All right. Okay. Thank you all for being here. My name is Matt Cost, Morgan Stanley US Internet team. Very pleased to be joined by Michael Komasinski, the CEO of Criteo. Thank you for being here so much.
Thanks, Matt. Appreciate it.
I have to quickly run through the disclosures. For important disclosures, please see the Morgan Stanley research disclosure website at morganstanley.com/researchdisclosures. If you have any questions, please reach out to your Morgan Stanley sales representative. All right. With that out of the way, maybe for those in the audience that are newer to the Criteo story, maybe give us a quick overview of the company, where you fit into the ad landscape and the opportunities ahead and maybe what's changing?
Sure. Yeah. Criteo is more than an ad tech company. We increasingly position the company around commerce intelligence and decisioning. Really, it's the sort of AI-driven commerce platform that's built to sort of power the next generation of shopping, right? Which, as we all know, is changing rapidly by the week, it seems like sometimes these days. In terms of where we fit into the ecosystem, we sit, like, right in the center of the commerce ecosystem. We sit in between brands, retailers, agencies, who are all trying to drive commerce outcomes across an increasingly fragmented ecosystem, and our cross-channel orchestration and decisioning capabilities help them do that, right? A couple of, I think, facts to ground the scale of Criteo for folks that are new to the business.
We see over $1 trillion of commerce transactions per year, or roughly $3 billion per day. We've got about 17,000 clients globally. We have a normalized SKU catalog of over 5 billion SKUs. And in terms of the daily active users that we can reach, it's actually 750 million per day, and actually extends up to 3 billion+ when you include the social channel that we now have access to. It's a really scaled business that in some ways it's bigger than the size of the company from a revenue and profit standpoint. In terms of where we're taking the business, I try to always simplify it down into really three things. One, we're pushing hard on agentic. We think this is the most powerful development in this ecosystem and maybe ever.
We'll come on to some of the recent events that underpin our focus on that. We have a performance media segment, and we are focused on really re-energizing that business and scaling it through three pillars that I think we'll come onto in more detail, which is full funnel, cross-channel, and self-service. Then in retail media, which is the fastest-growing segment of the media landscape, we've developed an industry-leading position with 235 retailers globally, and we are trying to continue to double down and accelerate that leadership position by bringing more monetization opportunities to our retailers, bringing in more demand partners, and just generally continuing to scale that segment to the market. Three really powerful pillars, and, you know, it gives us a lot of optimism about where we're headed.
I think you're about a year into the CEO role.
Mm-hmm.
Maybe, you know, in light of all of everything you just went over, what are some of the key learnings about Criteo that you've gleaned in that first year?
Yeah. Well, you know, I would equate kind of learnings and hypotheses because they're kind of the same thing in hindsight. I had sort of four hypotheses coming into the role. One, I had a belief, and I think it was true, right, that you can't do really anything meaningful in AI if you don't have scaled data assets and deep tech talent that is oriented around algorithmic and optimization-type programs. A new agentic would be big, although I certainly could not have predicted where we've landed at this point a year later, right? That all sort of goes in that AI and agentic hypothesis. Two, there's a convergence of brand and performance advertising, the benefit of that convergence disproportionately accrues to the performance platforms and advertisers in the ecosystems.
A lot easier to move up into biddable and addressable brand formats than it is to try to move down into performance. Criteo is certainly well-positioned there. I believe that the addressability landscape would quiet down. As we've all forgotten about the, you know, cookie deprecation issues and things like that. As a someone who's worked in performance and data and identity for, you know, over a decade, I knew that was going to get resolved probably in a pretty sort of quiet and non-disruptive way, and that did turn out to be true. Last but not least, you know, category leadership matters. that head start that Criteo has in retail media, I figured was something that could be built on and accelerated.
All four of those things have basically happened in this past year, sometimes in ways I couldn't have predicted. Those were my hypotheses coming into the role, like even when I took the job.
Maybe zooming out to the market backdrop for a second. I think on the last call you mentioned that, you know, trends remain stable through Q4 and the holiday season was solid. As you enter 2026, what are we seeing in the ad markets to date? Are there pockets or geographies that are coming out particularly strong or where there are headwinds? What's top of mind in your advertiser conversations as they, you know, plan for the rest of the year?
