All the disclosures that our compliance team has asked me to share. I'll stop sharing my screen. Again, most importantly, thank you, Melanie, from Criteo for doing this. I used to cover Criteo over the years, but I followed all the progress and it remains a very important read into just the broader advertising ecosystem and plays a very important role. Perhaps, given where the stock price is, I think there is a pretty significant opportunity provided, and I'm sure Melanie will provide us with their best pitch here. Just to set the tone and set the stage, perhaps maybe we'll just kick off with for investors who may still think of Criteo as a legacy retargeting company. How would you describe the company today, Melanie?
Maybe just talk about two or three biggest changes in the last 18- 24 months at the company.
And that's a perfect way to start. Thanks for having me. Criteo has evolved tremendously over the years. That transformation has accelerated over the past few years, and I would describe Criteo more as a commerce intelligence company today, not so more Retargeting is still an important tactic, but we're so much more than that. What that means is, probably in simple terms, we power intelligent commerce wherever it happens, across different channels, different touch points for the consumer journey. Think about shopping and commerce today being more fragmented, being more complex, and there is a need for some sort of orchestration and decisioning layer on top, that becomes more critical for that online shopper journey. That's where we think we have a competitive advantage. Criteo today really has two main segments.
Without diving into the specifics, I'm sure we'll have an opportunity to do that later. It's about performance media and retail media. In performance media, that's where we have expanded from just a retargeting tactic to becoming more of a full funnel, cross-channel and self-service platform, basically to help advertisers grow their business, drive performance for them, help drive revenue and customer acquisition for them. Expanded with broader portfolio of tactics across multiple channels and have been on a transition from managed service towards self-service. In a nutshell, that's performance media. That's an important part of our business. In addition to that, we now have an important retail media business. We're actually the leading independent ad tech retail media player. What that means is we power 235 retail media networks, basically emulating what Amazon has done.
We're the tech backbone for those retailers to drive advertising revenue. That's an important fast-growing part of the business. That's in a nutshell where we are today. Maybe adding to that, and we're going to touch on that, I'm sure, AI is very much part of the business, has always been, but importantly, today, that's a new channel for us where Criteo has been early and we have lots of exciting developments there. Lots going on.
Awesome. Thank you. I guess, yeah, the way I hope to structure the conversation is we'll talk about AI, we'll talk about agentic AI, we'll talk about kind of the data aspect in ad tech, and of course, we'll talk about new growth opportunities. That's how I'm structuring most of the conversations and having a common theme emerge. I guess when we think about AI at Criteo, maybe talk about where are you seeing a measurable impact on the P&L from AI today, and how would you characterize that 12 months from now? Just talk about kind of big picture AI-
Yeah
kind of impact on P&L.
Yeah, no. It does manifest across, I guess, multiple dimensions. Definitely in terms of revenue growth opportunities, product innovation, the performance we deliver for clients, and also some productivity.
So maybe I'll expand on a couple of examples here. It does expand our addressable market from the new channel that I was referring to. I'm sure you know where Criteo has been partnering with OpenAI. We were the first ad-tech partner.
We're on-demand partner for them, which means that this is a new powerful channel for our advertisers to, I guess, engage with consumers. From that standpoint, it expands the TAM, giving us access to budgets we didn't have access to before, including search budgets, which were part of more like, I guess, concentrated with Google. That's an opportunity for us where we've had exciting momentum for the past-
Three months now. New channel, that's an important one. We're also embedding AI in terms of new ad formats. That would be another tangible example to give you. Whether that's what we call conversational ads, basically the ability to engage within the ad and give you recommendations tailored to what you're looking for, which is exciting, and also sponsored recommendations within AI assistant chatbots that you can find on e-commerce sites, on retailer sites. That would be another example where AI has been, I guess, a driver of innovation and opportunities for us to drive more monetization opportunities.
