DV - NYSE

Greetings. And welcome to the DoubleVerify Second Quarter 2023 Financial Results Conference Call. At this time, all participants are in a listen-only mode. A brief question-and-answer session will follow the formal presentation. [Operator Instructions] As a reminder, this conference is being recorded. It is now my pleasure to introduce your host, Tejal Engman, Senior Vice President of Investor Relations. Thank you. You may begin.

Good afternoon. And welcome to DoubleVerify’s second quarter 2023 earnings conference call. With us today are Mark

Thanks, Tejal, and thank you all for joining us today. I am excited to discuss our second quarter results and to share the significant progress we have made across multiple fronts, from product innovation and channel expansion to new enterprise logo wins and international growth, as DV continues to fire on all cylinders. In addition, I’d like to discuss our pending acquisition of Scibids, a global leader in AI-powered digital campaign Optimization. Scibids AI makes DV data more impactful and accelerates our evolution from protection to performance. This transformative acquisition combines our media quality and performance data with Scibids’ AI-powered real-time Optimization algorithms to drive superior ad KPIs and tangible business outcomes for advertisers. Scibids will help make our sizable and successful Activation business bigger and even more essential for marketers and significantly differentiate our programmatic offering in the marketplace. But first, let’s talk about the quarter. We grew second quarter revenue by 22% to nearly $134 million, above the midpoint of our guidance range. We achieved second quarter adjusted EBITDA of $40 million, representing 30% adjusted EBITDA margins and exceeding the top-end of our guidance range. Revenue in the first six months of 2023 grew 24%, against an elevated 43% prior year growth rate. The key drivers of our first half 2023 revenue growth included continued strong adoption and expansion of our ABS product, increased social platform coverage and volume, and a growing number of international wins with key global customers. DV remained highly profitable and continued to generate strong cash flow from operations. We delivered approximately $25 million of net income and generated over $32 million of net cash from operations in the first half of 2023. Our customer win and expansion momentum continued in the second quarter. We won several large new enterprise logos including Uber Brand Marketing, Sam’s Club, Pizza Hut, Revlon, Liberty Mutual and Bose in the United States, Lululemon and Unicef in EMEA, Spotify in MENAT and the Ministry of Foreign Affairs in Japan. Notable existing customer expansions include DeBeers activating DV Measurement across multiple international markets, Swarovski activating DV’s Measurement and pre-screen social solutions in Japan and Avis adopting ABS in the United States. Our win rate across all opportunities remained above 80%, with 54% of our second quarter wins being greenfield, which we define as wins where the advertiser wasn’t using third-party tools for the business that DV won. This high rate of greenfield wins speaks to the underpenetrated TAM that continues to fuel DV’s growth. We expect new client wins to propel our successful land and expand strategy, through which we grew the number of advertiser customers generating more than $200,000 over the last 12 months by 20% year-over-year. In addition to generating strong core revenue growth, we remain focused on investing in long-term strategic initiatives that leverage our core verification offering to evolve our customer value proposition from protecting media spend to optimizing media performance. Our verification business continues to create a proprietary data asset of unparalleled scale and granularity. DV analyzes approximately 300 billion data transactions, including 60 years of video content, on a daily basis. Our investments in performance solutions have leveraged this vast data asset to create Measurement products such as DV Authentic Attention, as well as pre-bid, programmatic, tools such as ABS that are based on aggregated Measurement data and contextual intelligence. The acquisition of Scibids represents the latest and potentially most impactful of these evolutionary investments to unlock even greater value in these proprietary data assets. Scibids is a cutting edge, Paris-based startup founded by a team of data science PhDs and machine learning engineers who are driven to bring programmatic advertising into the AI age. Scibids builds independent AI that automates and optimizes programmatic buying of digital ad campaigns agnostically across demand side platforms. Their AI dynamically generates custom bidding algorithms aligned with client KPIs and desired outcomes. Scibids’ powerful and sophisticated machine learning technology can dynamically action DV’s granular Measurement data within a programmatic ad campaign. DV and Scibids’ combined customer value proposition is to deliver industry-leading campaign performance based on bespoke ad KPIs or business outcomes, by offering a continuous feedback loop between Activation, Measurement and Optimization. This loop is powered by Scibids AI technology and DV’s trusted media quality and performance data that is comprehensive, granular and actionable in real-time. For example, Scibids can dynamically optimize advertiser campaign performance against customizable KPIs and desired business outcomes using DV’s Authentic Attention Measurement data. The algorithmic bidding function that Scibids builds can deliver results that far outstrip static programmatic segments, ushering in a new era for results-driven advertisers. In addition to DV’s proprietary data, Scibids AI is able to ingest a wide range of data points, including first-party data, third-party contextual data, alternative media quality and performance Measurement data, programmatic bid stream data and pricing data, all to help inform advertiser’s programmatic strategies in a way that maximizes campaign performance. This means that even if an advertiser is not a DV Measurement customer, we can still help them deliver powerful results from their programmatic spend and that ad spend can be across any of the major DSPs where Scibids complements their platforms and has seamless integrations including, The Trade Desk, Google’s DV360, Microsoft’s Xandr, Comcast’s Beeswax and mobile platform Kayzen. We have seen the power of Scibids’ AI technology first hand. In partnership with Scibids, we recently launched the DV Algorithmic Optimizer, where Scibids AI ingests impression-level DV Authentic Attention data and campaign cost data, enabling advertisers to automatically optimize towards the best performing inventory at the best price, while maintaining scale. In the results of the initial tests, on average, we observed a 45% reduction in media CPMs, a 63% increase in Attention levels and a 95% increase in impressions won. The Scibids acquisition also has the potential to enhance and differentiate DV’s Measurement business. Not only will our existing Measurement data be significantly more actionable and impactful in programmatic Activation applications, but Scibids AI will also help enhance the precision and effectiveness of Measurement datasets such as Authentic Attention. By weighting Attention signals against conversions and other ROI measures, DV can construct customized Measurement metrics based on specific KPIs and/or verticals. Simply put, the results of algorithmic Optimization in Activation applications will help enhance the construction of a truly differentiated Measurement data set. Scibids shares DV’s values of transparency, independence and innovation. Both companies are ad server and DSP agnostic and do not buy or sell media. Both companies’ solutions are privacy-forward and don’t use persistent tracking technologies, such as third-party cookies or Apple’s IDFA. And both companies have a legacy of leaning into machine learning based AI. Scibids will bring nearly 20 data scientists and engineers to DV, which will augment and accelerate our existing machine learning programs with critical AI talent that is difficult to source. To demonstrate the power of Scibids AI and explore the future of AI driven innovation at DV, Scibids and DV’s executive leadership teams along with industry experts will be hosting an in-person Innovation Day for the investment community on Thursday, September 14th, from 1 p.m. to 4 p.m. at The Standard Hotel in New York City. The event will also be webcast live. As you can tell, we are excited about the opportunity Scibids creates to help our customers harness the power of their DV data to optimize and maximize campaign performance. Acquiring this AI-powered technology allows us to meaningfully elevate our customer value proposition across Activation and Measurement, while also increasing our potential TAM with performance driven advertisers, creating broader upsell opportunities for current and new customers. I would like to turn now to the key growth milestones we achieved since our last earnings call. I will discuss these in the context of our three core differentiators; our rapidly growing scale, our focus on market defining innovation and the deep level of trust that we build with our customers as an unbiased independent partner. Starting with our expanding coverage scale, we continue to grow our Measurement breadth across key social platforms with a focus on short form video, which has a steep growth trajectory and an impressive ad load that offers a unique opportunity for advertisers and DV. Hubspot’s State of Marketing 2023 study expects short-form video to see the most growth of all marketing strategies this year as one in five marketers plan to leverage short-form video for the first time in 2023. To that end, we are excited to expand our viewability offering on Meta to include Measurement on Facebook and Instagram Reels, a rapidly growing short-form video environment with a highly engaged user base. With Reels plays exceeding 200 billion per day across Facebook and Instagram, Reels is sought after by advertisers, who can now leverage DV’s viewability and fraud measurement technology for insights into which of their ads are being seen by real users and driving results for their campaigns. We are also excited to announce the expansion of our quality solution on YouTube to enable Measurement on YouTube Shorts. DV’s launch of viewability and fraud measurement across YouTube Shorts inventory will help advertisers ensure that their ads are viewable and safe from fraud, thereby providing performance insights across a greater volume of ad impressions on YouTube. To conclude on short-form video, we continue to build momentum with TikTok, which contributed nearly a third of our social measurement revenue growth in the second quarter as customers rapidly activated the DV Authentic Ad. We generated more TikTok revenue in the second quarter than we did in full year 2022 and TikTok remained our third largest social revenue generator, after Meta and YouTube. International markets fueled a good deal of our TikTok growth, where we continue to focus on maximizing our language coverage and plan to be live with brand safety and suitability in 25 additional markets in the second half of the year. Overall, we grew social measurement volume by 41% year-over-year in the second quarter. Our social revenue growth in dollar terms continues to be led by advertisers leveraging our solutions on Meta, which generates almost half of our social measurement revenue. In addition to offering a greater breadth of verification coverage across its platform with Reels, Meta remains committed to providing advertisers with a deeper level of transparency through brand suitability controls and verification. We continue to engage with Meta on brand suitability Verification and Measurement solutions and look forward to rolling out an expanded DV Authentic Ad in the coming months. Turning to CTV scale, we grew CTV Measurement volumes by 32% in the second quarter, outpacing the 14% CTV revenue growth rate expected of the industry in 2023, according to IAB’s research. And since we launched viewability verification and fraud protection coverage on Netflix’s ad-supported plan in March, we have seen four consecutive months of growth, both in the number of advertisers using our solution and in impression volumes. We expect long-term growth as Netflix continues to innovate its advertiser offerings. In addition, we continue to scale across retail media networks, which generated 81% year-over-year revenue growth in the second quarter, with revenue contribution across all three business lines. Our Measurement tags are accepted on over 50 of the key global retail media networks and sites, including 12 of the top retail media platforms and 39 major retailers. Moreover, our partnerships are global, with DV’s Measurement coverage of Retail Media spanning all major geographies, including the U.S., EMEA and APAC, with international partners such as Sainsbury’s, CitrusAd and

