How Criteo Is Bringing Commerce Media to OpenAI with Michael Komasinski

Summary

Podcast Summary

Marketecture: Get Smart. Fast.

Episode: How Criteo Is Bringing Commerce Media to OpenAI with Michael Komasinski

Host: Ari Paparo

Guest: Michael Komasinski, CEO of Criteo

Date: June 29, 2026

---

Episode Overview

This episode features a lively interview between host Ari Paparo and Michael Komasinski, CEO of Criteo, recorded live at an industry event. The discussion dives deep into Criteo’s evolution as a commerce intelligence platform, their partnership with OpenAI, the transformation of retail and performance marketing, and the implications of AI-driven commerce—touching on everything from practical implementation to future outlook. Listeners gain exclusive insights into how Criteo is positioning itself at the intersection of commerce, media, and artificial intelligence, with an emphasis on new advertising formats in Large Language Models (LLMs) and the broader impacts on retailers and brands.

---

Key Discussion Points and Insights

1. Criteo’s Evolution and Dual Business Model

• Tenure & Company Focus (02:06–04:04):
  • Michael Komasinski has been CEO for 16 months, describing Criteo as operating at the intersection of commerce, media, advertising, and AI.
  • “We organize and report the business in two segments: performance media and retail media... But the overall positioning is a commerce intelligence platform.” (04:04, Michael Komasinski)
  • Shared technical infrastructure connects both sides (performance and retail) despite different front-end products.

2. The Transformation of Commerce Media

• Transitioning to Retail (05:09–06:14):
  • Retail media continues to grow, but performance media is still the larger segment; interestingly, new products like Discovery Advertising (in partnership with OpenAI) blur the lines.
  • “Discovery advertising is not a performance or a retail product. That’s going to be both.” (06:00, Michael Komasinski)

3. Criteo’s Partnership with OpenAI

• Rapid Advertiser Adoption (06:30–07:04):

  • Over 2,000 advertisers are now live on Criteo’s OpenAI integration, doubling from 1,000 just a few weeks prior.
  • “It’s been one of the best things for our new business pipeline just at large. It’s been helpful to reposition the company away from retargeting.” (06:41, Michael Komasinski)

• Advertiser Experience and Capabilities (07:22–08:48):

  • Clients currently use mainly test budgets, expected to shift to larger spends for the holiday season or next year.
  • Key advancements: product feed curation, creative variants, and optimizing campaign inputs for the OpenAI API.

• Performance Optimization (08:42–09:24):

  • New to Criteo: passing “campaign hints” (extended keywords, conversational text) to help OpenAI algorithms target and optimize ads.
  • Budget completion and click-to-conversion rates are primary KPIs; early integration with client data feeds gives Criteo a competitive edge.

4. Budgets and the Evolving Digital Funnel

• Budget Source and Cannibalization (09:45–11:28):
  • The panel speculates new LLM ad budgets may come from both lower-funnel (search) as well as mid/upper-funnel (branding) budgets.
  • Notable Statistic: “80% of the referred traffic coming from paid advertising on OpenAI are new to brand.” (10:27, Michael Komasinski)
  • This high “new to brand” percentage may justify shifting brand-building dollars to LLM-based discovery ads.

5. The Future of Agentic and AI-driven Commerce

• Agentic Commerce (11:32–12:40):

  • Michael is skeptical of full autonomous buying (“staunchly against”). He envisions “ambient support” or heavy “assisted use cases”—like automated deal finders.
  • Interoperability across apps/platforms is key, but expects users will not roll their own agents; most will rely on integrated, behind-the-scenes helpers.

• Retailers & Brand Experiences in AI (13:16–14:38):

  • Retailers like Albertsons are moving their media offerings to chatbots, blending organic and paid listings within conversational assistants.
  • “If I feel like through the relevancy calculation I’ve got an opportunity to serve [a slot] as sponsored, I’m going to do it… I’m not going to lose a sale and I’m monetizing one of those slots.” (14:13, Michael Komasinski)

6. Brand and Retailer Needs from AI-driven Media

• Measurement & Incrementality (16:42–18:03):

  • Brands want more than just visibility; they need incrementality and sophisticated measurement frameworks to justify spend on new AI-driven surfaces.
  • “Most marketers believe that they’ve been over invested in Last Click ROAS for many, many years…. Discovery ads give them a new way to pivot out of that paradigm they feel they’ve been trapped in.” (17:24, Michael Komasinski)

• Funnel Compression (18:03–18:52):

  • AI/LLM interactions compress the purchase funnel, allowing shoppers to move quickly from awareness to transaction, exciting for marketers but challenging for measurement.

