DoubleVerify Holdings, Inc. (DV) Earnings Call Transcript & Summary

Yes. And I think it's also important to note is that we're still really underpenetrated, right? When you look at the global market. So there's -- I don't want to -- I hate to use word landgrab because it sounds like it's like chaotic. But there is a brand grab going on right now, right, particularly outside the U.S. So for us to lean into these markets. And it's not a ton of sales and marketing people, but we want more. This is the time to do it, right? Because we know average customer tenure for us of our top 50 customers is over 7 years. When we land a customer, they stick with us. And we've seen this investment pay off for us over time. 3 years ago, our #1 competitor was the same size we are. In 3 years, we're now $100 million bigger.

Well, you're winning 80% of new business, so like you have to go fly.

So why would we stop like leaning in and investing now when we lock these guys in, they become customers for life, and we continue that momentum. Like this is the time to do it, and that's what we're investing in R&D to make sure our technology is ahead of the marketplace and sales and marketing at some level to make sure that we're at the places where our brands are going so we can build that relationship with them.

Okay. Well, I don't really mind costs that are linked to revenue. Like if you're talking to sales and marketing, like I'm going to back off because if the guys don't perform, you're going to fire them, right? But when you talk about stuff like R&D, and I do worry, Mark, that this pivot that you talked about strategically and your goals from insurance, where you are 1 of 2 and you're winning 80%, which means your product is just a lot better compared to your pricing at $0.08, which I'm always giving you trouble about how low your price is. But the point is that is a too competitive market. The minute you go to performance, you're competing with a lot more guys and bigger guys and guys that have been there longer. So to me, that feels like lower margins. And now Nicola is telling me, I'm going to spend a lot of money. So please reassure me that what we're not about to do is take the returns on invested capital down because you're going into a more competitive segment?

No, we are able to manage to a margin that we feel is comfortable for the business. The investments we're making -- I'll give you an example. We invested early into the ABS product, right? We were able to launch it.

Tell them what that is. I know what that is.

Authentic brand safety and suitability is a product that allows our data that's collected on the measurement side to be used on the prebid side of the business. It creates a loop for the data that the advertisers are using for their entire spend, whether it's on the measurement side or on the pre-bid side. That product, we were able to launch well ahead of our competitor, and that was tied to investments that we weren't able to make in the product. So I think when we talk about moving to performance, these are products that are still tied to the data that we collect, so we're not sending people outside to do some research somewhere.

I see. And so when you make the data have another revenue strategy.

Exactly.

Okay. That's fair. Okay. That actually is helpful action. Let's go to Scibids. So I would say there was a sort of a misunderstanding when you bought Scibids. That what you've just done has gotten into direct competition with Google and Trade Desk. And then you did an all-day analyst meeting and 2 of the very high-profile people you either put on stage or were in the audience in a very profile way were Google and Trade Desk. So basically, it became clear that they actually think they benefit from the Scibids tech. So my question is, is -- this is the -- 1 of your goals for next year, is this new helping not only companies ensure there, but getting better performance. Talk about how you think -- I'm going to focus on the economics, but why don't you talk about strategically across the business, how you think Scibids helps overall DoubleVerify increase its market share and drive economics, which is what I actually care about.

Yes. For sure. I mean, Scibids is a really exciting acquisition for us, and it fits perfectly into our long-term strategic goals. I don't want to go too far in the weeds here, but if you think about what we do and what we call our activation business, which is over 50% of our revenue. Activation, as Nicola noted, is the ability to filter out bad behavior, bad stuff before someone even buys it. And we do that on the prebid side through platforms like Trade Desk and Xandr and Google and all the big DSPs, Amazon, right? . So -- and we've always done that in a very binary way. We started off with a brand safety filter, right? Good or bad, don't bid on this impression. That became advanced into ABS, which we launched in ABS with authentic brand suitability gave a little bit more flexibility. It allowed for dynamic campaign by campaign criteria. It took data from measurement and fed it into the tool set. And now ABS is by far our biggest product. We'll do -- how much will ABS generate this year?

Something around $117 million.

Over $100 million, still growing at 40% plus. I mean this is a product that was launched over 5 years ago. Going at 40%, net-net, ABS is a performance product, right? How this relates to Scibids is ABS is a performance product, but still kind of binary. It still says good or bad.

Okay. Says yes or no. Yes, we're not going to look at this [indiscernible].

[indiscernible] What Scibids does is it works with the DSP, and this is where the complementary aspect of this to a Trade Desk or a Google is. It works with the DSP to not say yes or no, but maybe.

Okay. .

Maybe you should bid on this impression based on the data that DV has, if it's cheap enough, and if it meets your outcome goals. So now we've moved from static in our brand safety products to dynamic with ABS to actually algorithmic optimization with Scibids. It's just an advancement of what we already do. It plugs into the same platforms that ABS does, but provides an advertiser the ability to look at a maybe answer. And when it comes to things like viewability, right, viewability is something that the industry has defined as this is viewable or not viewable. People don't buy viewable. But what if it's kind of viewable. It's right on the edge. It doesn't meet the 2-second criteria for video, but maybe it ran long enough that someone actually can click on it. And it's so cheap that it's worth me buying it.

