All right, good morning. Welcome to day two of the KeyBanc Emerging Technologies Summit. I'm Justin Patterson, I lead The Consumer Internet team. Really excited to have the DoubleVerify team with us here this morning.
Great. Thanks for having us, Justin.
Yeah, thank you. So Mark, maybe to just kick things off, for people who might not be familiar with DoubleVerify, provide a quick overview of just what you do, what problem you solve for the advertising community.
Yeah, it's a great question. For those of you who are new to the story, DoubleVerify is a software platform that enables advertisers to ensure that the ads they buy across a publisher, across a social network, across connected television, across a mobile app, is brand safe, viewable by a human, and delivered in a fraud-free environment. So basically, we ensure that a transaction between a buyer and a seller of digital media is safe and that actually can drive an outcome because a real person's on the other end.
Got it. Let's just dive right into it. There's a bit of an elephant in the room for the past week. I'll just ask it point blank. Is there a pricing war in the market? What's going on with pricing?
I never thought you were going to ask that question, Justin. Never. It was not even on my mind.
I know. You're throwing it back at a loss for words.
Yes. Let me be super clear. The answer is absolutely not. We've been in this space with the same competitive set roughly for almost a decade. And pricing has always been one part of a long list of dynamics that either win or engage a client or retain a client. But those dynamics haven't changed. If anything, the only thing that's changed is the nature of the competition, which is we have continued to build out a highly differentiated, very unique tool set which our competitors are struggling to keep up with.
Got it. So let's have Nicola unpack that a little bit more. It sounds like there's no price war, yet when we step back, we look at Q1 guidance, there's a bit of a deceleration. So could you just talk through some of those dynamics that are occurring within the first quarter and why you still feel good about a second-half reacceleration?
Sure. So we'll talk about the first quarter first. What we've seen is a reduced level of spends for some of our large clients. It's isolated to a few clients, so we don't feel that this is a macro issue at all. But based on what's happening endemically into each of these clients, they've started their ad spend levels lower in the first half of the quarter. We're obviously in conversations with them. We have a lot more clients than the ones that are starting slowly. So the rest of the year sees an assumption that that spend will come back and that new clients will ramp. We've signed very large enterprise deals, including Pepsi and Haleon. Those are ramping slowly in the first quarter, but we anticipate them to reach the levels that we expect from those clients in the second half of the year.
There's a little bit of a shift towards the second half of the year in terms of a mix of the revenue split, but it's really because Q1 is low.
Got it. We're thinking about just that degree of confidence in the second half. What's really providing that? Is it that new customer side? Are you seeing some of those ones that sort of slow, ramp back up? Let's put a finer point on that.
Yeah. So it's both. So we are anticipating not every single client will come back, right? Some have real issues internally where the ad spend will probably not come back to the levels that we've seen before. But we've accounted for that as a group. We expect the spend to come back. The confidence we have in the second half is the theory around our guidance is always core on core. So whatever is baked into the assumptions for our guidance doesn't assume acceleration around some of the core drivers that we've been discussing all along, including brand safety, brand suitability on Meta, accelerating faster than kind of a slow ramp that we've assumed. TikTok continuing to grow extremely fast now that we have more than just English language, which is all that we saw in 2023.
Scibids uptick being beyond what we've called out, which was $15 million-$17 million. So there's a lot of accelerators to the core on core assumptions that are in the guidance that are not in that number.
Got it. So you threw out the M word, so I'm going to throw that over to Mark here. Meta, short-form video, big theme across the entire industry this year, just launched in Q1. How are you thinking about just both short-form video as a whole and then Meta scaling up over the course of the year?
Yeah. So it's a really exciting development for us. We started digging into short-form with TikTok a couple of years ago, expanded that to Reels and YouTube Shorts last year. And now when you look at social engagement, almost 70% of social engagement now is on video. So the idea of providing brand safety and suitability, viewability, invalid traffic verification across that is a huge opportunity for us. We saw that with TikTok last year where our business from the beginning of the year to the end of the year tripled across TikTok. We've seen that with the uptick that we got when we launched across Shorts and Reels. And I think that will roll into a broader Meta opportunity that we have ahead of us. I mean, Meta was around $40 million of our revenue last year.
We think there's significant upside against that $40 million based on the addition of Reels and Instagram, as well as brand safety and suitability on the news feed itself, which we've only penetrated about half of our top 100 customers with that solution. So really nice upside there. And I think social continues to be where a significant amount of new dollars and ad dollars are heading despite some of the challenges. And one of the things that we are doing by providing brand safety measurement there is actually calming some of those fears and giving people confidence to be able to spend across social.