7. Retail Media Transparency and Control

• Ease of Use & Data Consistency (19:12–21:27):
  • Brands and agencies want easier cross-retailer buying and standardized, actionable data. Criteo’s dual position (sell and buy side) allows for better outcome-driven advertising.
  • “While they care about brand safety, transparency, and measurement… what they really care about is outcomes.” (20:46, Michael Komasinski)
  • Michael claims being less invested in UI is an unexpected advantage as agentic and prompt-based workflows become more prevalent.

8. Broadening the Commerce Media Concept

• Beyond Sponsored Products (21:59–23:34):
  • Criteo is scaling sponsored products and display, and investing in “page intelligence,” which helps retailers optimize both organic and paid listings dynamically based on user signals.
  • “Would I like to be in all those spaces? Sure. But right now, we’ve been focused on scaling sponsored products, getting into the display business…” (22:24, Michael Komasinski)
  • The company is prioritizing areas that are most adjacent and valuable to its retailer partners.

9. The Rosy (and Realistic) Future for Retailers

• Retailers’ Continued Importance (23:52–24:53):
  • Despite the rise of AI and alternative checkout experiences, Komasinski is optimistic:
    • “Retailers play a very important role in the ecosystem. Unless you’re willing to be in the fulfillment, customer service, returns, and joy of shopping business, it's tough to own a chunk of that commerce layer.” (23:52, Michael Komasinski)
  • He stresses that while retailers must innovate with best-in-class front ends (chatbots, AI, UX), their core functions (fulfillment, service, trust) are a durable advantage.

---

Notable Quotes & Memorable Moments

• “Discovery advertising is not a performance or a retail product. That's going to be both.”
  —Michael Komasinski (06:00)

• “80% of the referred traffic coming from paid advertising on OpenAI are new to brand.”
  —Michael Komasinski (10:27)

• “Most marketers believe that they’ve been over invested in Last Click ROAS for many, many years.”
  —Michael Komasinski (17:24)

• “If you can tick the box on [brand safety, measurement, transparency] appropriately, what they really care about is outcomes.”
  —Michael Komasinski (20:46)

• “As agentic and prompt-based workflow becomes the norm, you really just want to get to the back end.”
  —Michael Komasinski (21:27)

• “Unless you’re willing to be in the fulfillment business, the customer service business, the processing returns business, and just the joy of shopping business, then it’s a tough thing to try to own a chunk of that commerce layer.”
  —Michael Komasinski (23:52)

---

Key Timestamps & Segments

| Timestamp | Topic/Quote | |-----------|-------------------------------------------------------------------------------| | 02:06 | Komasinski’s tenure and introduction to Criteo’s positioning | | 04:04 | Breakdown of performance vs. retail business | | 06:00 | Transition products & OpenAI partnership introduction | | 06:30 | Advertiser adoption and the strategic impact of OpenAI | | 07:38 | Virtual demo of OpenAI advertising experience | | 09:45 | Discussion of budget sources and cannibalization | | 10:27 | “80% of referred traffic… are new to brand” statistic shared | | 11:32 | Skepticism and vision about agentic/automated commerce | | 13:16 | Bringing retail media into chatbots and shopping assistants | | 16:42 | Brand/retailer needs: measurement, incrementality, and channel planning | | 18:03 | Funnel compression and challenges of LLM-driven commerce | | 19:12 | Discussion on control, ease of use, and data consistency in retail media | | 21:27 | Move toward agentic/prompt-based advertising—UI implications | | 22:24 | Criteo’s expansion beyond sponsored products & priorities | | 23:52 | Future prospects for retailers in an AI world |

---

Flow & Tone

The conversation is insightful and rapid-fire, with industry in-jokes (retargeting terminology banter, UI complaints), candid CEO reflections, and a balanced mix of optimism and pragmatic realism about the future of commerce media and the impact of AI and LLMs. Notably, the speakers maintain a forward-looking, approachable tone, making industry shifts accessible while still drilling into the technical and strategic details that marketers and advertisers care about.