Okay, so it adds value to the equation, sort of yes, no, it adds value.

It's all ROI. So that becomes critically important to that basket of goods. So think of it's an advancement of what we already do. And then the second advantage of Scibids is it adds to our basket of goods. And this is critically important when we talk about competitive -- the competitive nature of the space we're in. We, like other software businesses have moved from point solutions to platforms. And advertisers don't want to work with 15 different point solutions that do 15 things. They want a broad basket of goods. . Scibids adds to that basket of goods, gives us a complete solution now from binary brand safety to measurement on the other end, prebid to postbid, gives us a unique set of solutions that our competitors and no 1 in the space has. And so, a, that gets a sticker to current customers, but gives advertisers another entry point into DV. We have advertisers that work with our competitors for measurements are coming to us on performance tools because our competitors don't have anything comparable. So it gets us in conversations with advertisers who said, "Well, we're kind of -- we're a blah, blah, blah shop. We're good with measurement right now with another competitive platform, but they don't have anything like this. Maybe we should be talking to you." So it fits into the strategy of protection to performance, but it also fits into our strategy of acquiring a bigger market share by having customers start to work with us with 1 product and then expanding them over time into all of our products.

So is this new product for performance mostly focused on the customers you have? Or do you see it as a nose under the tent to get new customers that you can then backward sell your insurance -- what I call insurance product, which 1 is it?

A nose under the tent.

It's a nose under the tent. So it's more insidious.

Nothing as an opportunity. Well, it's not insidious. It's a product that's not available...

You say insidious, we say it's good business strategy.

Yes. That's what I meant. I see it as a positive.

But it also goes back to 1 of the goals that Mark mentioned in the beginning, right, which is client-focused innovation, right? Because this -- what Scibids allows the customer to do is use highly customized data sets to put into Scibids along with other measurements that we're providing, right? So customers will have their own data set that they want to maximize against. They can put it against our own data and then create a perfect bid, right, through Scibids.

So we're using first-party data with this to create new customer demand that we can backward sell into the insurance product, which is [indiscernible]. Questions. I got 100 people in this room. Who has a question for these gentlemen before I go on. Oh, good. Okay. Generative AI. So generative AI is something you guys are using really extensively. So I want them to hear about that. Tell us how you're using generative AI today and how you think it -- what -- I know Nicola, what I liked best is what it does to your cost structure over time. Getting faster to market. But talk about generative AI.

Yes. So Nicola already mentioned this, is that we're using it. If you think about what DV does at our core, we're just a big decision engine, right? We look at stuff and we make a call on it. Is this viewable? Is this brand safe? Is this fraud? We classify content, we classify transactions. What AI does is it allows us to make those decisions faster, more efficiently and with less humans involved. So we talked about becoming more global. Becoming more global means we need to understand languages. Because when we look at the context of a page in a foreign language, we need to understand what a word means in French versus what it means in Portuguese versus what it means in English, right? Because the context of that page could be very different. We had translators do that in the past. We had human translators. Now we're using AI to actually do translation for us. When we look at video, so short-form video. We're identifying -- when we look at a short-form video, we look at the video itself. We look at the text that goes with the video. We listen to the audio that goes with the video, and we look at the metadata around that video, all in 1 20-second TikTok spot, we need to look at all of those things, right? It's a lot to say whether or not this is brand suitable or brand safe, because in many cases, the video has nothing to do with the audio. So if you ever watch a TikTok video, there's crazy audio that goes over video, you're like, what is this about. So you have to look at all those separately. What AI allows us to do, as Nicola noted, on the video side, rather than look at every frame, we can use predictive AI to determine what's going to happen in the next frame. So it cuts our costs rather than looking at every frame of the video, which is pretty expensive, right? Even people in the space have said, our costs are going to go up because we're going to do more video. We look at it as now, we can look at this more efficiently because we're only going to look at a portion of that video because we can predict what's going to happen next. That's AI at work doing that as well.

Well, what every CEO has said is generative AI makes more crappy content. And so now you have more demand for insurance.

That's the other part.

They're saying 10x -- they're saying 10x this year.

Yes, it helps us save money, but it creates so much demand. We launched an MFA product right? As made for advertising content, right? Made for advertising content went from a bunch of college kids typing out really bad articles about the Brady Bunch that they jammed out through a few websites to literally thousands of pieces of content a day across tens of thousands of websites that's clogging up the digital ecosystem. Advertisers don't want to be around that stuff, and that's all generative AI doing that. So it creates a demand for tools like our MFA tool that creates, it amplifies misinformation, disinformation, and hate speech which is terrible.

But that creates demand for your core product.

Creates demand for core product and we're running into an election year in which we run an election task force, which we launched in 2020 and then rebooted in 2022. And now we're just keeping it on all the time. And that election task force looks at the incidence of hate speech, right? The incidence of things that occur around political discussions. In 2022, we saw over a 20% increase in hate speech in November -- October, November, along with the elections.

