Why Retail Media Is Becoming the Intelligence Layer of Commerce with Yuri Bukhan

Yuri Bakar: AI is definitely making marketers much more proactive. But what it's really ultimately doing is we're moving from marketing automation to marketing cognition. Right. And automation is executing the rules, intelligence is recognizing the patterns and recommending the actions. So most marketing today is reactive.

Host: Welcome everyone. Today we're going to dive into one of the most interesting topics within retail. A lot of shifts happening within how retailers are approaching marketing media and largely organizing their digital programs. I'm joined by Yuri Bakar, who's the VP of Global Business Development at Growth Loop. He brings a lot of deep expertise at uh, the intersection of data, AI and media and marketing activation. Really excited for today's conversation. Yuri, welcome.

Yuri Bakar: Thank you. Surya. Great to be here. This is one of those moments where I feel multiple threads are converging retail media AI first party data. So it's a great time to really unpack what is really happening.

Host: Absolutely. I think let's start with the elephant in the room. I get more conversations, more confidence invites about retail media than anything else. It's exploded really. From your perspective, what's driving this growth? What's the reason behind this explosion?

Yuri Bakar: It really has at a high level. Retail media is growing because it solves three critical problems. Precision, proximity to purchase, and proof of outcome. For years, digital advertising gave a scale but not always real commerce signal. Retail media, uh, flips that. It's built on first party shopper data. It reaches consumers right at the moment of decision and it ties media exposure directly to product level sales. And the role of the digital shelf has changed. It's not E commerce anymore. It's where discovery happens, comparison happens, trust is built and conversion happens. Even if somebody buys in store, that decision is often shaped digitally first. And maybe the biggest shift, retailers are becoming media owners. They have proprietary intent data and that changes the entire balance of power 100%.

Host: Agree to that, uh, point, Yuri. I think if you look at some of the data that goes on behind the scenes with the retailers. Right. You've got greater insight about household shopping power, their underlying preferences, that they have, their means to shop as well as which all is not just relevant for CPG suppliers, but also non endemic advertisers. You spoke about how these signals come together and how this is a new way of looking at a fragmented industry. Where does this challenge come from with the fragmentation? How do we solve for the fragmentation challenges that exist today in building that unique commerce signal across these different systems and giving you that precision?

Yuri Bakar: Spot on. I mean, this is where things get a little bit complicated. Right. Because the Customer journey is fragmented and it's rarely managed as a whole. Brands can see behaviors across retailer sites, social search, apps, stores, but those signals are largely disconnected and they're owned by different teams within the organization. So you end up with media teams optimizing for roas, right? Return on ad spend, CRM teams focused on retention, e commerce teams tracking share. And they're all doing great work, but it's not connected work. And so the issue isn't the lack of data, it's the lack of operational coherence and fragmentation's actually increasing. So from my perspective, the real question becomes not how you buy more media, but how do you unify these signals and act on them intelligently?

Host: And why do you think there is a struggle to do that? Yuri, is this just operational structure or. There's some technical challenges that companies need to solve. What does it really take to bring all of these signals together and create like one unified action plan?

Yuri Bakar: I think what most companies have already done the hard part, they've centralized the data, they've built models, they've invested in analytics. But the challenge is that insight alone does not drive growth. From my perspective, the gap is in the last mile between analysis and action. Modern data activation solves that. It makes the data usable in production, not just visible in dashboards. So now transactions become audiences, product signals become triggers, and customer history becomes personalization. That's really the difference between a data asset and a decision engine. And from my perspective, this is really where things are evolving beyond traditional CDPs. It's no longer about unifying data, it's about activating data directly from the data warehouse across every channel. When you do that, well, data stops being an IT project and becomes a growth lever.