Thank you, Mark. And good afternoon, everyone. We delivered solid revenue growth and strong profitability in the second quarter and the first half of the year, in the face of elevated prior year growth rates. The business performed broadly in line with our second quarter expectations. Revenue came in slightly above the midpoint of our guidance range while adjusted EBITDA exceeded the top end. Second-quarter revenue growth of 22% was driven by 29% growth in Activation revenue, 16% growth in Measurement revenue and 4% growth in supply-side revenue. Advertiser revenue, which includes Activation and Measurement, grew 24% year-over-year and continued to be volume led. Second quarter volumes or MTMs were up 24% year-over-year, while price for MTF was unchanged. The price impact of the continued mix shift toward premium-priced solutions and the ABS price bifurcation that was implemented this year was offset by strong revenue growth in international, where our fixed fees are lower than in the U.S., commensurate with media costs being lower in international markets. Additionally, the price bifurcation of our standard programmatic products last year contributed to 10% MTF growth in Q2 2022, creating a substantial prior year comparison. Activation revenue was led by ABS, our premium-priced solution, which delivered 51% revenue growth in the quarter and comprised 56% of Activation revenue compared to 48% in the prior year period. ABS volumes were up 48% and the product’s fixed fee was 2% higher. ABS’ growth was largely driven by existing ABS users continuing to expand their ABS usage with new advertiser activations also contributing to the growth. Turning to Measurement, revenue grew 16% driven by 41% growth in social volumes with nearly half of our social revenue growth coming from markets across EMEA and APAC. Overall, International Measurement revenue grew 39%, as compared to 18% in the second quarter of last year, primarily driven by existing customers expanding their use of DV’s solutions on social platforms. International represented 27% of total Measurement revenue in the second quarter. We expect Measurement growth to continue to be driven by existing customer expansion on Social, as we roll out the DV Authentic Ad on Meta Reels and YouTube Shorts and expand language coverage on TikTok. We expect non-social Measurement growth to be mainly driven by new enterprise logo wins and therefore subject to the timing of when new customers sign, onboard and ramp their ad campaigns. Supply-side revenue grew 4%. The majority of our supply-side revenue is generated by platforms, which include SSPs, DSPs, marketplaces, ad exchanges and retail media networks. DV’s successful platform offering is well penetrated across the ad tech ecosystem and we expect incremental revenue growth to come from the adoption of new product offerings, such as DV’s Brand Safety Floor, by existing platform customers, as well as new wins such as Outbrain, who signed a multiyear contract for Brand Safety and fraud avoidance in the second quarter. Shifting to expenses, cost of revenue increased by approximately $7 million, primarily due to an increase in costs from revenue sharing arrangements with programmatic partners tied to higher programmatic revenue, as well as an increase in cloud services costs. Revenue less cost of sales was 80% in the first half and is expected to remain relatively stable for the remainder of the year as we continue to invest in scaling the infrastructure needed to support our growth. Similar to the first quarter, product development expenses increased due to our ongoing investments in AI and machine learning engineering resources. Sales and marketing expenses increased as we launched business operations in several international markets, executing on the international business expansion plans we discussed last quarter. G&A expenses declined year-over-year due to lower compensation and insurance expense, despite a $1.5 million increase in bad debt expense primarily related to a reserve for the MediaMath bankruptcy. We continue to expect both sales and marketing and G&A to benefit from our growing scale. Total non-GAAP operating expenses, which exclude stock-based compensation and other items for comparability, grew 18% compared to our topline growth of 22%, reflecting the efficiency of our operating model as we scale. We achieved second quarter adjusted EBITDA of $40 million and adjusted EBITDA margins of 30%, partially due to G&A savings and a pace of hiring that was slightly slower than planned. We delivered approximately $13 million of net income, as compared to $10 million in the prior year. Net operating cash flow was $32 million in the first half of the year, as compared to $27 million in the first half of last year. Capital expenditure was approximately $8 million in the first half of the year, as compared to $14 million in the first half of last year, which included the expansion of our global headquarters. We ended the quarter with approximately $295 million in cash on hand. Turning to guidance, we expect third quarter revenue in the range of $135 million to $141 million, which at the mid-point implies growth of 23% year-over-year. We expect third quarter adjusted EBITDA in the range of $39 million to $41 million, which at the midpoint implies an EBITDA margin of 29%. For the third quarter, we expect stock-based compensation to range between $15 million and $17 million and weighted average diluted shares outstanding to range between 173 million and 175 million shares. Following a second quarter topline performance that was broadly in line with our expectations and accounting for some friction as a few of our large advertisers transition to new DSPs, we are reiterating our full year 2023 revenue guidance range of $557 million to $569 million, a year-over-year increase of 24% at the midpoint. We are also reiterating our adjusted EBITDA guidance range of $171 million to $179 million, which continues to represent an adjusted EBITDA margin of 31% at the midpoint. Turning to Scibids, we have agreed to acquire this industry-leading AI technology for $125 million with a mix of cash and equity. Approximately $66 million of the purchase price is expected to be paid in cash, which we will fund with cash on hand and the remaining consideration will be paid in DV common stock. The deal is subject to customary adjustments and with potential additional consideration to be paid based on Scibids’ achievement of certain performance milestones by year end 2023. While we are not providing 2024 guidance today, we expect Scibids to generate between 15 million to $17 million in 2024, based on a 30% to 40% year-over-year growth rate. The company is near breakeven today and we see the opportunity to scale Scibids’ product across our customer base. We plan to enhance Scibids platform and value proposition through investment in people, software, processes and systems, and by building scalable data integrations between our two companies. We anticipate completing the transaction by the end of the third quarter and plan on updating 2023 guidance to reflect the impact of Scibids after the transaction has closed. To conclude, we delivered a strong first half and remain focused on successfully executing against our plan for the rest of the year. And with that, we will open the line for questions. Operator, please go ahead.