I would also mention, I guess, more in terms of how the campaigns are being set up and launched or the workflows. We have launched what we call MCP, which is our Model Context Protocol. We were pretty early with MCP, which is another important innovation to make things more scalable, have agent-to-agent interactions to use our tools and solutions. That's something that has been important for us in terms of the roadmap to keep scaling what we do. And then obviously what you would expect more internally in terms of productivity, automation. Thats an important, we have multiple work streams around that across pretty much all functions internally.
Okay. I guess like new channels and the new AI-driven ad formats that you mentioned, how should we think about the incrementality from, or maybe perhaps the contribution from them and specifically your ChatGPT as a new channel.
Yeah.
How significant is it today? How are you thinking about next 12 months?
This is still early days, obviously, but strong momentum as you would expect. It's really a very powerful discovery surface for users to explore, compare options, and consider, I guess, discover new brands and products.
We're lucky to be there first and learning quickly and, for us, I mean, your question about the opportunity there, we've had great traction from both existing and new clients. We've also had tractions with some agency partners to test in that new environment and grow.
what you can do there. Incremental budgets, I touched on that earlier. The fact that we didn't have access to search budgets before-
Yeah
That's a great way of searching for consumers. For us, that's an exciting opportunity to tap more into those budgets that we didn't have access to before and expand what we can do from a discovery perspective. The one thing that I haven't mentioned yet that's pretty exciting is we launched our Criteo GO product, which is basically taking all the different growth pillars we had in performance media, the full funnel cross-channel and self-service aspect of what we do into one product. GO is our solution to very easily set up campaigns across multiple channels for the tactics or to achieve the marketing goals that you want in as few as five clicks. Very easy. And we are incorporating, we're adding ChatGPT into that environment.
Okay. I got you. I guess just on AI impact on P&L and perhaps on the cost side of things, as in, I think most of the investors were focused on efficiency gains. Last year, you guys already have a pretty healthy set of margins, and good level of profitability. Maybe talk about where are we in that cycle where you're applying AI to improve productivity of employees, engineers, sales, and kind of getting more from the same dollar.
Yeah. I think the way we're approaching that is it augments what you can do. In other words, with the same amount of people that AI tools deployed internally, you can just achieve more. That has been really the philosophy of making sure that we're taking full advantage of that, and that applies across functions, obviously our engineering team, but also product team, finance team, et cetera. Really augmenting what we can do is probably the best way to put it.
we're always dynamically managing those.
Just for people who are online, I forgot to mention this earlier, with every session I'm doing a poll of same five questions. They're pretty quick. You may see something pop up right now. Would be awesome if you can answer those questions. At the end of this entire six or seven sessions, I'll collate all the responses and share them with everyone, including speakers like you, Melanie. I guess, just with regards to what's your latest take on this entire token maxing and are people inside Criteo also kind of buying a lot of tokens? And how do you think of that leading to efficiency gains of engineers?
Yeah, I mean, listen, I think that's something where, of course, you're always monitoring and making sure you're making the best of what you have. I don't think it's been a specific issue for us.
Okay.
Yeah.
Okay, cool. I guess, switching topics, with regards to, you already mentioned briefly about MCP and just taking a step back. With Agentic AI, there is a lot of discussion that kind of Agentic AI or agents will kind of disintermediate some of the advertising workflows. How is Criteo thinking about that shift, where we probably in the future are in a world where.
Yeah, no, listen, I think.
headless software is there. Yeah
I think for us, it's been really an opportunity to scale what we do, I guess, very seamlessly. Maybe I think it's worth going back to MCP and explain what that is.
You should think about MCP as being a secure bridge between an AI assistant, an LLM.