Host: And that also helps in bringing decisions in a centralized place as opposed to making decisions in a siloed way across across different channels. As you hinge more on the warehouse based data and decisioning strategy, this also opens the door for AI, primarily generative AI and agentic AI. We've seen the cost of generative AI drop significantly. It takes about a few thousand dollars to process millions of images right today. How do you see AI getting into this architecture where it is all warehouse centric, where the decisions are centralized, data are centralized?

Yuri Bakar: That's a great question. And I think AI is causing a monumental shift. AI moves marketers from doing the work to supervising the systems that optimize the work. So traditionally marketing is pretty linear. You define segments, you build journeys, you launch campaigns, you analyze the results. What AI really enables marketers to do is it makes that loop continuous. And this is pretty critical. AI can help identify new audiences, recommend journeys, detect fatigue, adjust budgets, often much faster than teams can react manually. But here's the key point. AI is a great co pilot. AI does not replace strategy. It compresses the cycle between hypothesis and learning. And in retail media, where everything is changing and evolving so quickly that speed of iteration is a massive advantage.

Host: And I think retail media on the whole is not just marketing to the end customer, but also making sure that your retail strategy, which the merchants follow in terms of pricing, promotion, high low price, everyday low prices, also talking to the same strengthenergic lens in retail media. Now when we look at, uh, the point that you made about AI will not define the strategy. You need to define the strategy for AI. And AI becomes the more the operational system that marketers can use. How is this making a marketeer's life easy? Is the role of the marketeer drastically changing from what it used to be? Is AI making them more proactive? Can you talk a little bit about. You talk to a lot of marketers day in and day out. Where do you see this going from a marketeer's roles and responsibilities of the future?

Yuri Bakar: AI is definitely making marketers much more proactive, but what it's really ultimately doing is we're moving from marketing automation to marketing cognition. And automation is executing the rules, intelligence is recognizing the patterns and recommending the actions. So most marketing today is reactive. We ask what happened last week, what worked well last quarter? AI lets us ask what's going to happen next in the future? What should we do now? In retail media, it could mean predicting churn, identifying demand shifts, spotting inefficiencies. So marketing really becomes less about channels and more about commercial intelligence.

Host: And um, this also needs to tie back to One of the interesting shifts that we see in this space is how retail media is not just a lever for generating new ad revenue, which is critical, but it also should talk to the end customer experience. It should also help with the overall enterprise marketing strategy. So there was a lot of need that we've seen in the retailer space, specifically on how retail media, along with enterprise marketing investments, is driving the overall benefit back to the retail bottom line. Now, um, where do you see the measurement space go from here? Are we looking at a space where retail media continues to be measured, siloed, or uh, is there an opportunity for AI to start improving the measurement itself, bringing together retail media spends, enterprise marketing spends, and so on? So how do you see measurement going?

Yuri Bakar: I think marketing has Always been very good at measuring activity impressions, clicks, conversions, but weaker at uh, measuring causality. The real question, what actually changed behavior? AI enabled measurement M connects exposure, customer history, product sales, long term outcomes. So we're moving towards incrementality, lifetime value and halo effects. Because what we see when we talk to customers right is if you're optimizing only for roas, that can actually lead to the wrong decisions. And the best companies really treat measurement as a learning system, not a reporting system.

Host: Now this reminds me that we've not really spoken about Growth Loop so far, but the philosophy that you just spoke about really aligns to what you promote as compound marketing engine. I really like that term that you use because it means that there is a system that's always learning and iterating on your marketing activation strategy. Can you talk a little bit about what is compound marketing engine? How did you arrive at that philosophy to start with?

Yuri Bakar: Sure thing. So the compound marketing engine, it really is a simple idea. Every campaign is an execution event and a learning event. Each cycle improves audiences, targeting, sequencing, measurement. Traditional marketing is episodic. We talked about kind of the linear approach the marketing has had. Compound marketing is really all about accumulating intelligence. Think of it as it's like a data platform. Each use case makes the next one better. So really we think that there is a shift from campaigns as projects to marketing as a continuously improving system. And the companies that are going to go and iterate and continually improve day by day are going to have the biggest impact on their bottom line and be able to drive the biggest beneficial outcomes for their business.