Thank you. [Operator Instructions] Thank you. Our first question is from Michael Graham with Canaccord. Please proceed with your question.

Hi. Thank you and congrats on the steady growth. I wanted to dive into the AI topic and Scibids for a minute. And Mark, you mentioned in the past that the interplay between Measurement and Activation is a key multiplier for your business. Can you just talk first about like how does Scibids fit into your business model? Should we think of it as another product like Authentic Attention that advertisers can use or is it going to be more integrated into your product roadmap? And then just talk about more broadly how do you think it impacts your long-term product roadmap?

Yeah. Yeah. Thanks for the question, Michael. And it’s -- as you can tell by the call, we are really excited about this acquisition. I think the short answer is both. It is a separate product for us that we will be able to leverage in Activation applications using our data for advertisers, as well as it has the ability to enhance our entire product set by using some of the outputs of the learnings from the Activation implementations to enhance our Measurement business. I’d like to think of it as having kind of three main buckets of impact on the business. The most obvious one is kind of on the revenue side, right? On its own, it’s a product that we will be selling in conjunction with them to leverage our data. But also it’s going to help us as a differentiated product set to sell enterprise customers, because the second bucket is really -- it’s a huge differentiator for us. No other company in our space has a product like it and it really sets apart our bundle of goods or basket of goods that we will be able to sell folks. So it will drive revenue, it’s a differentiator, and I think, finally, it opens up new markets for us. We get a question a lot, how do you guys deal with, for example, performance based advertisers or DR folks. This really speaks to them, because it is an Activation tool set that can leverage our data, but do so against specific KPIs and those KPIs can be things like acquisitions or new customers, things like that. So I think it really hits those three marks, which is revenue differentiation and kind of new market expansion. I’d like to think of it almost as a future ABS on steroids, right? It really has the ability to kind of drive Optimization in a really unique way.

