Integral Ad Science Holding Corp. (IAS) Earnings Call Transcript & Summary

All right. Good morning, everyone. My name is Matt Cost, from the Morgan Stanley U.S. Internet team, thrilled this morning to be joined by Lisa Utzschneider; and Tania Secor, CEO and CFO of IAS. Thank you so much for being here.

Thank you for having us.

Great. So maybe let's start out, Lisa, maybe at a higher level for people who may be newer to the IAS story, give us a quick overview of who you are, where you fit into the ad ecosystem and how the business has changed over the past few years?

Sure. Good morning, everyone. I'm Lisa Utzschneider, CEO of Integral Ad Science or IAS. Thanks for having us, Matt. We're a leading global optimization and measurement company. Our core customer base, our Fortune 500 brands. We have worked with over 2,000 advertising customers. And what our technology does is help brands find higher-quality media, both across the open web and social platforms. In terms of how we've evolved over the years, we've been around for 15 years. And earlier, as a third-party verification company, our technology helped advertisers and verified that their ads were viewed. So let's say, Nestle is running a 30-second video on YouTube, verifying that, that Nestle ad was viewed and viewed by human and not a bot. And over the last 15 years, as the Internet has evolved, as more and more users are spending a disproportionate amount of time on the social platforms in CTV, our technology has also expanded into brand safety and suitability ensuring that wherever that Nestle ad is running, it's adjacent to brand safe and suitable content, in particular, on the social platforms.

Got it. Maybe, Tania, let's unpack some of the puts and takes of the core business offerings, you've got optimization, measurement and publishing. Which segment has the most runway from here? And what are the key drivers that you'd call out?

Yes. So we've enjoyed very strong growth across our measurement and optimization and now even our publisher offerings. On the measurement front, we've recently launched our prebid social offering, which on Meta, and that is expected to support our growth in 2025. And another area of growth in terms of runway on the optimization side. Well, the prebid, the prebid social on Meta will drive the optimization front. But going back to measurement, we continue to see very strong growth on the social side. Social is 55% of our measurement revenue, and we've enjoyed strong growth as the industry has grown with expanded user-generated content, on the social platforms, and we have an industry-leading product that we've developed and launched called Total Media Quality. So still lots of runway there. And then on the publisher side, 15% of our revenue comes from our publisher offering. Just over half of that offering is our Publica business, which is a CTV ad server. And in the quarter -- in the third quarter, we saw 26% growth across our publisher business and that was really driven by our Publica offering and the expansion of our offerings as we're growing CTV is a very $30 billion-plus industry with 25% less growth, and we've benefited from the growth of the industry and our strong partnerships on the -- with our Publica clients.

Great. I want to zoom in on 3Q results and the 4Q outlook for a minute. So in 3Q, there were some slower volumes in retail and CPG. And then in 4Q, there were some puts and takes between the political cycle and product adoption. So maybe update us on what you're seeing today and your high-level expectations for what's coming in '25?

Sure. So we did see an overall volume slowdown in -- starting in the mid of the third quarter, particularly related to CPG and retail clients where we saw a reduction in their digital media budgets and some delays in their digital media budgets. As we moved into the fourth quarter, we saw a slowdown leading up to the election. And all of this has been factored into our fourth quarter guide. In terms of where some of the levers of growth that we're expecting in 2025, while we didn't guide for 2025, we did share that we are planning for double-digit growth in 2025. And what gives us confidence in that is some of the dynamics I talked about earlier around expansion and the launch of our prebid social optimization offering. We would love to talk today about the work we've done to really lean in to the opportunity to secure additional clients from Oracle. We've won 75 clients from Oracle and expect that to be a driver of 2025 growth. And then also, we had a robust year in 2024 of new products that we launched, particularly earlier this year, and we're expecting them to drive 2025 growth.

Lisa, a big development this year was Oracle announcing they were exiting the Moat business. And I think that you called out as you've gone after those clients, 75 -- over 75 customer wins and a 72% win rate with this former Moat clients. So tell us about the timing of the ramp on the business with those customers, the runway and upsell opportunities that you see there?