For sure. I actually think it's a really interesting time. It might even explain some of the Q1 dynamic. If you step back, you're just seeing a lot more of these CPGs really try to figure out creator marketing as a whole. I mean, I'm pretty sure most people were not thinking about MrBeast or Logan Paul PRIME being some of the fastest-growing CPG brands, yet that's the world we live in. More fun words that we get to say as internet analysts. But when you talk to CPGs, just how are they approaching social right now? Because there is that brand safety concern, and it does feel like this could be a pretty meaningful uplift for the next few years for you.
Yeah. It's a different world. I mean, we work with folks like Unilever, Mondelēz, Haleon, Colgate. I mean, these are some of the biggest consumer brands in the world. And they're companies that were used to buying TV to reach stay-at-home moms to sell them CPG products. That world is over, right? The TV world, as we know, is gone. And even CTV is becoming incredibly stratified to we can't reach that mass audience anymore. Where they can drive engagement is through social and specifically through video on social. They can demo products. They can show how real-life people are using their solutions to great, wonderful outcomes. So it's not just the idea of influencer marketing. It's also the idea of where do we need to be in a stream to actually influence purchase of our products.
And the wonderful thing about social, there's social video, and there's lots of terrible things about it too. But one of the great things about it is it curates a stream of videos. And if someone's really into cooking or really into home improvement projects, they're going to get a ton of videos there. And then the ads that run in that stream are going to be absolutely relevant to that user.
For sure. It also feeds into volume, which you're a volume-based model so far, so it seems like a healthy tailwind.
Yeah. Yeah. We've talked about CTV in the past, and we love CTV. We're engaged with the top 10 streaming platforms out there. When Netflix launched, they launched with us as a verification provider. But if you look at the load, the ad load in an hour-long session of, for example, Netflix is about 16 30-second spots. Look at the ad load in a one-hour TikTok session, and you're looking at 10-100 times that many spots that can go through that. So based on a volume-based model for us, short-form video is where it's all happening.
For sure. But to your point, CTV does still matter to the long term.
100%.
We have been talking about it for years. That does come from your background at Telaria previously. When you kind of look at the landscape today, where do you think we are on just that tipping point of the CTV opportunity opening up more for you?
Yeah. I think we noticed on our call or mentioned on our call, I think about 5% of our revenue and volume comes from CTV right now, which is pretty small considering the size of that market. There's a couple of considerations. I mean, we're a global company. CTV is not yet a global phenomenon. It is really a U.S. phenomenon when it comes to actually stealing a lot of linear dollars. So that's one thing to understand, which is CTV is huge. It's huge here. It's growing around the world, but I think the opportunity is still relatively here. Where we think that unlock is so right now for us on CTV, we focus a lot on two things: viewability and invalid traffic. There's a ton of fraud in the CTV universe. We just announced in partnership with Roku last week, we discovered a CTV fraud scheme called CycloneBot.
So we focus on that, and we focus on viewability. A lot of CTV, though, is still bought, not open market. It's bought direct buys. It's bought in what's called PMPs. Those are relatively protected from a lot of invalid traffic or fraud attempts. Viewability is still an issue. So we look at things. We have something called Fully On-Screen that relates to the fact that a significant percentage of ads are running while the screen is actually off. We found almost 25% of apps were not reporting back whether or not the screen was on or off. That's not viewable, right? So there's an assumption that all CTV ads are viewable. It's not. So MetNet, we looked at those things. Where we think the big unlock is, like we saw on TikTok, like we're seeing in social video across the board, is brand safety.
To date, publishers have been loathed to release program-level data into the ecosystem. We think that's changing because once you release program-level data into the ecosystem, we can provide brand safety and suitability on a program level, not on an app level or a genre level. I think that now puts CTV on par in transparency with short-form video, with social, and everything else. And if you see that there's one complaint from ad buyers today around CTV is lack of transparency in buying. They buy a genre across Netflix. They buy a channel across Tubi, but they don't get to buy show by show. At least if we can provide brand safety and suitability on a show-by-show level, it provides more confidence to buy through when there's those blind buys.
For sure. I agree that the momentum is changing there, and Netflix is now doing its semi-annual report of viewership across the platform. Slowly but surely, we're moving in that direction.
For sure.
So maybe just one more big picture TAM type question for you, Mark. Retail media, the other big secular growth trend, how does something like DV fit into that side of the world?
Yeah. We saw a really nice growth last year in our retail media business, about 60% growth year-over-year. Still relatively small as a percentage of our revenue, but it hits all three areas of our business. We have three ways we report revenue and kind of three main chunks. The first is activation, which is we help advertisers in a pre-bid fashion to filter out what they're going to buy before they buy it, so ensure that it's brand safe, viewable, fraud-free, etc. And then we have measurement, which is post-bid, in which we can block an ad that's already been bought if we feel it's not an appropriate environment. And then we have a platform business in which we provide some level of those services to platforms to filter their inventory before they even push it out. Retail media touches all three of those areas.