Okay. Thanks a lot for the extra color there, Mark.

You got it.

Thank you. Our next question is from Matt Swanson with RBC Capital Markets. Please proceed with your question.

Yeah. Thanks guys for taking the question. For, Nicola, you have got a history of being very pragmatic around guidance. Could you just kind of talk to us about how you are thinking about the macro environment for the rest of 2023 and then maybe just given the durability of your volume-based model, can you maybe help us think about how potentially a better spending environment shows up in that MTM, MTF equation?

Yeah. I will take the question, Matt. So as we have said since the beginning of the year, I think, from our perspective, the macro to us is something that is not providing strong tailwinds this year. We have taken sort of a conservative view and sort of a limited outlook view, right? So we are kind of looking at it on a quarter-by-quarter basis based on what we are seeing to be the spending pattern of our advertisers. So we are not seeing it as a strong tailwind. This is certainly not something that we put into our guidance consideration for the second half of the year. If there were improvements, I think, what you would see is an MTM growth number that would be above what we have be able to achieve in Q2, which was so 24% growth in MTM. I think what you would see is an increase in the MTM part of our equation. The MTF is actually a good story. It was unchanged this quarter just because our international growth is very strong and that has an impact on MTF, because the price of media outside of the U.S. is obviously lower than in the U.S. But when you say 39% growth in the rest of the world, on the Measurement side of the business, that’s actually a positive, which has an impact on MTF sitting at basically flat. So if the market were to really rebalance strongly, I think, our assumption of no tailwinds returning to the strong billings.

That’s really helpful. And if I could squeeze one more in for Mark. Greenfield wins still really impressive at 54%, but maybe like a slight step back than what we have seen in the last few quarters. Could you just kind of talk about the dynamics of the quarter, whether it be competitive, vertical, geo-specific that led to that change or were the last few quarters of being high 60%, 70%, was that more the anomaly from a greenfield perspective?

Yeah. It’s a great question, Matt. I mean, it was definitely a bit of a slide in the percentage of kind of greenfield wins. I don’t think -- I think it’s a bit of an anomaly based on where we think the future lies, particularly around our Activation business and the Scibids acquisition. That being said, we still beat a lot of competitors and I think that’s the good thing, too. And then the one thing about being a competitive deal and this may reflect a little bit in some of the numbers we shared is that, when you beat a competitor, you have to dislodge a product and it takes a bit longer to do so than just turning somebody on from scratch, right? He’s not using anything. So I think we love greenfield wins, we love competitive wins, all wins are good. Some take a bit longer to activate than others, and I’d say, those are, in some cases, the competitive wins.

Thank you. Our next question is from Laura Martin with Needham & Company. Please proceed with your question.

Thank you. Okay. My first one is on Attention. Mark, you are sort of been spearheading the pivot towards Attention has been really great resonance. Love the deal with TVision, because they are the other sort of leader. You guys are the tip of the spear in redefining Attention. My question is, can you bring us up to date on this metric pivot towards attention in the digital space. But secondly, my recollection is they are based on a panel, so I would love your thinking about why a panel plus your super high tech is worth more than just your standalone solution to Attention? That’s my first

That’s great. Great question, Laura. And yeah, we really are pushing the boundaries on the Attention definition and I think the TVision deal is a good example of that. And the combination -- we have obviously had the combination of a gold star panel with the reach of granular data at scale is probably -- is the best solution you can put together out there. And the reason why is that panel that TVision leverages can provide some really unique insights on behavior, particularly on the TV front that you just can’t get from a granular data set on its own. So I think it does add, we baked an incredible cake and TVision adds this frosting that just makes it so much better and gives us an attention solution that really is unmatched. And I think when you combine that now with what we have been able to do or what we will be able to do with the Scibids acquisition and what we have already shown, we launched, we announced that initial partnership with them a little bit over a month ago and our initial trials on Attention were really had some exceptional results when you are looking at the number of one impressions increasing by over 90% with attention increasing significantly as well. I think the combination of having a panel plus granular data solution on the metric side, plus an algorithmically based Optimization tool set on the Activation side for Attention, which no one else will have it kind of puts us really in a very unique position or our attention as we move from -- that whole area of being accepted by the marketplace to actually being commercialized at scale.

Super helpful, Mark. And then on Roku is all over your press release today, and my question here is that I love Ruko, great for me that you are doing all work with them. My question to you is, first is that, you charge $0.08 per Measurement, $0.21 for Activation. Why you are spending any time -- per impression per 1,000 impressions, why spend any time in CTV, because for the same price, you can allocate that time to rolling out another Meta product or rolling out another TikTok product where you are getting lots of new impressions from both existing and new clients, onshore and offshore, if you are just going to get a fixed fee, why spend any China connected television where CPMs are 35, but they are not that many thousands. There’s not that many impressions compared to social? So that’s my second question and thanks very much.

Yeah. It’s a really interesting question, Laura. And I think the short answer is, is because our customers are there and we want to create -- our driver has always been to measure every impression on every platform in any market anywhere on the planet, right? And CTV is part of that and we want to have a basket of goods, so when an advertiser is spending on Roku or spending on Meta or spending on the open Internet or buying programmatically that were there. So I think it’s important for us to be in CTV. For sure, the volumes are very different than what you would see, for example, in short-form video, which has been a real interesting growth area for us in areas like TikTok or Reels. For sure that the volumes are different there. However, an advertiser says, we want DV everywhere, you need to be everywhere and we want you across CTV. So I think it’s part of this idea of having a basket of goods that covers all the bases so that when you close an enterprise customer and enterprise advertiser, like a Sam’s Club or like an Uber and they are advertising across all these different venues, they want us there. I think it’s really important.