---

This summary encapsulates the full depth and color of the episode, and serves as a comprehensive guide for those who want to understand the present and future landscape of commerce media, LLM-driven advertising, and the evolving role of platforms like Criteo.

Transcript

[A]

Oh, hello, I'm Jeremy Bloom, co founder and CEO of marketexture Media. And boy do I have news for you. Market is back. And if you've been watching from a distance thinking that looks phenomenal, but a trip to New York is just a bridge or several too far. We've got great news for you. We're bringing Market live to the beating heart of Adland. That's right. On September 23, 2026, market coming to Chicago. We're going to be bringing the same sold out energy, sharp insights and industry defining conversations to the center of advertising's biggest transformations. From media and commerce to AI tech and modern marketing. This is where the people shaping what's next for the ad industry will be, want to be and need to be. Early registration is live now, so lock in your [e-mail] Again, lock in your ticket at chicago.architecture live.com Disclaimer I live in Chicago. It's not Chicago, it's Chicago.

[B]

Welcome everyone.

[C]

Hey Ari.

[B]

Hey, how you doing?

[C]

I'm very well.

[B]

So thank you all for coming to the beautiful, pretty o yacht. We're in the shade which hopefully is good enough for everybody. Got the fans, got everything like that. So we're excited. You know, I was talking to someone earlier about there's like a thousand panels a day, but most panels only have like two people at them too. Right. So I'm pretty, pretty excited about this turnout. Yeah, no, they must be interested in what you have to say.

[C]

Must be subscribers to Market Toxure.

[B]

I don't think we have this many. So.

[A]

So.

[B]

So you've been on the job for how long? A year, year and a half?

[C]

16 months.

[B]

16 months. You're not counting or any?

[C]

I'm not counting.

[B]

How is it?

[C]

It's great. Look, you know, it's a super interesting company. It's got great capabilities and we sit at the intersection of some of the fastest moving vectors in the ecosystem, whether that's commerce, media or advertising in AI, just performance and how that sort of form of marketing continues to evolve. So I mean those were the things that really got me interested in it and they've all turned out to be true. And then like along the way some interesting things have happened. And this will take you way back and show you how long those 16 months have been.

[B]

Okay.

[C]

I mean, privacy sandbox. Went away in my third month on the job or went out for privacy sandbox. Right. You know, agentic became a big thing. I think we all knew that was going to be big, but we didn't know exactly what the details and shape

[B]

of it would be.

[C]

And so that's become a huge part of the business.

[B]

I've noticed you're like almost weekly press releases on the subject.

[C]

Yeah, I do. I do talk about it a bit. It's good.

[B]

Keep hyping it up.

[C]

Yeah, yeah. Well, you know, and look, the Open Air partnership, probably one of our sort of flagship partnerships today. So there's been a lot that's happened. But again, that privacy sandbox one takes me like way back.

[B]

Right.

[C]

That was kind of hanging over this industry in a big way and then suddenly, poof, it was just gone.

[B]

So, yeah, I mean, we all dedicated years of our lives to it and now it's just a fading memory. Like last season was a dream. Don't worry about it. Right. Okay. Let's never talk about it again.

[C]

Okay.

[B]

Okay.

[C]

I'd like.

[B]

So. So I think Critio is a little bit Criteo. I always go back and forth. I apologize. But the. It's a little bit of an interesting situation because it's almost like two different businesses right now. It's what you. I don't know how you refer to it, but retargeting marketing side of the business and then the commerce side of business. Do you think about it internally as two sides of the same business?