So I'd like to refer as the summer of Oracle. So Oracle announced in mid-June on their public earnings call that they were exiting the advertising business shortly after announcing that, they also share with their customer base that they were turning everyone off by September 30. So it was an opportunity for us to pursue the customers. Oracle had three business lines: Moat measurement, which is comparable to our core verification, viewability and valid traffic detection; Grapeshot targeting, which is contextual targeting, both contextual avoidance, contextual targeting. Those are the two businesses that we went after. The third business line was Audience Marketplace, and we don't play in the audience data space. In terms of types of customers, three types of customers, brands. So we announced a few of the brand wins on our quarterly call, Heineken and Emirates. Publishers, they had a robust publisher business, both U.S. and EMEA. Team did a nice job putting wins on the board think of publishers like bloomberg.com and then third, platform integrations like OpenX. I'm very proud of the team and our performance and putting a 72% win rate on the board in addition to securing over 75 wins. We also announced that in addition to pursuing the business, Oracle also had a reputation in the marketplace of having very strong talent. We hired over 30 Oracle employees this summer. These are Oracle employees who are sales, technical support, business analysts, and we also hired the global CRO, Marc Grabowski from Oracle, a highly seasoned sales executive, highly respected in the industry. He joined IAS in mid-September as COO. So there were wins both on the talent and the customer side in terms of where we see cross-sell and upsell of those three customer types that I had described, the brands is where we see real upsell in terms of integrating their core viewability, invalid traffic detection, but Oracle did not have a robust brand safety suitability solution. So that's the real upsell for us is on ensuring that these brands adopt Total Media Quality across all of their digital advertising, across all of the markets that they're running. Great example of that is Heineken. The second area where we see incremental runway is in mid-market. Oracle had a robust mid-market business. We define mid-market as spending between $200,000 and $1 million annually. It's also a superpower of Marc Grabowski. He has a long track record of building robust inside sales businesses. He did that at Yahoo!. He built the inside sales business over $1 billion. So we're also pursuing mid-market both in the U.S. and EMEA.

Let's talk for a second about your announcement of your first-to-market optimization product on Meta. Maybe tell us why you were selected by Meta? How big that opportunity is and what your early learnings are as you launch that solution?

So Meta is a very important social platform partner of IAS. We have worked with Meta for years. Everything that we've run across Meta, YouTube, TikTok, related to Total Media Quality is all post measurement and Meta announced with the three badge partners, us and our two nearest competitors that they were launching an RFP. They launched an RFP process, very competitive in May to select one partner to co-build a prebid social optimization product. In June, they selected IAS and we co-built this product and the way to think about it is, it enables brands at a prebid level before the impression runs to select prebid content exclusion. What that means is, what is the content like L'Oreal doesn't want to run adjacent to on Meta and prebid inclusion, what are the types of content L'Oreal would want to run adjacent to on Meta, built the product, launch beta, September. The product is now live with IAS in fourth quarter. We've seen nice adoption rate from advertisers, and we were thrilled that Meta selected IAS. I would say two primary reasons Meta selected us. The sophistication, differentiation of our multimedia classification technology. It is more granular in how we classify video, image, audio and text than our competitors. It's more granular, because we are classifying frame by frame that's every single second of a video that's running in a live feed. We process more data than our competitors, which means that we're able to feed all of the data we're processing back into our models, to train the models, to be more accurate, more efficient in their detection, their classification. And secondly, as I mentioned before, Meta, we have a deep strategic partnership with Meta very strong track record. And again, we were thrilled and honored to have been selected by Meta.

And in terms of size, Matt, you asked about the size. Today, we have over $100 million of measurement revenue from our social platforms. And given the pricing that we've rolled out for this offering, we see an opportunity to upsell from just -- even just existing clients more than 2x that $100 million of revenue and it's incremental from an optimization perspective.

A big topic in advertising and culture at this point, AI generated misinformation and deepfakes. Obviously, we're seeing and hearing more and more about that. How are you positioning your solutions to detect and address fraud in the marketplace, particularly these AI-generated misinformation campaigns that are out there? And then what are the products that you offer that are most relevant there?