Thanks very much.

Thank you. Our next question is from Raimo Lenschow with Barclays. Please proceed with your question.

Hey. Thank you. Sorry to go back to macro a little bit. if you think what you are seeing out in the world there, like, there were other companies that have reported this afternoon as well, where it looked like it was incrementally getting a little bit worse. Like can you just kind of -- I know you talked on the call about it, but like if you think about how you would characterize like what’s going on out there, it felt like we had the profit at the beginning of the year, got a little bit better. Is it getting a little bit worse that you kind of increased your caution or like, how does it feel out there, sorry, like, I know you kind of tried to enter that already, I am just trying to dig deeper? Thank you.

Yeah. Raimo, what I would say is, we haven’t seen much of a change. I would say that, what we have seen in the market is also the benefit of lower comps for some of these companies, which as you know, our model is fairly steady in a good environment versus a bad environment, especially related to CPMS. So we have not seen a material change in the environment and we are taking a cautious approach towards it. The one number that we do track is Magna, they published in June and didn’t essentially changed our forecast for the year. So I would say those are the two things we are seeing probably some lower comps for some of the companies that are reporting and a general unchanged environment for us and our view that we will take the tailwinds only when they really appear.

Yeah. Okay. Perfect. And then how are you thinking about like setup in terms of like the growth internationally was really strong. As we go further through this and people talk about soft lending, how do you think about that balance of like maintaining that high discipline around margins and profit, which you do and well done and also getting ready for any potential improvement in underlying business? Thank you.

Yeah. Raimo, I think, we have been pretty clear that we have continued to invest even when the market has not been extraordinary and I think that investment is going to pay off. Whether it’s our R&D expenses, which we continue to lean into to build new products, new coverage or the acquisitions that we just did. We are ready, when the market tailwinds do materialize, take advantage of those, I think, even in cases where we are looking at some of the enterprise clients that we have added to our roster their spend has been relatively muted. I think when they are ready to start spending as well, we will be able to experience that. So closing lots of deals now, even if those deals aren’t huge and investing in our technology and spending money on acquisitions, all of those things together are going to put us in a great position when that tailwind does hit.

Okay. Perfect. Thank you.

Thank you. Our next question is from Youssef Squali with Truist Securities. Please proceed with your question.

Yes. Thank you so much. So maybe one question for Mark and one for Nicola. So, Mark, just going back to the Scibids acquisition, obviously, congrats on it. But just kind of at a high level and I think you spoke about TAM expansion as a result of this. Maybe can you just drill a little deeper into that how this kind of allows you to move into more kind of performance? And if you can maybe quantify kind of the potential TAM and who are you also kind of competing with for this type of offering now that you are -- you have doubled down on AI as part of the offering? And then, Nicola, can you maybe just speak a little bit to your guide for the year, particularly around the second half and your expectations for growth in Activation versus Measurement? Thank you.

Yeah. Some great questions, Youssef. So let me talk a little bit about Scibids and what it does for business opportunities for us. With regard to expanding those opportunities, the first thing which I mentioned previously is, it opens up a whole new set of customers and those are customers who are really driven by very specific performance KPIs. So direct marketers, those who are looking for who drive business based on a CPA or an acquisition-based data point. A lot of those folks, brand safety and suitability has been a consideration, but not an absolute, right? So this gives us a new tool set to approach folks like that, that we can put it together a unique offering for them. So, A, it adds a whole new customer toolset, new customer potential for us. The second is, if you think about what percentage of digital ad spend programmatic makes up, it’s roughly 85% plus if you take search out of digital ad spend. We are still not even close to getting to that level of our business being at 85% of our revenue coming from programmatic. So I think this opens up new opportunities to take advantage of customers who are spending a significant portion of their business across programmatic. And then finally, when you think about it, and one thing I mentioned in the earnings call, it’s great that we can leverage our data inside it and we have already done that, for example, across our Attention data sets and we are doing it, we are experimenting with some other data sets. But the great thing about Scibids is that it can work with any data set, even our competitor’s data sets to optimize. So it gives us an entirely new group of customers that may not even be Measurement customers for ours that we hope they will be down the road as we sell in a complete package. But there are customers who may be using competitive data that can be potential customers. So if you think of that as far as ratcheting up our programmatic tool sets, giving us an opportunity to talk to a different type of customer, deal our customer and then giving a chance to talk to customers that aren’t even DV Measurement customers, I think, it shows that there’s real expansion opportunity for us. What that looks like? I mean, programmatic is a multibillion-dollar business. This is a take rate business. The Scibids business is take rate business. So some percentage of that billions, we are not ready to put our finger on that just yet, but it’s -- I think it’s a big TAM expansion for us.

Youssef, your question about the second half and the mix of the business, we basically ended the first half with roughly 30% growth on Activation, roughly 20% on Measurement. I think we are continuing to be excited about the contribution of ABS in the Activation line and we are really excited about the contribution of social in the Measurement line. So I think those two are good indicators of what we are thinking will continue in the second half. As a result, obviously, you will end up with Activation representing a larger percentage of our total revenue. So the key growth drivers of those two lines remain the same as what we see in the first half.

Okay. That’s helpful color. Thank you both.

Thank you. Our next question is from Eric Sheridan with Goldman Sachs. Please proceed with your question.

Thanks so much for taking the question. I wanted to focus on slide 12 on the Retail Media Network disclosure that was in the deck for the earnings. Can you talk a little bit about maybe how we should be framing up the opportunity set you see when you lay out 12 retail media platforms, 39 retailers, the number of global partnerships? How should we be thinking about elements of both client growth, spend, scale building in that business on a multiyear view versus what you have already seen in the recent past? Thanks so much.