[C]

We kind of live in two different modes. So we do organize and report the business to the street in the two segments we report performance media and retail media. So those are P and ls. But the positioning of the company overall is a commerce intelligence platform. Okay. And that allows us to have sort of a common theme to investors so that they kind of know like what space we're getting into. And I think that is more representative of the like addressable market that we think of. And then when you get down kind of into the tech stack, there are a lot of shared components between those businesses. So we might report them separately. But there are relevancy models that span both businesses. There are audience capabilities that span both, their measurement frameworks that span both. So, you know, they operate sort of front end products pretty differently. But some of the componentry that's the most important part actually is shared across. Not to mention the infrastructure that it all sits on. Do you go to market that way?

[B]

Like do you go to a big meeting with a retailer client and sell both or is it different people in the retailers?

[C]

I think we know what that meeting setup is before we show up. Okay. You show up and sell people what they are ready to buy.

[B]

That's A good approach. How, how far along are you in your transition to having to have, having it be. I don't know how you describe as a majority retail business or retail, the retail commerce first business.

[C]

Well, I think that's a little ways off yet. Right. Because the performance media business is sizable and so while retail continues to grow at a great clip, it's got some years before it would overtake the scale of our performance business. But we are starting to work on some really interesting products that actually span both. Right. And so like to pick up one of them with, with OpenAI. I mean discovery advertising is not a performance or a retail product. That's going to be both.

[B]

Right.

[C]

So we're going to have retail media networks that sign up to use that as an off site play and of course performance marketers that are interested in as just another form of performance marketing. So I mean there's sort of like duality in the way that we think about it.

[B]

Yeah. I was going to get to the OpenAI stuff later in the conversation, but it's so important and sexy. We should probably just. We'll skip ahead. We'll come back, come back to commerce media. Let's just talk about OpenAI. So yesterday you had a press release that I think you're up to like 2 or 3,000 advertisers that are doing

[C]

just over 2,000, which was up from like a thousand sometime in May. So it doubled in about four, five weeks.

[B]

Are they clamoring to be part of the program?

[C]

They are actually.

[B]

Yeah.

[C]

It's been one of the best things for our new business pipeline just at large. And I think it's been helpful to reposition the company away from some of the things like you use the word retargeting. Right.

[B]

I did use that word.

[C]

We don't, we don't use that one.

[B]

Google doesn't either.

[C]

Nobody uses.

[B]

Never happened.

[C]

But you know, like it just allows customers to look at you in a new light.

[B]

Yeah.

[C]

And certainly like that form of, of advertising, Discovery is a new addressable market for us.

[B]

Right.

[C]

We didn't do any search advertising before. It was the wall of garden and now we're able to play in a, in a set of budgets and in a part of the funnel that we were blocked out of before. So it's exciting for a number of reasons.

[B]

Right. So probably a lot of the folks listening to this aren't regularly logging into your interface and know the ins and outs. Take us through. What's it like for an advertiser to start buying an open air Is it, Is it new budget? Is existing budget, is a new UI to these existing creatives. Give us, give us a little more of a.

[C]

That's a virtual demo. Great question. So the virtual demo would be primarily test budgets today.

[B]

Okay.

[C]

And I think that that'll start to evolve maybe for Holiday, certainly for next year. Right. So it depends on whether Holiday creates that urgency and scale or whether we're sort of locked into an annual planning process that then takes its natural course as we head into 27. But the setup identifies all the typical campaign parameters you'd expect, but in particular around product feed and then creative variants. All right. And those are a couple places where we've developed like new capabilities on our side to make sure that clients are like really optimizing the inputs that go into the OpenAI API to ingest all that, essentially ensuring campaign success. And then we are starting to develop methods to optimize performance once campaigns are running. So we can pass sort of campaign hints that help their algorithm sort of pick up like when something's going to be relevant.

[B]

Like what?

[C]

Conversation, text, kind of hints. Yeah, right. I think of it as extended keywords.

[B]

Right.

[C]

But basically things that it picks up on to help with burn rate and shut it up, basically, as well as creative variants that are also going to show up in a more relevant way. And it's all about budget completion. I think that's one of been one of the earliest challenges to get through is budget completion and. Yeah.

[B]

What other KPIs are looking at?

