MessageGears, VP of Product, Eugene Yukin, Exclusive Interview with MarTech Pulse on AI Personalization & Data

In an exclusive interview with MarTech Pulse, Eugene Yukin shares how personalization and martech evolution are reshaping customer engagement, enabling scalable growth through data, AI, and warehouse-native architectures.

Can you share your 15-year career evolution from building DTC ecommerce at Lamoda, to leading personalization at Carter’s and Home Depot, and key lessons in scaling martech across B2C channels?

I got my beginning in startups, spending about five years building products in the voice-to-text and education marketplace technology space. After a small acquisition, I found myself genuinely curious about how big companies build product at scale, which brought me to The Home Depot.

There, one of my teams was responsible for every triggered and transactional order communication – online, in store, and across multiple channels. The data complexity was immense: hundreds of dynamic templates, all dependent on how and where you ordered. One of my most formative experiences was running eye-tracking studies on our transactional emails. We wanted to know: when a customer opens an email, what do they actually see in the first ten seconds? What we found was humbling –  people weren’t looking where we expected. They would miss critical details like when their order would arrive and would call the call center instead. It taught me that even in something as simple as a transactional email, the customer experience and attention to detail is everything.

At Lamoda, the largest fashion e-commerce platform in the CIS region, my focus shifted to discovery and inspiration. The question was: what should the experience be for someone who doesn’t know what they want to buy? My team built an entire layer of discovery products, editorial content, curated looks, and stories. We were able to significantly increase traffic to discovery content, and the Looks product is still one of the most popular features on the platform today. My team also tackled creating a premium digital experience within a mass-market app. We used larger photography and different visual rhythm, which forced product, design, and editorial to collaborate extremely closely. The lesson that stuck with me is that marketing comes alive when it’s truly personalized to the consumer and the segment– and scaling content to make that possible is one of the hardest problems to solve.

At Carter’s, I built the team and technology stack driving personalization. We retooled the entire martech stack: a new CDP, CMS, DAM, and SMS provider. Through all of that, a few lessons crystallized: no matter what you build on top of poor data, you won’t get value out of it. Data transfers are extremely costly and compound over time. Previously poorly implemented tools cause massive downstream problems. I saw multimillion-dollar platforms go underutilized because the implementation didn’t match how teams actually worked. And perhaps the most overlooked lesson: most enterprise marketing teams don’t really have a clear picture of everything happening across their ecosystem. Everything is always go, go, go. So teams lose the ability to zoom out and understand what’s actually driving impact.

Read More – How a Composable CDP Lets You Treat 1,000 Customers Like Your Only One

What drew you to MessageGears as VP of Product, and how does leading the roadmap for warehouse-native data activation leverage your experience bridging B2C personalization with enterprise B2B martech needs?

MessageGears felt like the place where everything I’ve learned could come together. I’ve been the customer, and I spent years internalizing the martech buyer’s pain points – poor data, latency, vendor lock-in. I’ve built a marketing technology team, made the buying decisions, and lived with the consequences. So joining a company where I could apply all of that from the product side felt like a natural step.

What really drew me in is the scale of the opportunity. MessageGears is used by some of the largest brands in the country. I kept thinking: if we can get the marketer’s experience right at the enterprise level, we can impact the experience millions of consumers have with marketing every day. The possibilities genuinely feel endless.

My experience is serving me well here because the product discipline is the same whether you’re building a consumer shopping feature or an enterprise engagement platform – you must understand the customer deeply, find the friction, and remove it. At MessageGears, those users might be a data admin, a campaign manager, or someone building complex segmentation. And while every enterprise team looks a little different, the questions are largely the same. What excites me now is that the end user itself is evolving. Is it a marketer? Or is it an AI agent that a human is instructing? The questions are the same; who (or what) is using the tool is different.

Following MessageGears’ platform upgrades for multi-destination audience extraction and your quotes on cost efficiencies, what recent innovations are you most excited about for real-time cross channel engagement?

There are a few things I’m genuinely excited about right now.

We’re making a large investment in fully warehouse-native cross-channel journey capabilities. We currently have a powerful tool called blueprints that enterprises use for sophisticated audience segmentation and campaign orchestration, but where we’re headed next is enabling warehouse-native journeys at scale with stronger real-time capabilities. Most of our competition runs journeys entirely in the cloud, which limits personalization because data has to sync back and forth. That’s expensive, and it’s often too slow to keep up with consumer expectations – it also means your journeys can’t easily evolve as your data does. We’re investing heavily in changing that.

Our recent multi-brand release is another one I’m proud of. Enterprise marketing teams often have complex structures, including dozens or even hundreds of brands and product lines, with each having different marketing needs. We’ve completely rewritten MessageGears’ roles and permissions system so enterprises can now build marketing assets, segments, campaigns, and templates with full flexibility that aligns with how modern teams are actually structured and how they want to market each individual brand. For companies managing that complexity, this is a fundamental change in how marketing activation works.

On the AI side, I’m excited but intentional about how we approach it here at MessageGears. We’ve already built powerful predictive modeling capabilities our customers use today that drive intelligent behavioral insights and analytics. Now, we’re embedding AI directly within the product in areas that cause daily friction for enterprise marketing teams. Our current focus is on AI copilots that make marketers more productive and surface the right insights faster, while still keeping the marketer at the center.