Yeah. It’s a great question, Eric. I mean, obviously, Retail Media Networks continue to be a bright spot in our business, growing at over 80% this year and the nice part about it for us is that it covers all three line items of our business, right? So we work with folks, Retail Media Networks, both on the Measurement side, the Activation side and the supply-side. I think we have kind of given rough numbers on a quarterly basis, what this looks like. Generally, it’s grown. In Q1, it grew over 100%. This quarter, it’s grown over 80%. We are working with some of the biggest customer, Retail Media Networks out there like Target, Kroger, which we closed around this time last year, Macy’s, Dollar General, Walmart, Amazon, I mean, big companies. What does this mean for us is potential upside? I mean, this was a business now, which I think is approaching $30 million, $40 million for us a year, which I think is pretty substantial coming from zero just a few years ago, and I think as that sector continues to grow and as these companies continue to more sophisticated our business opportunity gets stronger and stronger. The other thing it does is, we have got a really strong footprint of CPG companies that we work with. So folks like Colgate, Mondelez and Unilever. It just makes our connection with those folks that much stronger, right? So those are traditional CPG companies, but they plug into these ad networks, these Retail Media Networks. So it creates a stickier relationship when we are touching both the Retail Media Network and the CPG company as well. And so, I think, our ability to grow here, I think, is continues to be strong, but it has a residual impact of really creating a stronger vertical offering for our overall business, particularly CPG companies.

And maybe if I could just ask one follow-up there, because this was kind of where I was going as well is to the extent to which this acquisition you are doing today talks about pooling sort of your own data with third-party data and you think about elements of data that come in from the CPG world potentially into Retail Media. Is this another AI optimization theme that we could see brought back to some of these retail media networks? Thanks.

Yeah. For sure. I mean, the way that Scibids works is it can look at multiple -- it’s multi-variant decisioning on a single ad impression and the ability to pull in some of this Retail Media data, for example, transaction data or shopping cart data is a real advantage for a model like that. So it definitely adds another play, and we continue to look at potential partnerships and customers that the Scibids and AI bidding will work for. And Retail Media is a nice one, because Retail Media is driven by selling, right? It is a pure response and performance-driven area and I think this fits quite nicely into that story.

Great. Thank you.

You got it.

Thank you. Our next question is from Vasily Karasyov with Cannonball Research. Please proceed with your question.

Thank you. Good afternoon. I wanted to follow-up on what you said in your prepared remarks. So one is, I think you said that the results of the second quarter were broadly in line. So it sounded like it was not quite as good as you expected. I wonder if you could go into a little more detail where the unexpected softness I assume, occurred? And then the second comment I wanted to follow-up on is, you said that there is some -- if I understood correctly, there is a disruption while some advertisers are switching DSPs. So two questions there. I understood that you are distributed by all major DSPs. So there shouldn’t be much friction there, is that correct? And also, when that process is completed, should we see a benefit or some incremental lift to revenue growth rate in the second half of the year, maybe Q4? Thank you.

Yeah. So let me take them in order. So, I mean, in Q2, we basically -- when we say we are broadly in line is, we slightly beat the midpoint of our guidance and let me talk about the drivers of Q2 revenue, right, which is we saw the same drivers as we did in Q1. We had over 50% growth in ABS, 41% growth in social volumes, 39% growth in international. We saw a very strong rate of wins, client wins over 80%. What we did see in Q2 is a bit of an elongated ramp cycle for some of the large client wins that we had on the Measurement side and so some of that revenue just took a little bit longer to get into the quarter and that’s essentially what we saw in Q2. And so that put us right at the midpoint of our guidance. And as we continue to sign larger enterprise clients, I think, we will kind of see a little bit of that elongated cycle to just see all of their campaigns coming up to speed, especially if the win is a non-greenfield client. In terms of what we see…

Okay. Thanks.

Yeah. So now in terms of what we see for the second half of the year, let me just talk about the friction point that we made. We did have some large clients that were using MediaMath as a DSP and those naturally just need to transition to another DSP, they will, we are working with them. But there is a bit of a natural transition there going from one trading platform to the next and so our guidance for the second half sort of takes that into account and we did have some large clients that were using MediaMath. So that’s a little bit of friction. As you pointed out, this is a point in time. So we don’t anticipate in the mid- to long-term, any impact to the overall revenue trajectory. It’s just for a point in time. It just might take a little bit of time to get to new DSP. And then in terms of what we saw is positive, yeah, Meta Reels, YouTube Shorts is about to launch as well. There are just one other vector of growth for the social platforms. So they were broadly kind of taken into account for what we were considering for social revenue growth. So the large -- and you know this already, obviously, the large opportunity for us on social is brand safety on Meta, which is hopefully for next year into the revenue contribution.

Thank you.

Thank you. Our next question is from Arjun Bhatia with William Blair. Please proceed with your question.

Perfect. Thanks for taking the question. I wanted to ask about new form factors, Mark, as you kind of expand into new form factors on social with Reels, right? So whether it’s short-form video or eventually to come with the news feed, what does the typical cadence of existing DV customers turning you on look like? Is there a friction point where they are deciding do we turn on DV or are they doing a cost benefit? What is that decision-making like and what does that time line look like?

Yeah. It’s interesting. What we have seen, for example, with the Reels rollout, the friction has been pretty low and I think because there’s a significant amount of demand against that area. So we saw kind of a nice cadence of folks kind of turning on that additional coverage. Now one thing that’s interesting to know is, when you look at our coverage of Facebook customers, I think, it’s better of our top 100 that are actually measuring Facebook. It’s only -- it looks slightly above 50%. So we still got a good number, a good amount of growth to just get folks to turn on anything across social. And I think that’s where, as Nicola noted, getting brand safety and suitability across more of our social networks is going to be key. Right now, even with what we are doing across Reels and we will be launching across YouTube Shorts, we are still talking viewability and invalid traffic. And I think the real value unlock on social is really around brand safety, which is what we saw on TikTok as we continue to grow across TikTok and short-form videos. So I think the cadence of the, let’s say, the easy products, the IBTN [ph] and the viewability products has been pretty good. But I think we still got some growth to do when it comes to getting the what folks really want, which is brand safety and suitability across social and I think we have got some nice opportunities there still to tap into.