Sure. Yeah. As you said, we did have a pretty solid holiday, you know, in our guide for Q1 indicated a little bit of softness in U.S. department stores and some year-on-year comps for our Asia Pacific region. But the macro is relatively stable, current events, you know, notwithstanding, and, you know, those would not be material to us right now anyway. But, you know, as I look ahead to the year, we're really trying to sort of get past some of the headwinds that we had announced last year with a couple of our retail clients and just really focus on driving the underlying business as we work past those comparatives. And, you know, I think that we're on track for that.
Performance media continue to accelerate across the year. Retail media, same thing. Product rollouts are very much on schedule. We think we're continuing to gain momentum as the year progresses quarter to quarter.
I think you touched on this a little bit in your opening remarks, but just to put a finer point on what your top priorities are for 2026. As you think about what's driving strength across the business, where are you leaning in?
Just kind of go back to those, maybe those three pillars in a little more detail on those. In agentic, that really I'd say shows up in three main ways. One, the ChatGPT pilot that we announced on Monday. Two, the product recommendation service that we announced a couple of weeks ago, which shows the power of our data set and our ability to create high-fidelity recommendations off the back of that. Third, the agentic enablement of all of our products. We are, like, quickly deprecating most of our user interfaces in favor of MCP-enabled APIs that make all of our products much easier to work with, more scalable, more accessible, and we think that that drives adoption and stickiness.
That's kind of the agentic bucket, and it's got a couple of really big initiatives inside of that. Inside of performance media, we talk about that trio of full funnel, cross-channel, self-service. Full funnel and cross-channel go together, right? They're two sides of the same coin, right? Competing for mid-funnel and upper funnel budgets is really the idea of the full funnel, and that convergence of brand and performance is the industry tailwind that enables that. Cross-channel really is the diversification of the supply base to support campaign optimization as we gain access to those budgets. You kind of can't do one without the other. To be clear, you take budgets in the funnel, and you enable them with a more diverse cross-channel supply base.
Self-service, that is again about making the product easier to use and easier to buy. We have a flagship product launching at the end of this quarter, so just in a couple of weeks, called Commerce Go, and that is a performance product that is focused on the SMB market, and it's full self-registration. Five clicks to campaign is sort of the tagline. It enables small and mid-size advertisers to access the power of Criteo's cross-channel performance engine literally with a handful of clicks, and they can be running campaigns cross-channel and have all the tagging infrastructure done automatically for them. Even the creative gets pulled in automatically. It sort of tails off of that trend of I call it the democratization of advertising.
This is our, you know, foray into that. I think we definitely have a fair share of market share to take with that. Really excited about those.
I want to follow up on Commerce Go, but first let's go back to the ChatGPT...
Yeah.
partnerships. I mean, that's really exciting news that just was a couple of days ago. It was this week. Tell us a little bit about what that is, what the opportunity is, and you know, what it means for Criteo?
Yeah. Here's the simplest way to describe it, and then I'll go into a little bit of detail, is ChatGPT scales discovery and Criteo optimizes downstream cross-channel conversion. Like, that's the simplest way to sort of like think about the relationship and who does what in the context of a customer journey. How it works practically, you know, we've been invited into this pilot program as an API partner. We're now integrated into that API, there's a tech integration where advertisers with Criteo now, we can pass campaign information to ChatGPT, where they then can determine contextually relevant opportunities to surface that ad. That will allow our advertisers to be discoverable in contextually relevant situations. You know, discovery really is like kind of the new growth area in the industry.
You know, what clients get for that besides more access to that important surface is the convenience of just working with Criteo. We already have MSAs in place. We've got data pipes in place. We've got agreements in place. Now we just turn that on, and we let some of that demand flow into ChatGPT as those contextual opportunities arise. Our clients can benefit now from additional discoverability. In, like, the fastest-growing platform on the planet. Really excited about it. The inbound interest has been unbelievable since Monday. Our sales and partner teams are just having a lot of really great conversations.
Sure.
It's been good.
No, that's really exciting to be kind of the first ad tech partner there.
Yeah.
