This article was based on the interview with Romain Fouache, CEO at Akeneo by Greg Kihlström, AI and MarTech keynote speaker for The Agile Brand with Greg Kihlström podcast. Listen to the original episode here:
The pressure on enterprise marketing leaders to deploy AI is palpable. It’s a mandate that echoes from the boardroom, reverberates through shareholder meetings, and lands squarely on your desk. The directive is clear: innovate, integrate, and do it yesterday. The market is awash with dazzling generative AI tools promising to revolutionize the customer journey, from hyper-personalized recommendations to fully conversational shopping assistants. The temptation is to grab the shiniest new object and plug it into your stack, hoping to declare an early victory in the AI arms race. But this rush to the front lines of innovation often bypasses the most critical territory: the supply lines.
In this landscape of hurried implementation, the greatest risk isn’t falling behind the technology, but rather getting dangerously ahead of your customer’s trust. The foundational work—the unsexy, meticulous, and absolutely essential task of getting your data house in order—is what separates a successful AI strategy from a catastrophic one. As leaders, we understand that technology is a means to an end, not the end itself. The goal is to build stronger, more meaningful connections with customers. Roman Fouache, CEO of Akeneo, brings over two decades of experience in the data and AI space to this exact conversation. His perspective is a grounding one: before you can dream of an AI-powered future, you must first master the data-driven present.
The First Rule of AI Club: Get Your Data in Order
Before a single line of code is written for a new AI shopping feature, before the project kickoff meeting is even scheduled, there is a fundamental question every leader must ask: Is our data ready? The answer, according to Floch, is the absolute determinant of success or failure. The allure of AI can create a kind of strategic amnesia, where we forget the simple, immutable law of computing: garbage in, garbage out. An AI model, no matter how sophisticated, is only as good as the data it’s trained on.
“Don’t even try if your data is not in the right place. You know, we we share a lot about AI trust, about hallucination, about all of these things, but to begin with, if you don’t know what your products are, AI is not going to make it up for you… if you get it wrong once, you will kill the trust of your consumers.”
This is not a call for delay, but a call for proper sequencing. Floch’s point is a critical one for any leader feeling the pressure to “just ship something.” A single bad AI-driven recommendation—a piece of clothing suggested in the wrong size, a spare part that doesn’t fit, a product description that misrepresents a feature—can instantly shatter the trust you’ve spent years building. The core function of product information is to create certainty for the buyer. When AI, fed by inconsistent or inaccurate data, introduces uncertainty, it’s not just failing at its task; it’s actively working against your brand. The deliberate, often unglamorous work of ensuring your product data is accurate, consistent, comprehensive, and structured is not a barrier to speed; it is the accelerator for meaningful and trustworthy AI adoption.
Beyond Automation: Avoiding the Plague of “AI Slop”
Once the data foundation is solid, the next temptation is to unleash AI on content creation at a massive scale. The logic seems sound: if generating a product description or a marketing email used to take hours, and now it takes seconds, why not generate a thousand variations? Floch warns that this path leads to what he aptly calls “AI slop”—a deluge of low-value, generic content that dilutes your brand and ultimately harms performance. The promise of AI is not simply automation; it is augmentation.
“I think that is one of the biggest risk potentially of the year or the next couple of years. It’s to think that AI is going to be autonomously able to generate large amounts of contents that have added value. Because ultimately, yes, from the title of a product, I can generate a 50 page description with open AI. But it’s just going to be AI slop. And the products you’re selling, you don’t want them to be slop.”
The most effective marketing organizations are reframing the role of their teams in this new paradigm. Your marketers, copywriters, and merchandisers are not being replaced by AI; their roles are being elevated. As Floch describes it, they are becoming “managers of agents.” Their job is to provide the strategic direction, the brand voice, and the core truths about a product. The AI agents then execute against that direction at scale—translating, adapting, and formatting content for different channels. The human remains firmly in the loop as the strategist, the editor, and the final arbiter of quality. This collaborative model ensures that AI serves the brand, not the other way around. It allows you to maintain the unique voice and perspective that differentiates you, while leveraging technology for efficiency and reach. The goal is not to produce more content, but to produce better, more relevant content, more efficiently.