All right. Perfect and that’s super helpful. And then going to Scibids, it sounds like as we move more into performance, especially with the Scibids acquisition, it seems like you might have a role to play in directing ad dollars and ad spend to different DSPs and correct me if that’s wrong. But how does that impact DV’s role as an independent player in this ecosystem where you are not a part of any conflict if advertisers decide to move ad dollars from one network to another?

Yeah. It’s a great question, Arjun. And it would be really clear that, the acquisition of Scibids and Scibids’ role in kind of the digital ecosystem is create no additional conflict and is actually as independent as we are in the media transaction. They work with Trade Desk and with DV360 and Xandr and Beeswax and probably more DSPs down the road. The idea is that there are tools. There are a tool that can be implemented by an advertiser. It’s technology that can be moved similar to our Activation tools are currently can be moved. Advertisers can move from platform-to-platform. So for us and for Scibids, the key will be to be everywhere, right, to be agnostic to the platforms that we are on and to make sure that wherever our advertisers want to spend, we have technology there. Very similar to our current Activation business, so whether it’s ABS or standard brand safety and suitability segments or our contextual targeting segments. We want those to be everywhere and we want them to be fueled by DV data. So I think the independence of Scibids -- or the acquisition Scibids doesn’t change either our independents or theirs and they have got great relationships with those platforms and have since the day they launched across.

Okay. Got it. Thank you for taking the questions.

You got it.

Thank you. Our next question is from Mark Kelley with Stifel. Please proceed with your question.

Hey. Great. Thanks very much. A couple of quick ones just on Scibids. I guess you talked about getting access to new customers. Can you give us a sense for the types of customers you now have access to? Is it more small and medium business types of clients, is it geographic specific, given that they are based in France? That’s my first one. And then as a follow-up to that, I guess, your 30% to 40% growth rate assumption for next year for Scibids, does that bake in accelerating growth, given that you have got a large sales force or is that consistent with the current trends for the Scibids business? Thank you.

Sure. So on the customer side, I think, it’s less geography, even though it’s always nice to have some more resources across EMEA, considering that there is a little bit of a bounce back as we have shown in our numbers earlier. It’s more -- and it’s not really size either. So what it is, is goals and KPIs. Advertisers who are looking for specific drivers of results that may not be as concerned about viewability, for example, or be as concerned about brand suitability whereas brand safety maybe just be fine if their impressions go. It’s an advertiser who’s looking for less of an on-off and more of a fluid, hey, I can take a slightly less viewability or slightly less suitability for slightly higher performance. The way that the Scibids’ algos work is it’s a dial as opposed to a directive, right? Where it’s not black and white saying, yes or no, it is a continuum. And what that does is it opens up to bunch lots of advertisers who, again, are very focused on performance as opposed to being 100% viewable or meeting specific viewability criteria. It’s just a different type of advertiser, which I think is kind of interesting for us. The other aspect, I did mention geography, it is interesting that 60% of their business comes from the rest of the world, which is a very -- a different ratio than our business currently. So it is nice having some contacts outside the U.S. or some touch points outside the U.S., which this helps us as well.

And on the question for the growth, the 30% to 40% growth is, it doesn’t assume an accelerate growth. We are buying an early-stage technology. We are going to spend a lot of time just making sure we understand the technology and how we integrate them to our own technology. So there will be growth above the standalone growth. But understand this is an early-stage technology and their current revenue pattern is really dependent on a few clients. So this is about the technology, understanding how we are going to use it on our side. The revenue will grow in the future years, but this is really about technology first and foremost.

Okay. Great. Thank you both.

You got it.

Thank you. Our next question is from Andrew Boone with JMP Securities. Please proceed with your question.

Good afternoon and thanks for taking my question. I will keep it to one. I wanted to go back to Facebook. It sounds like social has some real tailwinds, and as we think about brand safety coming on to the feed for next year, Mark, there’s the real potential that you guys could have a step function change in terms of revenue. And so the question is really, you guys have held EBITDA margins at 30% and that’s kind of how you guys have managed the business. If there was a step function change as we think about the potential of Facebook, would you be willing to let margins expand or would you maybe reinvest now back into performance. How are you thinking about controlling kind of the EBITDA profile with the potential of a significant supply source coming online? Thanks so much.

Yeah. Thanks, Andrew. It’s a great question. I think we have been pretty adamant around this idea that we are going to continue to reinvest in the business. Now it doesn’t mean that margins won’t expand a point here or there if we do open up a step function when it comes to an inventory source or a platform like Meta. But those platforms also come with some work and some scaling around it, too. So I think we continue to focus on revenue and topline, keeping that EBITDA margin right in the sweet spot where we have it. So we can continue to invest. We talk a lot about land and expand, and we are still focused on landing lots of customers. And I think as we have noted several times so far today, some of those customers may be totally new opportunities for us through algorithmic, fee-based solutions. So we want to continue to invest there, too. So I think we are going to keep that band fairly tight like we said. We have done a pretty good job over the last three years, keeping in that zone, and I think, we are going to stay in that zone.

Thank you.

Sure.

Thank you. Our next question is from Brian Fitzgerald with Wells Fargo. Please proceed with your question.

Thanks, guys. Maybe two really quick follow-ups on Scibids. Congratulations on the deal. I wanted to know maybe if you could explain a little bit how their customer algorithms work alongside the bid algos from the DSPs, how they deal with advertisers who do use cookies and other identifiers? And then quick clarification, did you take a bad debt charge regarding MediaMath in 2Q or should we expect that in 3Q?