It will be interesting to watch how that develops. I guess going back to Commerce Go, I mean, you gave a little bit of an overview of what it is, but I guess, what is the cost to serve difference for Go campaigns versus traditional managed service? How should we think about the revenue opportunity as you launch the full self-service, I think you said just a couple of weeks?
Yeah. Here's what's interesting about Commerce Go. We've been transitioning existing clients to Go since late last year. What launches here in a couple of weeks is the net new customer acquisition program. We learned a lot of really important things as we were transitioning existing clients. We found that they churned less, they spent more on activated media spend, so they increased their budgets. They're getting 20% better performance. Like, those are incredible statistics, right? It gives us a lot of confidence that the product, like, really works.
Now it's all about how effective are we in reaching potential customers, and in some ways, doing the work that we do on behalf of our clients to go out and reach new customers, get them into our sales pipeline, and hopefully get them onto the platform where that ease of use and the results will speak for themselves and hopefully drive incredible results for them.
I guess on the performance media side, I think you grew 2% in that business line, constant currency in the fourth quarter. Commerce Go was strong. I think retail was an area where that was slightly more of a headwind. Going forward, how are you addressing kind of those different category trends? Can you help us understand the offsets from Commerce Go and CTV versus retailer category?
Yeah. Yeah, sure. Here's how I would sort of like back up and look at the business right now. Performance media as a segment is kind of in re-acceleration mode. I think before I arrived, the company was dealing with some of those addressability concerns across the ecosystem. Those settled down early into my tenure, we were able to really press on product expansion and roadmap acceleration. That's really where we are right now. It will take a couple of quarters for that to show up in the results, but Commerce Go will be a catalyst to that. We'll be rolling out a new discovery product, at least in alpha and beta in the first half of this year, hopefully general availability in the second half.
Of course, the ability to invest client performance dollars in the agentic channel, we think will drive interest and performance for clients. That's in re-energizing mode. In retail media, you know, we announced a couple of scope reductions on some large retailers last April. We had to announce those, you know, pretty far in advance before they actually happened. Now we're in that period of working past those comparatives. That will happen across the balance of this year. It does start to tail off, especially into Q4. Really, we're just focused on driving the underlying business. Continuing to scale the auction-based display product. We launched a conquesting product actually two months early, so that's in market now.
We continue to make significant investments in our Commerce Max platform to drive the demand side of that product set. New insights module, new cross-retail campaign support, things that make our tool easier for brands to buy the 235 retailers that we represent. That really is like focus on the underlying business, let the comps come off as we progress through the calendar, and hopefully come out here at the end of the year and show, like, a really strong business in both segments, with some other, you know, simplification of the business that I think we'll probably come onto later in this talk as well.
Yeah, of course. Maybe shifting to AI and agentic commerce. You've obviously been active in the early stages of agentic commerce. You've talked a lot about your commerce or your agentic commerce recommendation product that you're working on. Some, you know, some really great stats on that in the early stages. How are you thinking about monetizing that product? 'Cause I believe it's not part of the ChatGPT partnership. How are you thinking about monetizing it as you expand surfaces that it can touch? And what advantages does Criteo have when it comes to developing that recommendation service?
I'll start with the monetization question first. That path is not determined yet. It could go a number of ways. It could show up in a SaaS format. It could be some kind of a per query basis. The beauty of that offering is that it can be monetized in both ad and non-ad formats, right? Because the sort of target market for that is going to vary in terms of how they're thinking about monetizing use cases that have product recommendation at the core of them. What we're focused on right now is proving effectiveness and driving partner engagement and frankly, industry awareness. We've had this strong thesis that we have the best commerce data set in the industry outside of Amazon.
You kind of need to prove it. The commerce recommendation service is the attempt to do that. It says, look, you can enter product queries into this, see the carousel that you get back, analyze that for accuracy, fidelity, relevancy, put that up against any other carousel that comes back from any other platform, and it's better hands down. In fact, if you were to commingle that service with, say, an LLM environment, you get a 60% uplift in the relevancy and accuracy of the products, which is powerful. We think that as not just LLMs, but other platforms that are gonna be competing for daily and monthly active users, where product recommendation is a use case that drives a differentiation, that makes us an attractive partner, therefore then there's a path to monetization from that.
There's 2 leaps of faith, but, I think standing up the service and really being able to put it out there in full general available form shows the power of that conviction. It will lead to some interesting places.