An agent, and the Criteo platform. The client, the agency, they would tell the AI assistant in plain language what they want to do, like show me the campaign performance, set up a campaign, et cetera. That goes through the Criteo MCP, which means our MCP translates that request and is going to call the right Criteo API in a way that's secure and structured. That will let the user create, manage the campaign, monitor the campaign performance and think about it as like a layer or universal adapter between the AI assistant and our tools or data. Why it's important, it's faster answers, faster campaign set up, less manual work, more optimization with the AI assistant. Our MCP let the AI agents use our tools and our platforms in a way that's-
I see
secure and effective. That's important because, especially in our conversations with agencies, that just helps make our tools and solutions accessible in a way that is very scalable. That's been a very exciting. Think about another way for us to do more of self-service-
versus managed service, that's less UI heavy, more but very much API first and having that layer underneath. That's something we've been talking about for a little bit over a year now and-
Okay
continues to be front and center in agency discussions.
Okay.
We actually mentioned the first campaign with them last quarter, and we have more this quarter.
Okay. Is that bringing in new Dollars, or is that bringing in new actors into the Criteo ecosystem? Is it shifting Dollars from one bucket to another? How do you view kind of the adoption of MCP?
I think it just simplifies the workflows. That's really the logic. I would say it's just another way for us to deepen partnerships with agencies, for example. It just makes our tools easy to use.
I see.
in an agentic environment.
When would you say that? Are we still in the early kind of pilot phase of this adoption curve, or are there more full kind of deployment budgets and now it is just more kind of a game of penetration?
No, I think it's still early days, but good traction and yeah, we're moving along nicely, but early days in the grand scheme of things.
Okay. This comes up, since you mentioned agency a couple of times, but this comes up in investor conversations that kind of some people would argue that AI agents could eventually do more and hence disintermediate human in the loop as such. Is there a risk? If there were a risk, it kind of straight goes towards kind of agency buyers as such. Is that something that, from your experience that agencies are worried about and there could be some hedging here or how they use kind of these new products?
Yeah. None of this applies to us. Listen, it's meant to be like working agent to agent.
Okay.
We're ready for any scenario.
Okay. I got you. I guess, switching over to data advantage, maybe level set. There used to be kind of in the legacy retargeting world, you clearly had a data advantage. How does that translate into the new kind of AI first advertising world?
Yeah
kind of data advantage does Criteo have?
Oh, absolutely. Yeah, no, you can have the best AI. If you don't have the data to train your AI, then you're-
Yep
missing an important part of the picture. Our data advantage comes from, I guess, the scale, the granularity and the normalization of our shopping data.
By that, I mean we work with 17,000 clients globally. We see billions of daily active users across multiple channels. Our product catalog is over close to 5 billion SKUs now. We do have scale of data and when I mention normalization of the data set, it's like we have a curated view of the relationship between people and product, right? Which is very foundational insights for us to predict intent and therefore deliver then the right product recommendations in-
advertising. That's the data foundation that basically everything we do is leveraging that important data at scale. The precision of it is really what's important, right? If you think about all the changes over the past couple of years with LLM platforms specifically. Most of what you get in those platforms is based on publicly available product descriptions or data that's publicly available to anyone, so not differentiating. Our data is more granular and knows at the product level lots of specific characteristics to the product, what's in stock, the latest pricing, in which location. It's a lot more granular view of what you need to know about that product specifically, what's trending this week, what people are buying right now.
That enables us to provide more precise recommendation that are going to end up performing better. It's been core to how we've been approaching partnership discussions, frankly, how we operate our business, and something that remains different, like a core asset to Criteo that-
I see. I guess like-
What we do. Yeah.
Yep. With regards to how does that build, having that data and having that kind of contextual information, how does that build upon incremental return on ad spend that you may be able to showcase? Can you give us a couple of examples around why having that data today is much more valuable than what it would have been a couple of years ago?
Don't get me wrong, it has been, over the years, very critical to our ability to drive performance for clients, right? You want to predict what people are going to buy, and that's how you serve the right ad at the right time with the right product for the right person. That has been very core to how we operate for years. But in today's environment, I think, lots of chatter around AI assistance and asking for product recommendations, and that can lead to frustration if you get products that are not in stock or you're not sure that that's the right pricing or that this is actually not based on the product review from six years ago. That's what I mean in that sense.
I see.