We work with folks like Macy's, Best Buy, Walgreens, Target, Amazon, a significant number of retail media networks on all three of those. So for some of them, we're providing a base level of brand safety and invalid traffic protection. For others who have their own DSP implementations, we actually integrate our data directly in there so that when advertisers do buys, and retail media networks are really the one thing that it's important to understand, retail media networks are just that. They're networks. Whether it's Amazon's or Target's or Macy's, they use data from their own sites, but really most of the media is bought off-site, right? There's only so many users that go to the Target website. Where you're buying is users that go to the New York Times or someplace else. So those buys need to be protected.
We insert our data into the buying process so they can filter out things that aren't brand safe, not viewable, etc. We see it as a growing area. It experiences the same challenges as every other media buying platform. We've got great partnerships in that space. It continues to grow. The nice part about it, similar to CTV dollars that have come in from linear that we never would have touched, a lot of retail media dollars are coming in from non-advertising sources that we would never have been able to verify, whether it's co-op dollars or in-store dollars, things like that. They're coming into retail media and digital, and that provides just more fuel for our transaction fire.
Got it. So shifting away from some of those big secular drivers, maybe a little more tactical industry, you're going to be shocked by this one. Third-party cookies get replicated this year. So you're not necessarily a cookie-dependent company in there, but you still deal with a lot of publishers at the end of the day. So when you think about signal loss, what potential opportunities might this create for you?
It's a great question. I appreciate the T-shirt. I was going to wear my Lehman Brothers T-shirt as a contra to you today, the cookie T-shirt, so. But I think when you think about cookies, we've always said it's either neutral to positive for us as a business. We don't rely on cookies to deliver our service. We don't rely on PII or individual-based information. We look at the context of a page. We look at what happened in a transaction, but we never really look at the user. That aside, I think when you lose the ability to target a user based on tracking them with a cookie, you have to look for other proxies. And those proxies right now include things like context, which is what is happening on a page, which we have tons of, and things like attention.
We launched an attention metric almost two years ago now. Continues to gain steam. I think advertisers are going to be looking for ways to connect some level of measurement of an ad and the effectiveness of that ad to some outcome, right? They've always used the ability to target a person on demographics or the fact that they went to another site as a proxy for that performance because they could tie that 1-to-1. Those will be gone with cookies. The ability to target users will be gone at scale with cookies. I think things like context, attention, areas where we've invested in over the last few years will become more important.
Again, as Nicola says, we don't bank these into some trajectory of growth, but I think there'll be nice icing around the cake as our contextual business grows and as our attention business grows.
Yeah. It's funny you bring up attention. I feel like I should do something to measure attention in the audience right now, but I need a little more caffeine to do that. So in the interest of time, let's talk about just the future state of the business. You had the Scibids acquisition last year. I think there was some initial investor confusion around that. How does Scibids build upon what you've already built with ABS?
Yeah. So if you look at the core fundamentals of our activation business, it's always been about trying to help an advertiser avoid certain types of content that aren't going to work for them, right? Get garbage out of the system, and what's left will work better. We started that with what we call, let's say, basic brand safety, which was a very stiff yes or no based on very specific criteria. Was this safe or not safe, right? ABS was the evolution of that. So we said we went from static to dynamic, and ABS allowed us to say, "Okay, let's create brand suitability tiers," so not just brand safety but suitability on a customer-by-customer basis, on a campaign-by-campaign basis. So it became dynamic, right? And then it learned with our measurement to say, "These are things that worked. It didn't work.
Let's feed it back in." So you think about the evolution of brand safety and pre-bid. It went from static to dynamic, and now Scibids allows us to take data points in and make them algorithmically defined. So something like viewability, which was seen as a black-and-white deciding factor, viewable, not viewable. If it's not viewable, don't bid on it, right? We can now use a tool like Scibids to say, "Well, okay, let's move this on a spectrum. Let's be dynamic and say maybe we'll accept what's considered a non-viewable ad if it's actually driving a KPI on the other side," right? So let's optimize our bid based on driving this outcome.
So it's just the actual natural evolution of what we've been doing, which is saying yes or no to a bid, but doing so in a way that becomes much more fluid, much more dynamic, and driven by a KPI on the other side. So it leverages our data to do that, and it's another way that we take core data points on brand safety, on suitability, on context, on viewability, on attention, and leverage those in a way, in a pre-bid fashion, to help optimize an advertiser's spend and help them drive more spend through DSPs.
Got it. So Nicola, could you put some financial context around that? So how big is ABS today? How has that been growing? And when you look at Scibids scaling up the next few years, how are you thinking about that as a contributor?