[C]

Click through rates or click to conversion. Yeah, exactly. Yeah, exactly. And I believe that they'll probably continue to evolve that roadmap. Right. They introduce their CAPI ability. So you can kind of see, I think, a pretty clear progression of how that's headed to a true performance offering.

[B]

And customers use your feed into their cafe or they have to implement that separately.

[C]

They can use ours. Yes, they see ours.

[B]

That's a pretty big advantage because it's a pain in the ass to set that up. Right. So where do you see, I mean, if you put on your prognostication hat like where do you see advertising in LLMs going?

[C]

Well, that's a great question. I was. I almost might ask this back. Right. So what I find really interesting as it becomes not test budgets as it goes into BAU and it's part of the plan, where does that budget come from? Sure, Right. I think the conventional thinking is that a lot of that comes from search or PLA type of budgets. And that actually I'm probably pretty confident about. But here's a statistic that'll get you to think differently about where other budgets could potentially be cannibalized. Statistics that we found just last week. So 80% of the referred traffic coming from paid advertising on OpenAI are new to brand.

[B]

Okay.

[C]

Right, right. Which is a really powerful statistic. And when you think about it that way. Right. Why wouldn't you carve off budget from other brand type of budgets to fund that? If you get an 80% new to brand, I would take that from other sort of brand building platforms and put it into that. So I don't know if that's exactly how it'll turn out. But I mean those types of learnings start to make you wonder well, is it just cannibalizing sort of lower funnel search or does it start to cannibalize other sort of mid or upper funnel brand type of budgets? I think that's a super interesting question.

[B]

That is really interesting. I bet a lot of that has to do. I bet to be contrary here, the issue really is that search should be like that. But so much of search is paying for branded keywords. Right. So it muddles it because you're of people want a Burberry thing, they search for Burberry. People are not typing Burberry into their LLM when they want Burberry.

[C]

Yes.

[B]

Right. It's fascinating.

[C]

Yeah.

[B]

What's your general take on agentic commerce is it's been. I've written about it in the fact Mark Dixer has published at least three articles saying it's the next big thing and at least three articles saying it's never going to happen.

[C]

Yeah. So I would kind of probably somewhere in the middle. I've been very staunchly against a lot of the autonomous buying scenarios. So yeah, very much not in that camp. But I do believe in these kind of other words around like ambient support or sort of augmented. I mean imagine like something you could set to always be searching for the best deal on or hard to find items and then it's doing work while you're away from your interface or your agent. I think those are quite useful. And so I do believe in sort of like a lot of heavy assisted use cases whether those are active or in some ways like time delayed and a big believer in like inter app operability. I think that's like a really important use case that those platforms need to figure out. Not open claw but something much more like manageable and tame.

[B]

Consumers are never going to spin up their own situation.

[C]

No. No. But most people use Like Google, like AI answers and think that they're using AI. Right. Like most of the people that are not at this event think that they're using an agent when they're getting like AI overviews.

[B]

So bringing this together, a couple of points you made, you put out a very helpful press release this morning about Albertson. So I'm not just trying to promote what you're promoting, it actually is relevant. Bring it up. Right, so I think what I understood is that you're taking your retail media relationship with Albertsons and moving it into their chat bot and potentially also into other chatbots.

[C]

Yeah, exactly. So I don't, I don't think that like we talk a lot about OpenAI and, and the big platforms but like agentic advertising that have to be the sole domain of those platforms. Right. So if you're a retailer in particular, you're looking at this saying, well, how do I bring AI assistance into my shopping experience? Yeah. And I think the real benefit as well, and this is our Albertsons case study. I'm used to serving shoppers in a way that blends organic and paid listings and doing that in a way that's actually additive and useful to their shopping experience. And so why wouldn't I do that in this new surface? So if I've got a shopping assistant, I'm going to do the same thing. Now I might not show them a product listing page that has 12 slots in it, I might show sort of three products. But if I feel like through the relevancy calculation I've got an opportunity to serve one of those as sponsored, I'm going to do it because I'm getting a great shopping experience, I'm not going to lose a sale and I'm monetizing one of those slots. And that's a sort of shopping experience that everyone has become accustomed to. So it's the same relevancy engine that drives that decision making. It's just a tighter sort of set of slots. And we're really happy to announce that with Albertsons and I think other retailers will follow that and similar use cases as they build out that kind of shopping assistant on end.