The integrity and the freshness of the data is pretty important.
I see.
Even more so today than it was before.
Okay. I guess just switching over to your retail media, commerce media offering, it has been one of the strongest growing or maybe strongest secular growth areas in internet advertising. What is the current state of the market, and is it like the first phase of growth is now behind us and now the next leg of commerce media is going to come from where?
No. Yeah. It remains one of the fastest growing part of digital advertising, and really on the back of Amazon reporting advertising revenue, I think for the first time Q4 2021.
Yep.
That has been pretty eye-opening for, I guess, pretty much everyone else to be willing to build high margin revenue stream from advertising, and that's where we come in. Today I mentioned 235 retailers, but to give you an idea of the scale, we work with 75 of the top 30 retailers in the U.S. We work with 40% of the top 50 retailers in EMEA, and we're growing our footprint in APAC as well pretty steadily. Basically it's kind of a land and expand type of dynamic. You win retailers, those retailer relationships are multi-year relationships, for the most part exclusive, and then it's a playbook where you have different ad formats that you can put on the retailer side, sponsored ads, display ads, shoppable video ads.
You have the ability to do that for your marketplace if you have one, and then the ability to expand offsite and monetize your data as well. That playbook is a playbook that's working really well across all those retailers. What we do at Criteo that's pretty important to help those retailers grow is we bring demand. Today we have
15 API partners that are an important part of our demand strategy. We work with all the major holding companies. We have 4,100 brands working with us directly and really solving for fragmentation in that landscape and making retail media easy to buy. That has been
I see
continues to be the playbook. Our media spend in that segment grew 30% in Q1.
continues to outpace the markets and, yeah, lots of opportunities ahead for our retailers.
Okay. I guess, you have announced partnerships with DoorDash and a few others.
Yeah
How should we think about those partnerships, and where do you add value?
Yeah
the growth algorithm there?
I'm actually glad you mentioned this one because it tells you that we're working with a lot of traditional retailers, but the playbook expands beyond traditional retail across what we would call adjacent verticals. DoorDash is a perfect example of that. DoorDash also illustrates the value of the demand we bring because that partnership specifically is about Criteo bringing demand, incremental demand to DoorDash to complement what they already have or continue to drive themselves. They see value in the demand we bring, and that helps continue to grow their business.
We had launched that relationship last year. We recently announced an expansion in Canada.
That's an exclusive demand partnership it's a multi-year exciting opportunity for us.
Okay. What types of incremental demand are you able to bring into a company like DoorDash?
Yeah. It's complementary from a brand relationship perspective.
Definitely have relationships with a number of brands, and we have built that strong demand base. I mentioned the 4,100 brands and the agencies we work with. We also have the 15 API partners that we work with. It's a pretty broad demand footprint that we have.
I see
that we can bring to our partners and just complements nicely with what they have.
Okay. I know you mentioned Criteo GO, your self-service offering.
Yeah.
Maybe talk about.
What it is?
why is it strategically important?
Yeah
You've always had a balance or more tilt towards managed. Now it is more.
Yeah
self-service. Maybe talk through what is going on there.
Yeah, no, that's a very exciting milestone for us. We launched GO at the end of Q1, as planned, and maybe to take a step back, we've been building on something we've done over the past year, which is transitioning existing clients towards the self-service workflow, especially important for the smaller advertisers.
