The Modern Data Company

Three weeks ago, I wrote about the Bullwhip Effect, how misaligned teams turn a 12% forecast adjustment into a 28% overorder and a margin disaster. The solution? Alignment. Providing visibility of the demand signal to merchandising, finance, and ops. But here's the harder question: What happens when the signal itself changes FAST? Alignment and visibility solve internal amplification. They stop your teams from distorting demand as it moves through the organization. But eCommerce in 2024-2025 didn't just suffer from internal misalignment. It suffered from external volatility: - A supplier delays shipment by 3 weeks - A tariff announcement hits overnight - A micro-trend spikes 300% in search volume, then collapses in 8 weeks - Weather shifts and your seasonal inventory becomes deadweight When the environment moves this fast, visibility isn't enough. You need to understand relationships. How a change in one place cascades everywhere else. This is where most eCommerce operations break down. Your systems can tell you what you have. They can't tell you what it means. They know you ordered 10,000 units from Supplier A. They don't know that Supplier A shares a dye house with Supplier B, that both feed 14 SKUs generating $3.2M monthly, and that a 2-week delay forces markdowns on your highest-margin category. That's not a visibility gap. That's a context gap. The fashion industry shows what happens without it. Nike's markdowns jumped from 19% of assortment in 2022 to 44% in 2024 costing 350-400 basis points of margin in a single quarter. LVMH and Kering reported €5 billion in combined excess inventory in 2023. Discount rates increased 5 percentage points in H1 2024 alone. These companies have visibility. They have dashboards. What they lacked was the ability to trace how a supply disruption cascades into margin erosion fast enough to act. The next level: Semantic Intelligence Visibility shows you the demand signal. Semantic intelligence shows you what happens when that signal, or your supply changes. Traditional system: SKU-4521, Supplier: Acme, Cost: $12.50, Lead time: 6 weeks Semantic model: SKU-4521 shares materials with 14 SKUs generating $3.2M monthly, has one qualified alternative requiring 4-week requalification, 15% margin erosion if marked down before November. That context is the difference between reacting in weeks and responding in hours. Building on alignment: 1. Map hidden dependencies: Which SKUs share suppliers or materials? 2. Where does one failure ripple across your catalog? 3. Quantify cascade effects: If Supplier A delays, what's the revenue at risk? What are the markdown implications? Enable "what if" in hours, not days: Can your team model a supplier delay or tariff change in a single meeting? Bottom line: Visibility quiets the internal noise. Semantic intelligence lets you hear the external signals and act before they become markdowns. The Modern Data Company Srujan Akula Animesh Kumar Saurabh Gupta Pam Kramer Dinker Charak

"AI-Ready: The Foundation Before the Leap" Every company today wants to be “AI-enabled. But here’s the paradox, AI doesn’t fix weak data foundations. It amplifies them. Before we talk about models, predictions, or automation, we need to talk about something more foundational: data readiness. In most retail and e-commerce organizations, merchandising, supply chain, and sales still operate on disconnected realities. Each function is operating on its own, but inconsistent in when see together, leading to problems that constantly repeat: inventory gaps, reactive planning, and costly markdown cycles. Readiness isn’t about adding more technology. It’s about creating one version of truth - shared definitions, connected data, and transparent lineage. When that foundation is strong, intelligence will become a second nature. Clean, connected data doesn’t just improve accuracy, but it shifts focus. Efforts move from reconciling numbers and proactive planning. "AI becomes contextually aware and outcome focused, not just computationally powerful." #Retail #eCommerce #RetailTransformation #EcommerceStrategy #SupplyChainIntelligence #ConnectedData Srujan Akula, Sagar Paul, Sanjoy Roy, Sundeep Gopal, Animesh Kumar, Bhava Kompala, Pam Kramer, Dinker Charak, Sachin Dharmapurikar

Eight years ago, thanks to my friend Sanjoy Roy (and Dinesh Sharma), we launched the world's first sense-and-serve AI platform for fashion. The goal was to provide better value to our end customers, and capture more of their attention, which resulted in more insights, better zero-party data, and an increased conversion rate. Now the old business model built on attention is finally starting to lose ground. The web's power structure is breaking. AI agents don't follow the rules we built for humans. They don't click ads. They don't browse pages. They don't wait for search results. A new pyramid is forming where NVIDIA controls compute, Microsoft/AWS control orchestration, specialists control domains, and data owners control context. Organizations should really transform their data infrastructure to be AI-ready now. Companies that organize and control their data will win twice: they'll operate more efficiently, and build (totally) new revenue streams. Reach out to Sanjoy to learn how The Modern Data Company's DataOS can help your organization prepare for this shift. Today's Gennaro Cuofano's Business Engineer newsletter, "How AI Agents Are Reshaping Digital Power," is a crystal-clear analysis of this huge shift.

6 weeks ago, we met a mid-sized fashion brand in our office. We were shocked to learn how they had throttled their aspirations of using AI just because they felt they were too small and unsophisticated with Data. We know how many countries skipped the land-line era to jump straight on to smartphones. Why can't we leapfrog, learning from each other and getting access to market ready data infrastructure? Introducing : AI-Ready Reference Network You're not just implementing technology. You're joining the AI-ready data transformation movement. Our CDO-as-a-Service reference partners get exclusive access to: → AI-ready data best practices from other premium brands → Early access to new AI-optimized data capabilities → Peer benchmarking on AI deployment success metrics More importantly, you get at-cost pricing because we're building an AI-ready data community, not maximizing profit margins. European brands that get AI-ready data right will dominate 1. Personalization 2. Inventory optimization 3. Customer experience. Are you leading the conversation or watching from the sidelines? The Modern Data Company Matteo Petruzzellis Sanjoy Roy Srujan Akula Dinker Charak Pam Kramer Saurabh Gupta Animesh Kumar #AIReadyData

The AI-Ready Data Dilemma This is how it looks when one is hungry, desperate for some finger-licking food, have paid the money but the restaurant says "Sir you have to wait for 30-45 mins atleast"... with my dear friend Dinesh Sharma somewhere in Mumbai while making a few sales calls few years ago. Large e-commerce brands in Europe might have AI budgets but not AI-ready data. The hard truth: → Senior CDO + Supporting team of engineers and analysts + Infrastructure build-out = 800,000 € at least. Time to #AI-ready data: 12-18 months minimum. Slow and expensive. So a lot of brands remain non-AI-ready only focusing on doing PoCs but little outcome. That's why we at The Modern Data Company created CDO-as-a-Service. For a fraction of the above costs, you get AI-ready data infrastructure 80% faster and 50% cheaper: ✓ A dedicated team of senior data professionals ✓ DataOS platform optimized for AI consumption and your industry ✓ 85% faster AI project deployment ✓ Data products built for machine learning from day one. Want to learn more about this new offering? Ping me or Matteo Petruzzellis on the comments below and lets talk. Sagar Paul Dinker Charak Dinesh Sharma Sachin Dharmapurikar Saurabh Gupta

Most data infrastructures were designed for human reporting, not AI. That’s why so many initiatives stall. 👉 Use our AI Readiness Scorecard and Framework to see if your data foundation can fuel enterprise AI.