[B]

Yeah, I mean we've seen Amazon and Walmart roll out a chat bots and the claim great success. I don't know offhand if they're putting in product listing ads in those, but I expect everyone would do this.

[C]

Yeah. And it'll be interesting over time to see the results in like sales, performance and gmv, because that's the other thing too is like how much of that is correlated to sort of high propensity shoppers, first adopters. So I read those same statistics. I get excited about them as a practitioner. But I do in the back of my mind say, okay, like, does the like growth in the business actually change? Right.

[B]

Yeah. You're moving the bottom of the funnel

[C]

just a teeny see how that plays out. Right. And companies that are able to invest in those types of capabilities, do they start to separate themselves from their competitors? They might, I believe in the compounding of technology capabilities. So it's a very real possibility.

[B]

Now take to the next step. Do you see something like and not necessarily Albertsons. Do you see an Albertsons like chatbot now showing up on OpenAI's chat when you're looking for Albertsons? Because I think they've done that with Walmart and they've said they're going to do that with other folks. And would there be an ad in that?

[C]

That's a great question. You'll have to ask Brian Monahan. I mean OpenAI has had, I think kind of a slow start with their app store.

[B]

Okay.

[C]

And so I think there are a lot of apps that are sort of standing up and going into that environment, but I don't know that the usage statistics are like off the charts yet. And then of course they walked away from the sort of owning the checkout.

[A]

Yeah.

[C]

They just seemed like just a smart move reacting to market feedback at the time.

[B]

Right, right. Well, that's definitely an area to watch is who controls that and who makes the ads work. So let's switch to the buy side. What do brands want from this environment? I think we've heard many times that brands really want to understand and increase visibility in AI results. Is that it or do they want more than that?

[C]

I think they want more than that. I mean, I think the biggest thing for them is going to be understanding how it fits in ChannelNext. Right. So I think, you know, we've all been talking about measurement and incrementality for years up and down this strip. I think that issue continues to take on even like more and more importance because now you have yet another sort of layer in the funnel, yet another major platform that's sort of begging for budget. How do you make those decisions? But what I believe in, most marketers believe that they've been over invested in Last Click Roas for many, many years.

[B]

Okay.

[C]

Right. Harvesting demand that they were already going to get anyway and not cultivating new demand or new customers. And I think that discovery ads give them kind of A new sort of way to express that desire to pivot out of that paradigm that they've in many cases feel they've been trapped in. But they've got to have measurement frameworks that prove that out. So I think it will be. It just takes on a new challenge, it's not a new topic. But now you've got to have just more sophistication to that because what's the bright look back window on an OpenAI ad relative to other channel touch points? Right, that's got to get figured out.

[B]

Yeah, the funnel question, I mean a lot of people don't like funnels, but the funnel question you brought up earlier is really fascinating about how the LLM answer could be at any point in the funnel, whereas it's much less so in search.

[C]

No, I mean the funnel thing is still a useful concept. Right. But I think you'll have to always use the descriptors to caveat it. People move through funnel potentially much, much faster in many cases. It varies by product category, but I think you get sort of a broader aperture at the top in terms of the types of questions and places that you can show up as a brand and that's exciting for marketers. But then people can like move to, from consideration to transaction, like really fast. Right. And so you know that it just is what it is.

[B]

So, so stepping away from OpenAI and LLM specifically and just about retail media are is the same sort of conversation with brands going on around control versus transparency. Do they want to just an easy button where the algorithm, the AI does everything for them or do they want to control everything at a bid and SKU level? A little bit of both.

[C]

But it also depends on in the retail space whether you are thinking about it from a trade perspective or a true sort of media planning perspective. And either way what they're looking for is just more ease of use to the channel. Right. So cross retailer campaigns or buying just needs to be easier, so less friction in that and then better measurement and incrementality because every retailer's got a slightly different take on that. As critia, we try to help standardize the approach to it. But you know, I've got a great looking roas, but I'm not shipping any more cases than I was a week ago. How do I get those two data points to line up? So I think they want to have a better understanding of cause and effect and again, depending on whether they're a media buyer or a trade buyer, they have sort of different things that they're paying Attention to given that one set of budget is rather fixed and the other one is a little more performance determined.