Host: Completely subscribe to the thought process of compound engineering and I think we call it test and learn platform. There are different variations of this out there, but ultimately the philosophy is can we make marketing and continuously improving system rather than a one off best decision possible. Now as we go into the space, there is a need for all of these systems to talk together and you have a great product and growth loop. But in order to accomplish this end to end, I'm pretty sure that there's a lot of partnerships that are involved across the board to execute this mission. And as uh, a VP of business and development at Growth Loop, how do you see partnerships evolve? How critical are partnerships for a retail media business? And where do you see the true benefit of all of these different technologies coming together?

Yuri Bakar: It is not a one person game. Partnerships are absolutely critical because it is an ecosystem problem. Retail media involves retailers, data platform, activation channels, AI systems. No single company can solve it. And bringing the expertise of someone like Treatence, the domain knowledge, the systems knowledge, it really helps to ultimately be able to deliver value to customers by bringing those systems together ultimately into a solution that solves the customer problem. But really strong partnerships, they reduce the time to value, they improve interoperability and enable real workflows. Enterprises have technology investments, they have complex stacks. It's absolutely critical to be able to plug into their existing systems to be able to drive value quickly. And for enterprise clients, integration really matters more than features. So partnerships aren't just go to market, they're part of the product, they're part of the solution that customers end up adopting.

Host: And we do see this more often than not, especially in the top retailers that have adopted this philosophy. Well, have significantly outperformed in terms of the ad revenue that they bring to the table, but then also the experience that they deliver to the customers where uh, multiple such products come together in a composable architecture that's more API friendly, that can do these real time interventions and corrections as needed and give the best experience possible to the customer across on site and off site channels. Now we spoke about retail media quite a bit. We spoke about how it needs to evolve into this continuous learning system and the partnership's criticality. What are some top trends that you're seeing in this space? Where do you see retail media go from here into next three years? Where is the investment heading?

Yuri Bakar: I think there's three shifts. So first, expansion beyond on site media to commerce media across ctv, social and in store. The second is around AI driven discovery, so optimizing for algorithms, not just humans because we're going to see, if we're not already seeing agentic based commerce. And then third is industry consolidation, better measurement, better governance, interoperability. Retail media is really going to become less of an ad channel and more of a core intelligence layer for commerce.

Host: That's uh, a fascinating closing thought on how retail media is becoming the core intelligence layer for commerce. As we look at the future of retail media and how it enables that backbone for commerce, what are some thoughts that you want to give to the leaders in this space? Someone who's looking to break away from a third party model into a more in house retail media stack. We've seen newer retailers take uh, interest in retail media programs and invest heavily into programs. We've got all different types of users in the space right now. What are some of your key advices on someone who's starting from scratch in this space? What should they do? What should they look out for in uh, the future?

Yuri Bakar: I think one of the things is it's critical to start with an operating model, not channels. The second thing is treat first party data as a strategic asset. It's unique. You have it, nobody else has it. It's your differentiator. Align around business outcomes and build for learning velocity. The winners are not going to have perfect plans, they're just going to learn faster than everyone else. So this constant iteration is critical for figuring out what is working, what is driving the biggest impact and then investing in those particular motions now.

Host: Couldn't agree more. Fantastic insights, Yuri. Thanks for sharing all of these great details with us. We covered wide range of topics today, but I think we were very precise on the recommendations that came out of this conversation. So appreciate your time and thanks for being here.

Yuri Bakar: Surya, thank you so much for your time. Really enjoyed it. We're going to be at Google Next in Las Vegas later this month. Treatnce is obviously going to be there as well. Come check us out. We're at booth 2611 and also join our joint happy hour at Stripstake at 5pm on Thursday, April 23rd. Thanks again Sariya.

Host: Thank you.