Yeah. I will answer the second question first. We did take a charge in the second quarter. Bad debt increased by $1.5 million in the quarter versus prior year and that includes the impact of MediaMath.

Okay.

Yeah. On the questions around Scibids and their interactions with the DSPs. So the custom algo works in conjunction with any additional Optimization that the platforms may have embedded in themselves. So the leading platforms, so The Trade Desks and the Googles of the world have been really clear on it. They love this idea to bring your own algorithm and let it run against what you -- what are important to you as an advertiser want to bring and then they can help kind of clean up and optimize on top of that at the end as well. So there’s a really nice complementary existence between these platforms and between the tool sets, then it plays well. With regard to the data points that the advertisers are using, for the most part, these, well, they are not for the most part, they are not cookie-based data points and particularly around the data that we are now plugging into these systems, so things like Attention, data points like viewability and others are identifier free and cookie free. So they maintain the same level of privacy friendliness that we have had to-date, yet they still drive real KPIs and I think it’s a nice, again, complement to our business. We think this is a growing area that’s going to not only kind of enhance our Activation business. But as we said, I think, it has the ability to actually enhance our overall Measurement business as well.

Excellent. Thanks. Thanks, Mark. Thanks, Nicola.

Thank you. Our next question is from Nick

Hey, guys. Thanks for fitting me on here. Just a little bit deeper on Scibids, if you don’t mind. I mean, so you mentioned Scibids, it automates and it optimizes programmatic buying across DSPs. I am envisioning Scibids technology is designed to deploy and allocate ad spend on behalf of the brand, maybe following a rule set across KPIs as dictated by the brand or agency. So effectively, I am thinking AI trading here, but I just want to understand if I am correcting that, if I am understanding that correctly or not or perhaps this technology is more designed to act like an AI assistant for traders. Just any further deep dive here on how Scibids is deployed would be helpful?

Yeah. So it’s the former versus the latter of what you mentioned. It’s really focused on optimizing the bid, right, and ensuring that the media spend is allocated to the impressions that drive the highest return. So based on the KPIs that the advertiser is going after, how can I drive those KPIs, right? So how can I drive them based on the lowest potential price, right, that drives the highest ROI and that’s what it looks like. And the main part about it is, it does it on every impression. So think of a decision engine that doesn’t look at the campaign or doesn’t look at the advertiser. It looks at every individual impression against that specific KPI and looks to maximize the bid, or in this case, minimize the bid to deliver the maximum return. That’s what the logic does. That’s the algorithm does. So it’s actually plugged into the DSPs to make those decisions. So it’s less of a -- it’s not an assessment at all to a trader, it actually does the trading on its own. And does the bid Optimization on its own and runs and makes those decisions. And like I said, the cool part for us is that we can start to load our data, our unique and proprietary data like Attention Data into these algos and have advertisers start to optimize against things like Attention at a very low -- a much lower rate than they would if they were just trying to target a static attention segment, for example.

Got it. That’s actually super helpful. And then just one follow-up, if I could. Just regards to the allegations that Google has been facing for some video ads delivered on third-party websites in small windows and without audio, as I am sure you know. I am just curious if there are any implications here for DV. Like are you seeing advertisers potentially pull back on YouTube and redirect or perhaps it triggers advertisers to actually utilize your viewability solutions more so, just given the concern? Just any thoughts on implications of the recent development here?

Yeah. Look, obviously, it’s definitely a challenging place for advertisers to be to try to figure out what exactly is the next step for them as they look at -- they want to continue to spend across Google properties. But it really begs the idea and the question around transparency and the role of third parties in that. We have worked really well with Google to provide an independent take on GDP, YouTube kind of the issue that was surfaced. I think we have got a pretty extensive blog post in which we outlined what we were able to deliver and what we see out there in the marketplace. And I think it always is -- it always begs the question as to making sure that the companies that are measuring and that are doing the analysis of the transaction are truly independent outside of the media transaction. And I think we have always taken that position, which is we work with everybody. We do so in a way that is accredited and we do so in a way that is independent and I think our relationship with the advertiser is tantamount. That’s who pays our bills and that’s what we have to worry about every day. So we focus on trust. We are focused on making sure that we are independent and we make it folks are providing as much transparency as we can. So we haven’t seen any significant pullback on across Google properties. Obviously, they had a great quarter as well. So our relationship with them remains strong and we remain both focused on the same thing, which is delivering great results to advertisers and doing so in a way that’s transparent and leverages the technology that we have built.

Great. Thanks, guys.

Thank you. Our next question is from Mark Murphy with JPMorgan. Please proceed with your question.

Hi. This is Arti on for Mark Murphy. Thanks for taking the questions and congrats on the quarter. First question is, I know you talked about the macro a little bit earlier. Anything worth calling out in terms of divergences among customers in terms of the sector?

No. I mean, as you know, we are fairly diversified across most sectors. So we haven’t -- we didn’t see anything that was significantly outweighing other sectors. We still remain very strong in CPG. We still -- we saw some good uptick on auto, which is one of our smaller segments. But in general, for us, we are still diversified that whether one segment goes up or down, it doesn’t really impact our financials and we didn’t really see anything that was materially different than the prior quarter.

Great. Thank you. And then one more, I will step back into the queue. In terms of the ABS price bifurcation, I know last time you guys said that was kind of frictionless and rolling out nicely. Does that continue to be the case or just anything worth calling out given we have a couple of more months behind us? Thank you.

No. It’s as we had planned. It was frictionless. It was a small amount based on the price of the product. So nothing new to report on that.

Thanks.

Thank you. There are no further questions at this time. I would like to hand the floor back over to Mark

Thank you all for joining us this evening. We delivered solid business momentum and couldn’t be more excited about the opportunities that lie ahead. We look forward to seeing many of you at DV’s Innovation Day in New York on September 14th, where we plan to share our long-term vision and strategy for AI-powered innovation. Have a great evening, everybody.