Yeah. I mean, it's not hard to imagine how it could be commercialized. Obviously, Criteo is in a unique position given the data that you have access to and the other products that, you know, I think are relevant to the process of developing it. I guess, what are the milestones investors should be watching to get to the point of saying like, "Oh, this is really heading to something that's gonna be commercializable at scale"? What are the milestones that you're watching?
Yeah. We've been trying to provide some disclosure on some of the testing activity that's happening with large partners, so we'll continue to update on that. After that, it really comes from those testing efforts. I think by being really public about it's also starting to create other inbound interest in it. I guess it's really about testing updates and then what other types of, you know, partner activity do we wanna, you know, do we get and that we wanna report on. Those would be the near term milestones.
Got it. Maybe, talking about competitive positioning for a second. You know, as you position Criteo as this AI-driven commerce intelligence and orchestration platform, how do you differentiate from competitors, who are also investing heavily in AI and in commerce data? What are the unique advantages of Criteo?
One of the things that we've not talked about is the cross-channel setup. We're one of the only platforms out there that has the ability to hold performance constant across channels. There just aren't very many or frankly any that I can think of that have that same capability. That's the way that marketers want to market. Marketers don't want to be channel marketers. You lose audience fidelity, you don't have frequency capping, you have all kinds of problems that occur when you get locked into sort of channel investments to run media. The cross-channel setup is how marketers think about it. I wanna run discovery campaigns to find, you know, to attract people to my brand for the first time.
I wanna run customer acquisition campaigns for people that are giving off some kind of a high-value behavior signal. I wanna run remarketing campaigns to capture demand. I want to run those things cross-channel seamlessly, where I'm getting efficiency in how impressions are served, how audiences are served, and how I measure. We map better to how marketers want to market, and that is unique. Walled gardens by definition are walled gardens. That ability is sort of deep in the Criteo DNA of being a performance marketer for the last 20 years. We learned how to identify identity in low signal environments. We learned how to do attribution across fragmented customer journeys. All those are sort of IP that create that cross-channel setup. That is unique and differentiated.
Hopefully with Go, we can start to scale that to a new client segment and I think really demonstrate that, as well as expansion into channels like CTV, and then of course the agentic channel, which is how we think about it. It's just another channel that sits among others, and depending on the campaign type and the advertiser, it's going to be entitled to a certain level of investment to drive to an overall outcome.
A lot of exciting things to offer customers. I wonder if we could talk just for a second about your customer life cycle.
Mm-hmm.
You know, you've had changes in scope from some large retail media clients in the past, but overall, I think your client retention as of last quarter was over 90% still. Very, very strong. I guess, what's driving that retention rate and how is the process going of meeting these customers with these new offerings?
Sure. Yeah, we do have a high client retention rate, which is great. I think it comes from, you know, one, driving performance and being a valuable part of their P&L, right? Clients count on Criteo to deliver predictable, valuable results. That drives part of that. Maybe more on the retail side, it's similar, but we are an essential monetization partner. Our retail clients really count on us to not just give them tech that provides scaled, you know, ad serving for a retail network, but to drive the demand side of that as well, right? Criteo's proposition in retail is really that of a full business partner. We stand up the tech for you, but equally, we will go out and surface demand that actually puts revenue into your P&L.
I sometimes will call it like a turnkey P&L, which is really differentiated from other competitors that might just offer tech only. I think it's then just like more cultural things, like listening to your customers. We're tight with our 235 retailers, and we listen to what they want. We incorporate those things into our roadmap. We meet with them frequently. It's a very client-centric organization in that regard. Yeah. I think those things kind of all add up to contributing to that retention rate.
Great. Maybe on the process of re-domiciling. You're in the process of moving from France to Luxembourg, I think in the third quarter this year, with the goal of getting to the United States from Luxembourg in the first quarter of 2027.
Mm-hmm.
I guess walk us through the strategic vision underlying this move, and, you know, the costs, and benefits associated with it.