So misinformation detection has been one of the top requests from our brands, especially leading into the U.S. elections. It was in such high demand. I kept hearing it over and over at the start of 2024. Spending time with the brands. They said, please get your detection live in the social platforms, leading into the U.S. elections. So based on their feedback because we build for our customers, we pulled that product. It was scheduled for the back half of 2024. We pulled it earlier into 2024. We launched misinformation detection on TikTok and Meta first and then we launched on YouTube right before the U.S. elections. The brands were thrilled. We launched it earlier, because of what it meant is we could train the models on misinformation detection, train them to become more and more accurate. In terms of how the technology runs and works is, it's both tech layer with human layer, the reality of misinformation is it's incredibly nuance. So you need a human eye to confirm what the models are detecting in terms of misinformation. I'll give a quick example. Let's say, hypothetically, there is a live interview on Russian television with Putin, running on TikTok. He's speaking in Russian, and let's say, hypothetically, there is a banner below in English that says Putin supports Biden's position on, let's say, Hamas. Our technology, real-time can classify what Putin is saying in Russian, can real-time classify the text that is running below Putin and marry it up and say what he's saying and what's showing in text is not matching up. We're flagging this as misinformation. And then at the human level, we work with a third-party company actually based in the U.K. called GDI, and a human as it gets flagged, we'll say, yes, I'm confirming that's misinformation. There's a real-time feedback loop back for the social platforms that we flag it and say, "Hey, we're detecting misinformation". And what it means for the brands is that their ads will not run adjacent to the content that we're flagging as misinformation. So we've seen tremendous adoption rate by the brands. As I mentioned before, the models are getting more and more accurate as they're detecting more misinformation. And it's a really important product that we have in the marketplace on the social platforms and the open web. In terms of deep fake detection, and again, this is a reflection of how the high caliber of our data science team, we are launching a deep fake detection beta in the open web. What we are detecting is face swapping. So face swapping is when I'll make it up. It's Taylor Swift's face and let's call it, Obama's voice, like it's just not an accurate or a human actually speaking, and we'll continue to train the models on deep fake detection.

Got it. It's very, very important work. CTV, obviously, remains an incredibly important topic for advertisers. Tell us about the latest trends that you're seeing there and what you think will need to change for CTV to be a larger place for your advertisers to go in the medium term?

So with CTV, we see CTV as a long play, tremendous opportunity. As I mentioned before, the two places especially coming out of COVID, where users are spending the majority of their time. I know, I see it in my house. social platforms and connected TV. With connected TV, we have, for example, viewability in valid traffic detection, running in platforms like Netflix. It's -- we launched with Netflix last year when they launched their ad support tier. We launched our solutions, but where brands are really interested is as CTV opens up the programmatic channel. And Netflix, for example, they were running just with Xandr, Microsoft's DSP, they opened it up to other DSPs, but we expect the floodgates to open up. It's going to take some time on platforms like Netflix, Amazon Prime, and that's when the brands will be really leaning in to our brand safety and suitability offerings, especially in programmatic CTV inventory. The other side of the coin of CTV is shortly after IAS went public in 2021, we announced the acquisition of Publica. Publica is a leading CTV platform where we are helping major OEMs like Samsung, which is an important strategic partner. We're helping them on multiple fronts. With Publica, we have a unified ad auction, header bidding, helping Samsung drive up their optimization in yield of their programmatic CTV inventory. We're serving the CTV ads on Samsung's platform, and we're also stitching the creative real time in Samsung, which addresses the frequency capping issue where consumers -- they often every time I get the cat ad watching any form of streamed content, I think I don't even own a cat, this is a poor user experience. So being able to stitch the creative real-time makes for a better consumer experience. We're thrilled with the growth that we've seen with Publica. Q3 highlight, for example, in our publisher business, we saw a 26% growth rate due in a large part to the tailwind growth that we're seeing with Publica.

Tania, as you look to capture budget share and new opportunities in '25, how should we think about margins and OpEx growth? How are you looking at headcount expansion into next year as well?

Yes. Overall, philosophically at the company, we're very focused on continuing to invest in the growth in the business. We've got a very large total addressable market. We're investing in product and innovation to drive top line growth. But at the same time, we've also been able to deliver on efficiencies to protect our margins overall. We reported 38%, record 38% EBITDA margin in the third quarter, and our full year guide in 2024 is at the midpoint, is 35%. As we move into 2025, it's very important that we continue to make these investments. Headcount is expanding. We really leaned into our hiring in the third quarter of 2024 with 30 plus Oracle hires, the new leadership, investments in R&D, talent, data science and sales and marketing. So we do expect that to continue in 2025. But overall, between driving efficiencies and investing in the business, we anticipate to have similar margins in 2025 similar to the midpoint of our guide in 2024 of 35%.

I'll address this to both of you, as you want to carve it up. But how are you thinking about M&A opportunities to further develop your existing solutions or extend into new areas of the ecosystem? And maybe if you could touch on capital allocation, frankly, more broadly.