What we've seen over the past year, as the proof that this is working for us, is that when you transition to that self-service workflow, you tend to have, on average, better return on ad spend thanks to all the optimization that goes with it leading to higher spend and lower churn. We have done that across, I guess, thousands of clients. We mentioned in the U.S., two in three campaigns for performance advertisers are already on Go. Building on that, we opened that to net new clients at the end of Q1. Excited to, getting good traction, excited to continue to roll that out. We have a pretty robust go-to market plan around that. We're going to continue to build on what we have. I mentioned adding ChatGPT inventory in there, in addition to
the cross-channel setup we already have. By the end of the year, CTV. Really good for us to take, I guess, all the important pillars of the growth, I guess, the broad pillars to re-accelerate or re-energize that business
in a way that's scalable. Self-service is just by nature scalable. For us, really, frankly, it's been a way to The important part of that, of self-service offering is that you can address the SMB segment in a way we haven't been able to before, and just that alone is pretty powerful. When you're managed service, it's less scalable. Self-service is perfect for us to address the SMB segment, which we know is a big TAM, and we hired the right talent for us to address that. Courtney MacConnell coming from Google. She was running Performance Max. Think about
Go as competitive product for Performance Max, Advantage plus, et cetera. We have the right setup now to go after that market.
Very cool. I guess if you're wildly successful with Criteo GO, is this something that can be half of your business in the next couple of years? Or how do you see the ramp in the adoption, and the impact it could have?
We're pretty excited about the potential of that offering, right? It's meant to be very easy to use.
We know it's working because it's building on that year of transitioning clients internally towards that workflow.
Pretty excited about the potential there. We'll report on the progress.
Okay. I know you guys work with ChatGPT and from the outside in, we are not sure how important is ads for them and having building partnerships with important ecosystem people like you.
Yeah
is going to drive more dollars.
Yeah.
Based on your experience, how important can that channel be for you and even for ChatGPT?
For them? I think it's going to be important for everyone, is the short answer. No, listen, as well as we do the monetization has to be a big part of their strategy. They need the revenue. They're spending a lot, so.
Yeah.
They have ambitions to, or reported to have ambitions to go public. All those things are going to continue to remain a key part of the financial picture for them. That's our core belief. For us, I mentioned the TAM expansion we get from that. We're very well-positioned as the first partner. We've been a big part of their international expansion. We started in the U.S. initially a few months ago, but we are now live in a number of countries, including Australia, New Zealand, Canada, the U.K. just launched, and soon we'll have Japan, Korea, Brazil, Mexico. Criteo is a scaled player, and we're a good partner for them to continue to scale what they launched as quickly as possible and very complementary to what they can do themselves. It's been a good partnership. Yeah.
Okay. All right. Cool. Some specific questions, tactical ones. I know you're going through re-domiciling the company. Maybe talk through what do you want to tell investors? What is it? What is the strategic rationale and the timetable?
Yeah, absolutely. No, that's actually a very important corporate development for us. We're excited to re-domicile eventually to the U.S. in basically a two-step process. We're expected to re-domicile in Luxembourg early in Q3. That's the first milestone. The second milestone would be the U.S. as early as beginning of Q1 2027. That's coming up. Main rationale for that, we want to be easier to invest in and a couple more specifics on that. Together with Luxembourg, we're moving to a direct listing of shares, replacing the ADR program that we have currently. That's going to be easier. When we get to Luxembourg, we're lifting the restrictions we have around share buyback. That's really something that has been, I guess, we've had some specific rules being a French company. That's no longer going to be the case.
We're going to have a lot more flexibility on that front when we get to Luxembourg. When we get to the U.S., we'll be basically eligible for index inclusion in a way that we're not today. That should make us part of the benchmarks, and again, easier to invest. We're pretty excited about that, and this is all moving according to plan and coming up pretty soon.
Okay. There's a question in the chat here on the timing of your share buybacks. Can you just clarify what you have mentioned on the timing and how the re-domicile process impacts buybacks? I think you addressed it, but just maybe repeat.
Yeah. Maybe to just elaborate a little bit. When you're domiciled in France, you can't own more than 10% of your treasury shares. That limits how much you can repurchase.
Oh, okay.
That's the current situation. We're maximizing what we can do based on those constraints and continue to repurchase what we can. When we get to Luxembourg, we no longer have that 10% limitation. That's the nuance there. That, yeah, that's I guess part of.
Okay
what we are contemplating when we get there. I would just add to that, the buyback has been an important part of our capital allocation strategy. That's not going to change.