[B]

Yeah. And Criteo, because you have a sell side product for retailers and a buy side bidder for brands that enables sort of roas or you know, testing in a way that if you were just on the buy side it'd probably be difficult is. Do customers actively take you up on that?

[C]

They do. I mean I think that, I think that being on both sides of sort of supply and demand in today's age is a real benefit for Critio because I think more and more people are looking to buy outcomes.

[B]

Right.

[C]

And while they care about brand safety, they care about transparency, they care about measurement. If you can tick the box on all those appropriately, what they really care about is outcomes. And the best way to deliver outcomes is to have a connection point to both sides of the trade. And the other thing that I think has positioned us well in this space because like the way that you interact with a platform is also changing quite a bit as a company. Like we've never been really overly invested in user interface and workflow and so

[B]

I'm lift holding comment on that.

[C]

Yeah. Okay. It's a real benefit actually because as agentic and sort of like conversation based workflow becomes the norm, you really just want to get to the back end.

[B]

Right.

[C]

What supply do you have? What algorithms do you have to drive performance against that? And spare me sort of all the layers in between that I don't want to pay for. And so that's kind of our vision of how people are going to come into our capabilities. So that sort of agentic and prompt based interface has actually been a pleasant surprise from a roadmap perspective.

[B]

Makes Todd's job easier. He could just cross out all that stuff. Don't need UI anymore.

[C]

Yeah. I'll remind Todd that his job got easier. I don't think he'll agree with that.

[B]

I don't think he would. So we talk about commerce media, retail media as if it's one thing, but it's actually like 10 things. So a lot of things we don't talk about are like, you know, branding ads, non endemic ads in store ads. Is. Are those just separate topics or this all fold into the way you think about commerce media?

[C]

Well, yeah, I mean there's two sides of this. If you are, if you're a retailer, you think about that all. Yeah, for sure. Right. That was, that was my lunch meeting earlier today. They absolutely think about it together on our roadmap and we reprioritize the things that fit together logically. Would I like to be in all those spaces? Sure. But right now we've been focused on scaling sponsored products, getting into the display business, which has scaled nicely. We've started to reinvest in our front end to make cross retail campaign easier to run and manage. And then also starting to get into the page intelligence space which helps retailers basically optimize organic and paid results given sort of the signal that they're picking up on that page visit, which takes us into a new space because that's sort of a typically owned by the merchandising team or the E Comm team. Right. But it's a really important adjacency for retail media because why would you just have three fixed slots all the time when in some queries you should go as deep as 9 or 10 and maybe in some cases you should only have one. But that's a determination that should be made holistically for a retailer. So that's a really exciting space for us to be getting into.

[B]

Yeah. So to kind of close it out, you know, retailers and commerce companies, they're under a lot of pressure. It's tough out there. And now they're saying, well, maybe people aren't going to shop anymore. People are just going to say and you know, so is there a rosy future ahead for, for your customer base, for the retailer community?

[C]

There is. I look, I have to be glass half full on this. I think retailers play a very important role in the ecosystem. In fact, we saw E commerce traffic rise in the last quarter or two. Right. Not, not declining. And as you see like AI platforms pull back from checkout features, I think it reinforces that unless you're willing to be in the fulfillment business, the customer service business, the processing returns business and just the joy of shopping business, then it's a tough thing to try to own a chunk of that, of that, of that commerce layer. Right. And so I think retailers have a real right to win in this new world, but they can't sit still. They need to have best in class front ends that are the same type of thing that people are used to using on either Amazon or their favorite AI assistant. If they do that, then all those other sort of incumbent capabilities keep them locked in and I believe they've got a really strong future.

[B]

Sounds great. Well everyone give a big round of applause to myself, Komasinski, the CEO of Prettio. Thank you. Thank you.

[C]

Thanks everybod.

[B]

Sam.