The New Frontier: Generative AI Engine Optimization
For the past two decades, Search Engine Optimization (SEO) has been a cornerstone of digital marketing. We’ve all become experts in keywords, backlinks, and domain authority. But the very nature of discovery is undergoing a seismic shift. Customers are beginning to bypass the search bar in favor of conversational interfaces like ChatGPT, Gemini, and Perplexity. This shift demands a new discipline, what we might call Generative AI Engine Optimization (GEO).
“I think when it comes to AI in the e-commerce space, currently, the highest impact feedback loop is understanding how you’re ranking and being presented on chat GPT. You know, it’s it’s kind of SEO but for the… generative AI engine optimization… I think the way we’re going to be driving discovery and ranking uh whether paid or organic in a year or in two years is going to be radically different from what we do today. And to be honest, at the core of it, it’s going to be one simple thing, your product data.”
In this new world, discovery is not based on a user matching keywords to your webpage. It’s based on an AI agent understanding a user’s intent (“I need a waterproof jacket for hiking in the Pacific Northwest in October”) and matching it to the structured data it has about your products. The comprehensiveness and accuracy of your product information—specifications, materials, use cases, compatibility—become the primary currency of visibility. This makes the foundational data work discussed earlier not just good practice, but a competitive imperative. The feedback loop for leaders is no longer just about your Google ranking; it’s about actively monitoring how and why your products are being recommended (or not) within these generative AI environments. The work you do today to structure and enrich your product data is what will fuel your brand’s discoverability in the commerce channels of tomorrow.
Preparing for Disintermediation: When Your Website is No Longer the Star
Looking further ahead, the evolution of AI points toward a future of “agentic e-commerce,” where AI agents, acting on our behalf, handle transactions with minimal human intervention. Imagine telling your home assistant you need to restock on your favorite coffee, and it handles the entire process—finding the best price, ordering, and scheduling delivery—without you ever visiting a website. In such a world, many of the traditional levers of e-commerce marketing become less relevant.
“There is a world in which there is no higher impact on your performance than your product data because ads might not be there anymore and your website might actually be a secondary or tertiary uh channel because many of it is going to go through perplexity and… open AI and Gemini.”
This is not a prediction that websites or search ads will disappear entirely, but an acknowledgment that their role in the customer journey may be significantly diminished. When an AI agent is the “shopper,” it won’t be swayed by beautiful web design or clever ad copy. It will make its decision based on cold, hard data: price, availability, shipping times, specifications, and perhaps historical conversion rates for similar intents. In this disintermediated future, the single most powerful asset you have is the quality, freshness, and comprehensiveness of the product data you can feed into these ecosystems. The brand that provides the most reliable and complete data is the brand whose products the agents will trust and select. This is the strategic long game leaders must begin playing now.
The path to AI excellence in commerce is not a sprint toward the newest, most impressive technology. It is a deliberate, strategic march that begins with a solid, unshakeable foundation of data. The principles are straightforward: establish a single source of truth for your product information before you even attempt to deploy customer-facing AI. Elevate your teams to become strategic managers of AI agents, using technology to augment their expertise rather than replace it. And finally, recognize that the very channels of discovery are being rewritten, with your structured product data as the new language of visibility.
As leaders, our role is to navigate this transition with a clear-eyed focus on the ultimate goal. The true competitive advantage in the AI era will not be measured by the number of AI features on your website, but by the depth of trust you have with your customers. That trust is built on a bedrock of accuracy, consistency, and relevance—all direct outcomes of a world-class product data strategy. Resisting the pressure for superficial, short-term wins in favor of building a resilient, data-first marketing engine is the defining challenge—and opportunity—of our time.