Okay. Last call for people who are still online. If you want to answer those poll questions, I want to receive a lot of responses. I'll collate all of them, as the week goes by, I'll share with people who have signed in. Again, Melanie, with regards to just another big picture question is just new growth areas. You talked about retail media. You talked about performance media, Criteo GO, self-service. Beyond that, from a secular growth potential in internet advertising, what are you most excited about? Maybe it could be all three of the above or.
Yeah, I understand that.
Something that you.
Yeah. We definitely have a pretty robust roadmap across all the things that we discussed today, right? The objective is really to continue to grow our retail media business faster than the market has, we have done so far and reinforcing our leading position there. I think that part is important. For performance media, that's a segment where we're very focused on re-accelerating the growth.
from where we've been over the past few years. The different pillars we discussed today are what we're focused on to get there. That's maybe those are the most obvious ones. AI and continuing to innovate there and grow what we've put in place, ChatGPT, new ad formats that we talked about for sure. Then maybe the one that we haven't talked about is CTV, where CTV has not been a big part of the strategy in the past, but CTV is a channel that we truly believe, or Michael, our CEO truly believes that can be a performance channel. We're very focused on performance. We drive performance for clients across channels. That's, for us, an opportunity where we're spending more time. I mentioned briefly earlier that that's something we want to incorporate into the GO product by the end of the year.
We have a dedicated team working on that.
That's an opportunity for us. That's also pretty exciting, given it's obviously an important part of digital advertising where that continues to grow.
Okay. Fantastic. There's actually one more question in the chat. What's the main learning since launching OpenAI 3 months ago? Anything more you wanted to add? What's been the main kind of, if you had to iterate on the product or
Yeah
anything that you would share?
Yeah, maybe I'll share that it's a very, I've called it a powerful channel. What that means is that it's a very high-intent channel. We've seen that it's a channel where we're seeing 1.5x to actually 2x better conversion depending on the category you're in. That's a very powerful channel in that respect, right? If you think about when you do search, you search with a keyword, you have a very short query that you put there. In a conversation in ChatGPT, you get a lot more context around what the user is looking for, and you end up getting very high-intent users. When you click on that ad, those ads, the clicks rates are higher and, yeah, those tend to convert pretty well. Powerful channel in that regard, which is important if you think about the longer-term opportunity-
Yep
beyond just the initial test and scaling the channel.
Okay. Fantastic. I know we are budgeted 35-40 minutes. We are at 40 minutes, I want to be careful of your time, Melanie, and everybody who's logged in. Just to end it on a quick light note. Quick word association, Melanie. First.
Ooh
Without blinking, first thing that comes to your mind, and without any kind of layer of thinking as such. We are in an AI bubble. Yes, no, maybe?
No.
Okay. I like that.
It is exciting.
Walled gardens are losing market share. Yes, no, maybe?
Maybe.
Maybe? Okay. ChatGPT.
Depends how you define walled gardens, right?
I know. It's like.
Everyone has some walls somewhere.
Yes. ChatGPT ads?
Very exciting.
Of course, I have to ask you.
Discovery. Yeah, I would say discovery, too, is kind of how we see those.
Okay. Criteo in 2027?
Exciting. I mean.
Okay. Bullish.
Yeah. I put exciting everywhere. No, I mean, yeah.
All right. Cool.
2027 is going to be a good year for us.
Okay. Fantastic. This was great.
I should say U.S., just to tie that back to
Oh, exactly.
We're going to be a U.S. domicile, U.S.-listed company.
Yep. Yeah, I think, in my opinion, that in and of itself is a pretty significant catalyst.
Yeah, that's an important one we've been working on for some time, so we're happy we're getting close to it.
Great. Awesome. Good luck. Thank you for spending the time with us.
Thanks.
Again, thank you for everyone who joined. We have a session tomorrow on LiveRamp M&A. Feel free to join. Thank you, Melanie.
Thank you
have a great rest of the week. Thank you.
Sounds good. You too. Bye.
Bye,