And so assuming that is successful, that we get these sort of open Internet take rates down from the 40% level towards the 20% level, let's say, does your take rates suffer and do you get injured by this SPO?

For Mountain, absolutely not. And the reason why is we already had private marketplace deals with pretty much every source of inventory in the industry, streaming network, fast network, and so forth. So it is the publisher's decision as to if they use an SSP and who that is. It is not our decision. We don't pay publishers. So that might show up in the pricing we get from the publishers if they choose not to use an SSP. But it doesn't at all affect our mounds economics. And so we were already kind of set up for this. I think in the, you know, you think of a more classic DSP platform with a media buyer, a programmatic media buyer, where, you know, they view their value as cutting costs and things like that. That's kind of a different scenario where, yeah, I agree with you. There's a lot of, like, are SSPs becoming DSPs and DSPs becoming SSPs. Mound is a platform for small, mid-sized businesses to drive results through performance marketing on television.

So they can stop spending the money if they're not getting performance. They don't care what's happening with these fees over now.

Yeah, it's not a topic that they're aware of, and it's kind of just they're insulated from. And then Mound, we have direct relationships, so we're somewhat insulated from it also.

On the supply side.

On the supply side. I'll give you a very specific example. The NBCUniversal owns Freewheel. If they want us to buy through Freewheel, execute those buys through Freewheel, great. If they say we don't need Freewheel anymore, great. It doesn't change any aspect of our relationship with that specific publisher. That's their decision, and we're happy to buy wherever we get the fastest execution.

The highest return on ad spend. Because your publisher spends more if you get five times return on the dollar versus if you get a two times return on the dollar.

Now, if you're a big agency, your role with a big advertiser is to cut costs wherever you can find it. It's a different relationship. Yeah, that is.

Questions from the audience for Mountain? Okay. So is your view that in the near term open Internet benefits from generative AI disruption versus walled gardens versus open internet, who wins and loses?

I think open internet benefits, and the reason why... Now, let me step back a second. In order to benefit, you have to embrace change. Okay, so if you're a company that is not tech-forward, that is not change-forward, that is trying to protect your position rather than expand your position, then you know that that's a choice you're you're operating out of fear and that's a choice you're making if you're a company that's leaned in to being a market leader then it's entirely different scenario the in that scenario where your market leader you benefit why do you benefit because AI is democratizing the models across the industry like the targeting models now can leverage AI technology, LLMs. I'll give you a very concrete example. When we serve ads, we want to serve ads to specific consumers that are going to respond to a specific product, right? So example, the Onewheel. They sell single-wheel skateboard. Looks really cool. You see it in Manhattan, but it also looks very dangerous, right?

Yeah, it does. Looks really dangerous.

Okay, so how do you find the consumer that's not only going to like seeing someone on it, but want to spend almost $2,000 to risk their life on it, right? So the way you do that is you apply, first you need an LLM in order to say, who would buy this? What's the profile of this consumer? And not like age, gender, geo. This is an adventure seeker. This is someone that probably owns a surfboard if they live in California. They might go helisking in Canada. Because they have to have a lot of money, too. So you build a profile using LM. You couldn't do that two years ago, right? And then you apply machine learning models. Once you have that profile, then you take large pools of data and you run those models to find those individuals using machine learning models. And then you use our relationships with the networks to deliver that ad directly in that person's home. The first two steps two years ago, that was a very different technology stack than it is today. So that's an example of how AI is leveling the playing field because Google, people are just literally typing in what they're interested in. Meta, obviously, I think they have something like 30,000 engineers, but now AI technology is leveling the playing field in terms of targeting. it's laying the playing field and just apply that example across the entire tech stack for advertising and so but to be that winner you have to be heavily invested in technology and that part of it is obviously self-serving that's why one in two people at mountain on the engineering team and me personally i'm a coding engineer i don't get to code unfortunately as much but i certainly can hold up but everybody's saying that i have a lot of ceos like the reddit ceo the tabula ceo I know I'm supposed to be leading people, but I spend four hours a day on Claude,

and it's the most exciting. I'm so happy to be back learning new things in code.

I'm not spending four hours a day on Claude, but I'm definitely in four hours of conversations about our products and technology.

I think a lot of people are software.

Obviously, like everyone else, I use a lot of AI technology, but I do wish my retirement job many, many years from now will be coding. A lot of CEOs love the coding stuff. yeah um okay but in the so in the near term you think open ai bet you think the open internet benefits from because walled gardens are cutting people as they move towards these llms and they're accelerating revenue growth because they have cloud a lot of them have clouds so that's helping i also think brands want just like investors want a diversified portfolio they want a diversified marketing portfolio right okay the the it's true the you know mark zuckerberg's vision of like just give us all your money and we'll you know you could just be a drop shipper and we'll do everything else i i don't think builds brands and and i don't think is how people view it um they i think they view it as we understand the customer and we leverage all this technology to lower the cost to reach that consumer and then connect with them yeah okay well i definitely see the niche aspect of what you do for the $2,000 skateboard.

I totally see how that is a value added because there might be, in the world, there might be 3,000 guys. But if you sold even 500 of these, that's an awesome return on investment.

They sell 10,000 a year.

Do they really? There's 10,000 lunatics a year.

I own one.

You own one? Is it as dangerous as it looks? It is as dangerous as it looks. I thought wheelies were dangerous.

I don't get to ride it very often. My fiancée is kind of like not as risk averse.

She's got you on hourly. Okay, you could do two hours a month. That's as much risk as I'm willing to partner with. As ad buyers demand more transparency and large publishers gain negotiating leverage, how would you think about your take rates over time?

I think they're very stable. For us in particular, our customers, the way they view Mound's platform is, Let's say they had $10,000 they could spend on social, $10,000 they could spend on search, or $10,000 they could spend on performance TV. And none of those conversations is take rate, additional fees, anything a part of it. It is just money allocated to drive a revenue.

And how much revenue lift do I get? So performance TV is the highest, which it should be, because they've spent the least there. If 95% have never spent a dollar, then as long as it performs, all this money, in theory, can move. and it isn't a take rate.

That's not part of the conversation. In the enterprise side of the market of agencies, that's a very different scenario.

Okay, but that's not a very big part of your business, right? That's a zero part of our business.

We are small and mid-sized.

So it's all performance, 100% performance. Okay, that makes sense. Okay, any other questions? Okay, I'm going to call it there because we have eight seconds left. We're right on time. Thank you very much.

Thank you very much.