It's got to be the war. It's got to be the Hamas and Israel war.

And we saw in competitive markets, in markets where there was a competitive Senate or governor's race in '22, political, we saw over a 30% increase in hate speech. So we know 2024 is going to be a banter year for really bad stuff.

Well, and today, GroupM raised their estimate again to $17 billion of political ads. Every 3 months, they took it up, it started at $12 billion, then it was $15 billion, now it's $17 billion of political. So how much could that benefit the P&L of DV this year, the political?

So we don't directly benefit from political because we don't sell to politicians because, a, they're fly by night. B, they [indiscernible].

But the hate speech product doesn't sell?

Not to politicians. It sells to advertise.

But that's what I mean. I'm going to call that political derived. So that's what I'm trying. Is there incremental revenue to you because there's all this political hate speech this year?

I think there's incremental demand.

Awareness and demand.

Yes. It's hard to kind of say how much more freaked out our advertisers are going to be because they're freaked out every 3 months, right? Something always occurs.

There's a new war, election is just another thing.

It just creates stickiness, let's put it this way because it creates greater need for, hey, I definitely need insurance.

Stickiness for insurance. Okay. Cookie deprecation is going to be the big issue in ad tech this year. So let's talk about that. So Google has already started deprecating, I think, January or maybe it's next week, January 24 and they're going to be fully deprecative cookies. Tell me how that affects targeting the ad tech ecosystem and DoubleVerify specifically.

So I think -- someone asked me the other day, like what was CES like this year? And I was like, "Oh, people were more positive. They thought versus last year, where everyone was like worried about a recession." But like I think the other thing about CES was the actual cookie deprecation like -- storm had actually hit, people were like, they saw the Google privacies. Yes, they're like this is actually happening, like Google had like a privacy sandbox meeting room, like they had an open suite just for privacy sandbox, right? So like -- it's actually happening.

By the way, it started in 2019 Mark, and they keep putting it off. So it's not 2024..

It's kind like 1 of those things like, oh, my kid is finally going to leave home. They finally left the door. Like I've been waiting for this well, so it actually happened. And I think that for us, it's important to note that we don't traffic in cookies, like our measurement doesn't use cookies. Our analysis doesn't use cookies, because we look at the what, the how and the where, not the who. Cookies are always tied to identity and the who around it. So I think it's going to create challenges for folks who have not built some type of privacy safe proxy for an individual, who've not leaned into things like contextual advertising and other areas like that. And I think that for us, if anything, it's neutral to positive because we do have some contextual solutions. We look at context with a -- we have a contextual targeting tool. It's not a big revenue driver for us, but I think it just creates more interest in those things. It also creates more interest in non-user or non-individual-based proxies like attention.

Well, and the other thing I was going to say is I think it is possible that money moves from open Internet to Walled Gardens and you measure both, you're sort of hedged in a year where there's going to be a lot of uncertainty about what happens to demand and targeting, pricing and ad spending in the open Internet, you're sort of indifferent if the money moves, you benefit $0.08 per 1,000 in both places. So I would say that's a key thing. I mean I don't know what's going to happen, but I'm going to go back to questions. Come on. Somebody must have a question that's smarter than mine. Come on. Come on. One question.

They thought they were in the [indiscernible] room.

Yes, exactly. That's why everyone's here. That is such a big [indiscernible]. EU regs. You just said this will be our last one. EU regs, you said that you think international growth is 1 of your primary goals. I wrote it down, and I will hold you to that next year. The EU is about to lock down with the Digital Services Act. So it's going to make targeting and return on capital and attribution much harder in the EU. By the way, they're also going to regulate generative AI. So talk about as 1 of your 3 key goals for the next 12 months how the EU expansion actually can happen with all of this EU regulation that's coming?

I mean I think, look, the growth that we're getting internationally, just to be clear, is advertisers that we haven't really spoken to before. So it's -- there is expansion from existing advertisers, right? But there's also just the fact that we're on the ground, we're able to participate in RFPs and we're going after greenfield opportunities as much as takeaways. And that's much more pertinent for outside of the U.S., right? Greenfield opportunities there's a lot more outside of the U.S. than they are in the U.S. So whatever the environment is, I think just the organic growth from us just being there present and winning RFPs is really going to help drive the growth internationally. That's kind of how we see it. And that's part of that 80% win rate, right? That's kind of how we get there.

And the regulatory environment in Europe has never been a challenge for us because, again, we don't traffic in individual data, we don't have traffic in -- really in targeting. If anything, a lot of these privacy regs and new regs tend to favor Walled Gardens as you've noted down. And our relationship with the Walled Gardens is very strong and will continue to grow. And I think it goes back to that second thing I said, we want to make sure that we expand our presence across platforms because wherever dollars go from an advertising perspective, we want to be there and verify it so that we don't get concerned about a shift from 1 platform to another or from 1 country to another. And I think that's where we want to be.

Okay. I'm going to call it there. Thank you very much, everybody, for coming. We're right on time.