With M&A, so since I've been at IAS, we've made four acquisitions, three have been tech tuck-ins. And the fourth, as I mentioned before, Publica, was a strategic acquisition. We put the lens of build, partner, buy on everything that we do. We have a really strong track record with these tech acquisitions to acquire technology that is additive to our existing technology, quickly integrate that technology to drive incremental profitable growth. A great example, the first acquisition we made when I joined was ADmantX, within 5 short months. We took the ADmantX technology, what that is, is text classification. The tech reads the text of a page like a human. We took that tech, embedded it into our tech stack and launch context control with the Trade Desk, that was March of 2020, which now has been such a tailwind, represents half of our programmatic revenue. So strong track record. And in terms of future M&A, that's that we apply the same lens, where look at technology that could be additive to our existing, that could shorten our time to market, that would drive value for our customers, looking in areas like CTV, like data enhancements, also the whole area of optimization performance. So we're always out exploring potential M&A.

And we have a very strong free cash flow. We're well positioned from a balance sheet perspective, with net debt, net of cash of only $8 million. So we have -- and we've been able to pay down debt with our free cash flow. So we have -- we're well positioned with debt capacity for pursuing acquisitions.

So we're about 2 years since AI exploded as this huge theme in the investing world and frankly, society more broadly. How important are these advances in AI to your business? I know that there are some very important initiatives you've been working on, but how have they materialized? How are those investments panning out? And what do you see as the path forward?

So science is in the name of our company. 30% of our engineering team is made up of data scientists. And many of the products that we've launched are powered by AI in addition to things like Total Media Quality; MFA, Made For Advertising. Those are clickbait sites, right? So the sites that are highly dense with ads. That's a great example of a product that we launched, high demand from the brands because the brands are looking for greater efficiency, greater ROI and being able to leverage AI to be able to detect these map or advertising sites drives value. Another way we've leveraged AI is with the rollout of Total Media Quality language translation. It's really important for global brands like Coke, the faster our technology can be launched across dozens of markets, dozens of languages, because an account like Coke, they have such a massive global footprint. They want the classification to be running everywhere. And when we first launched the product with TikTok, we leveraged open AI and we were able to switch on language translation going from a product offering, seven languages to within 24 hours, offering over 90 languages with high accuracy rate. So AI is core to our company, core to how we build product, and we'll continue to leverage AI.

Let's talk about international. I think one thing that's really standing out right now is kind of the difference in economic growth and even inflation between different markets around the world right now. So I guess, talk about your international footprint, the regions you plan to kind of double down in '25 and why?

So our international footprint is a big differentiator for IAS. It represents 30% revenue. We have deep roots in Europe and we've been in Europe and APAC for many years. We continue to invest in emerging markets like Latin America, Southeast Asia, India, and I'm surprised you haven't brought up China. And we announced last week that we've launched in China partnership with RTBAsia. And what it means is that we're offering invalid traffic detection, brand safety and suitability for our advertisers, large advertisers running digital media in China, and we're also able to offer our verification solutions for Chinese advertisers, advertising outside of China. So continuing to invest in international, continuing to build products that are global and scalable. And we're able to run our solutions wherever the Fortune 500 brands. One is to -- is really core to our business.

Maybe let's close out on a longer-term view. If you look at IAS over the next 3 years, what would you say are the top couple of highest priority initiatives and market opportunities that you think will determine IAS' success?

I would say a few areas, we will continue to offer what I call insurance. So that's protecting brand equity, brand reputation for the largest customers that we have, like a Coke, L'Oreal, Verizon, wherever they're running digital media, but also we're very focused on a shift of insurance to performance as we help L'Oreal run their brands adjacent to higher-quality media also helping L'Oreal drive efficiency, drive higher ROI and sell more L'Oreal product. That's really important to shift from insurance to performance. Continuing to leverage AI wherever possible to drive greater accuracy, greater efficiency of our classification detection, prebid social optimization, really important runway for our business, it's something the brands have been asking for, for quite some time, continuing to invest in CTV. And then also, we talked about in our Q3 earnings call data. We process a lot of data tied to the what and the where the ad ran. Media quality data is in high demand in the industry and we're now double down, focused on leveraging our media quality data and partnering up with third-party data sets and also first-party data from our brands.

Great. Well, that's a great point to leave it at. Lisa, Tania, thanks so much for being here.

Thank you. Thanks for having us.