Yeah, sure. The principles or the drivers are relatively straightforward. In the French domicile setup, we're forced into an ADS structure for the equity shares. We had no passive index inclusion, and we had certain capital controls that limit share buyback activity. The move is really to address those three things, right? The move to Luxembourg gives us more flexibility on capital allocation, and then the subsequent move to the US will give us more index inclusion. We're very much on schedule for that. We just had an exceptional general meeting last Friday, where we received over 98% approval on all of the recommendations.
We do expect to land in Luxembourg in early Q3, and then there will be a subsequent vote and move to the U.S. that could conclude in Q1 of 2027. It's all about removing complexity from the stock and making it more complementary to the way all of you and the way the market buys shares in public companies today, which is at least partially through passive indices. Of course, flexibility in capital allocation should be table stakes.
Of course. Before we go to capital allocation, I do wanna talk about your P&L for a second. When you're thinking through margins, how are you thinking about the puts and takes between leverage versus investing in strategic opportunities like agentic AI, increased productivity? How should we think about OpEx discipline longer term? Do you see more efficiency coming in 2027 and beyond?
There's definitely a lot of puts and takes in the business on this right now. I think you could see it in the profit guidance that we gave for the year, right? In light of a $75 million headwind, we were able to guide to EBITDA targets that are not too far off a relatively high performance last year. What that shows is that we're able to drive efficiency in the business while still investing in accelerating the roadmap and in new channels like agentic. I would look for that to continue. Commerce Go creates a new lever that's kind of interesting in the self-service that reduces our cost to serve.
We might reinvest some of that efficiency on deeper service models for large advertisers. Equally, we could take some of that to the bottom line in efficiency or redirect it to maybe even more investment in agentic or into the retail roadmap. We'll figure all that out as we go into 2027. The good news for us is that we've got multiple levers to pull, right? We've got a self-service product that we think is gonna go well, that spins off efficiencies by its very nature. We've got a broad-based AI efficiency program that runs across the company that just makes us more efficient. Then, you know, we've got a nice tight roadmap for the two segments, so there's not any sort of wasted effort in terms of the investment that we put into those.
I would look for us to sort of be in and around that ballpark, but we'll get onto guidance for 2027 kind of when we get there later this year.
Right. You're pursuing this re-domiciling to the US, which creates opportunities, including buying back stock. In the meantime, you know, as we think through 2026 before those new opportunities potentially become available, how are you thinking about capital allocation this year, especially given the investments that you're making kind of on the operating side of the business?
It's always the same list and in the same order. We invest in the core business to drive organic growth, then we're looking for high-value opportunistic acquisitions. Then we return capital to shareholders through buybacks. That has been the philosophy since I got here. Sarah's good at repeating that often, it's in the same rank order. You know, I guess on the second one with M&A, we haven't done anything since I've been here, it's just we've been selective, right? We actually do have an active M&A pipeline. We look at all kinds of things that are on market or in some cases not in market, it's gotta be the right fit. Synergies have to be there.
It's gotta be the right profile for where we are as a business. The landscape changes quickly, so you also have to have, like, you know, the real conviction about how it fits for the midterm, and we do. That's sort of why we've done what we've done.
Great. Maybe we can close out with a big picture question. When you think about the conversations that you're having with investors on the topic of AI, what do you think feels like the most underappreciated opportunity for Criteo, and then maybe an underappreciated challenge that you're working on executing through?
Yeah. Yeah, I think the, I think the underestimated opportunity is the kind of incrementality of agentic. I think people don't appreciate. We tend to get into a lot of zero-sum game conversations and the incrementality element gets lost. I think you guys actually put a paper out late last year that talked about sort of the accelerated contribution of agentic to e-commerce overall, and we very much agree with that thesis. I think that is underappreciated even across this conference probably, that the power of those engines unlocks commerce that otherwise would not have happened in legacy search 'cause it was either too time-consuming, too inefficient, or price discovery was too opaque. A lot of that gets addressed with what I'll talk about then in my second point, which is the underappreciated challenge.
That unlock's not gonna happen unless agentic platforms have high quality data feeds feeding high fidelity relevant product recommendations with accurate product data in price and availability and other elements like that that make it high quality. Again, that core thesis of, you can't do great recommendation without good data, I think is still underappreciated. You know, time will tell, and we'll be proven out.
Great. We can leave it there. Michael, thank you so much.
Thank you, Matt. Thanks everyone.