Right. So you mentioned ABS, which is our premium-priced product. It is now five years old, and last quarter, it grew over 40%. What's great about that product is that two-thirds of that growth came from existing customers using the product on more and more of their impressions, right? So we continue to see adoption of the product over more and more impressions. That's now a product that's last year, it was about $180 million, a real success for us. The reason why it's a success is because it's differentiating the market. It really isn't something that looks like ABS. And it's a success because it ties the measurement data into the pre-bid data, right, seamlessly for our advertisers. So it's obviously been a great success.
Scibids, when we announced the acquisition last summer, we said that we thought it had the potential to be a $100 million product in the next five years. So closely following ABS sort of in the trajectory, the initial response that we've gotten from new clients who want to try it is very high. 40% of them want to try it. And so I think we're off to a good start with the product. And similar to ABS, Scibids is very differentiating the market.
Got it. One more for me before I open it up to the audience. Let's talk about margins. You've been very consistent at 30% margin since you've been a public company. When you look at just reinvestment priorities right now, what are you leaning into? And as you get past that, how can we think about the margin opportunity for the business down the road?
Yeah. So our business is fundamentally very profitable, right? Our gross margins are over 80%. We've invested in that line over the last few years to maintain that level of gross margin, even though we're now doing a lot more short-form video content verification, which is a lot more volume to verify. But we've been able to do it in a way that we maintain a pretty nice industry-leading gross margin. The investments that we're making are in R&D, right? We've moved away. We're now three years into our life as a public company. The growth in G&A is diminished at this point. And then sales and marketing, the investments that we're making now are opportunistic in certain regions where we don't yet have people on the ground. So it's all R&D.
And within R&D, it's data scientists tied to the acquisition that we made in Scibids and try to figure out exactly how we're going to integrate that into what we do. There's obviously a cost component to AI and investments that increase your margins. We are still investing into the business. And the reason we continue to do that is if you look at our top line, we're growing much faster than our peers, right? And so to the extent that that continues, we're going to continue to do what we're doing. Over time, we are a very profitable business, right? So you could see that margin grow for sure. But to the extent that we're growing revenue, we grew 29% in the fourth quarter. We don't see a reason to stop investment.
Got it. Great to hear. All right. I think we have time for one or two from the audience if there's any takers.
How has your program and what are driving the brand advertisers on the front bucket of Meta and then our direct response of almost advertisers? How do you think about that? What about the actual advertisers on Meta? What do you think you can make a difference with this?
Yeah. It's a great question. So we already have a very strong footprint of brand advertisers who advertise across Meta. As we noted, 50 of our top 100 are using some of our solutions across the Meta landscape. So when we look at that opportunity, we know when we look at platforms like YouTube, for example, over 90% of our top 100 customers are using us for brand safety measurement on YouTube. We think we can see that level of penetration on Facebook as well. It's the same customer set, same group of folks. They're advertising. We know they're advertising on Facebook. They just haven't implemented our solutions there yet. So when we look at long-term growth trajectory, yes, we focus pretty heavily on brand advertisers. Those brand advertisers are already on Meta. They just aren't working with us yet.
So there's a large opportunity, I think, for us to continue to penetrate that group. Plus, outside of our top 100, the next 500 customers as well, of which a good portion of them are also advertising on Meta. So I think the opportunity is one in which we've said is multiple is larger than it is today. It'll take time. We have upselling to do. We still have additional volume and growth to go after. And the iteration of the product that we've launched in brand safety on Meta is just iteration one, right? That's important to note. Any product that we launch across a walled garden usually starts with a basic version and then ends up in a much more granular, detailed version down the road as we move into areas like brand suitability, more granularity on reporting back on suitability violations, etc.
All right. No one's taking the last one. Capital allocation. You've got a growing cash balance. How are you thinking about deploying that?
We've been successful with M&A. The way we've been successful at that is we've been able to acquire technologies or groups that have allowed us to accelerate our roadmap, which leads to the industry-leading growth on the revenue top line. That's our first priority. We have enough cash to also invest through R&D and operations. Those will remain our first priorities for sure for 2024 for sure.
Awesome.
We have enough cash to also invest through R&D and operations. So those will remain our first priorities for sure for 2024 for sure.
Awesome. With that, we are out of time. Mark, Nicola, pleasure as always.
Thank you.
Technologies or groups that have allowed us to accelerate our roadmap, which leads to the industry-leading growth on the revenue top line. So that's our first priority. We have enough cash to also invest through R&D and operations. So those will remain our first priorities for sure for 2024 for sure.
Awesome. With that, we are out of time. Mark, Nicola, pleasure as always.
Thank you.
Thank you.