For 2026, what trends do you foresee in warehouse-native martech, particularly AI-driven personalization, reverse ETL scalability, and how enterprises will consolidate stacks for faster customer data activation?

The trend I feel most confident about is data continuing to consolidate into the warehouse as the centralized source of truth. MessageGears was founded on this premise – that the data warehouse is your marketing database – well before warehouses started positioning themselves that way. What’s accelerating it now is AI. As models get more capable, the enterprises that keep their data centralized and accessible will be able to train and activate against it far more effectively than those shuttling copies between disconnected tools.

On the AI personalization front specifically, I think we’re at an inflection point. The conversation has shifted from “use AI to write subject lines” to “leverage AI agents that orchestrate entire campaign workflows, audience selection, channel mix, timing, and creative variations,” with the marketer providing strategic direction rather than manual execution. That’s a meaningful shift, and it requires your AI to sit directly on top of your data, not a stale copy of it.

For reverse ETL and data movement more broadly, I think the honest answer is that the traditional model – extract, transform, and load data into a cloud tool, then activate – is showing its limits at enterprise scale. Every new data variable or model refinement amplifies transfer costs. The composable approach, where activation happens closer to where data lives and you plug in best-of-breed tools as needed, is gaining real traction. We’re seeing enterprises move away from monolithic suites toward architectures that give them optionality.

The trend I’m watching most closely is the conversation around on-prem. Training models and running LLMs requires moving significant data back and forth for context, and that’s expensive. The question is how fast and what will be the breadth of capability when models get integrated directly into the warehouse. Survey data shows that many companies have already started moving sensitive infrastructure and AI workloads back on-prem. Ultimately, the enterprises that think long-term about data ownership and flexibility will be in a much better position.

Read More – The Business Impact of AI Personalization in E-Commerce

Drawing from your product leadership at Carter’s and MessageGears, what top advice would you offer martech teams struggling with siloed data to achieve real-time, compliant personalization at enterprise scale?

Start with the data, and be honest about the state it’s in. I have seen it take nearly a year of focused effort just to clean and modernize our data foundation before we could do anything meaningful with personalization. That’s not glamorous work, but it’s the work that makes everything else possible. AI has introduced capabilities that can speed this effort up, but you still need the work done.

Architecturally, move your engagement tools closer to where your data lives. If your data is in the warehouse but your tools are in the cloud, you’re paying to move data back and forth constantly, and those costs grow every time you add a new variable or improve a model. And here’s something that doesn’t get discussed enough: if your cloud vendor is storing a copy of your data and running AI on it, you’re failing to grow the intelligence of your own centralized data store. That lock-in has real long-term costs.

Be intentional about what actually requires real-time personalization. Not everything does. Some journeys should be hyper-personalized and triggered in real time, but scaling that across every touchpoint is extremely hard and expensive. Think carefully about where it drives real impact versus where it’s a nice-to-have.

And finally, focus on your content pipeline. How does your content get created? What taxonomies do you apply to image assets? Do those taxonomies persist through your CMS and other downstream tools? Can you tie user attributes to content attributes? These sound operational, but they determine whether personalization actually scales or breaks down in production. Getting content scaling infrastructure right is just as important as getting the data right.

Reflecting on your shift to MessageGears amid rising composable martech, what’s the most transformative career insight you’d share with product leaders navigating data warehouses and customer engagement evolution?

My most transformative insight is simple: understand how your teams actually work before you buy anything. I’ve seen organizations spend millions on platforms that went underutilized, not because the technology was wrong, but because the implementation didn’t match how people actually collaborated. Do you have deeply technical data teams alongside non-technical marketing teams? Are they working toward shared goals or in silos? The right tool often isn’t the one with the most features, it’s the one that fits how your organization actually operates.

The same rigor applies to how you think about your customers. Does your business truly require real-time messaging, or is batch fine for most of what you do? Is one-to-one offer decisioning genuinely effective, or does your team lack the analytics infrastructure to run it well? What channels do your customers engage on? Do you need every single marketing channel to be available as part of the tool even though your business does not require that amount of channel depth? Being honest about those answers saves you from buying complexity you can’t operationalize.

And increasingly, we need to ask: what does marketing look like a year from now? Are we building for a world where AI agents make recommendations and purchasing decisions on behalf of consumers? If that’s where things are heading, and I think it is, the tools and architecture we choose today need to account for that. The basics of understanding your customer never go away; what’s changing is what that customer looks like.

If advising your early-career self on martech’s shift to warehouse-native platforms like MessageGears, what one question would you pose about future-proofing personalization, and your answer today?

If I was advising my early-career self, the question I’d ask is: am I building personalization around a vendor, or around durable architecture?

Earlier in my career, I focused heavily on platform capabilities and feature velocity. Over time, especially as I saw the shift toward warehouse-native and composable models, I realized the more important question is about data ownership, flexibility, and interoperability. Who owns the data? Can it move freely between the tools you use today and the tools you might need tomorrow?

Future-proofing personalization isn’t about choosing the right tool for today. It’s about designing systems that give your organization real optionality as technology evolves – because it will evolve, likely in ways none of us can fully predict. If you’ve built around architecture that keeps your data centralized, accessible, and interoperable, you can adopt the best tools as they emerge without starting over. That’s the lens I apply to every product decision I make now.

Write to us [⁠wasim.a@demandmediaagency.com] to learn more about our exclusive editorial packages and programmes.