Ecosystem Architecture: The Missing Discipline in Composable eCommerce Projects
April 30, 2025

Composable Commerce’s Silent Crisis: Who Owns the Data, and Who Should?

May 2, 2025

The Illusion of Control in a Composable World

In the age of composable commerce, where agility and modularity are seen as cornerstones of retail innovation, there’s a critical fault line quietly deepening beneath the ecosystem. It’s not the complexity of microservices, not the evolving front-end interfaces, nor even the pace of integration. The real crisis — one barely acknowledged in boardrooms and platform roadmaps — is the data. Specifically, who owns it, who controls it, and what happens when everyone wants to be the data backbone but no one truly is.

Composable commerce, especially as expressed through MACH principles (Microservices, API-first, Cloud-native, and Headless), has given birth to extraordinary flexibility. Retailers can now cherry-pick best-of-breed services to build tech stacks that are tailor-fit for speed, innovation, and customer delight. But this architecture, as empowering as it is, has brought a dark passenger along for the ride: siloed data fragmentation at a scale and complexity we haven’t seen before.

What used to be a centralized monolith has now splintered into dozens of specialized services — each one a master of a niche domain, and each one carrying its own slice of the data pie. The customer doesn’t see this. To them, it’s one store, one experience. But behind the curtain, it’s a fragmented mess. Every integration is another potential silo. Every SaaS vendor is another potential data fortress. The very APIs that make composable commerce so attractive are also accelerating the uncontrolled propagation of customer and business data across an increasingly tangled web.

When Everyone’s the Backbone, No One Is

Ironically, this isn’t just a technical mess. It’s a strategic failure.

Retailers, the very entities who should be the ultimate stewards of their customer data, are often the least in control. The ambition of every third-party MACH player to become the “data backbone” of the ecosystem — under the banner of analytics, AI, personalization, or operations — has quietly turned into a land grab for data sovereignty. Everyone wants to be the central nervous system, the single source of truth, the default database where insights are generated and value is extracted. The only one not stepping up to claim that role is, alarmingly, the retailer.

And while GDPR and other regulations provide a framework for privacy and data protection, they do little to address the deeper issue here: governance and access across ecosystems. This is not about protecting customer information from hackers; it’s about deciding who has the right to use that information in the first place, how it’s shared, where it lives, and who governs its flow. The fact that this governance is often implicit — buried in contractual fine print or loosely defined API specs — should worry anyone serious about long-term platform resilience and customer trust.

Platforms, particularly headless commerce solutions, are in a unique position to solve this. They are the bridge where all systems meet. They see the customer first. They orchestrate much of the traffic between services. They have both the vantage point and the infrastructure to become true stewards of data on behalf of their retail clients. And yet, many of them are either oblivious to this responsibility or too slow to recognize the opportunity. Instead of becoming guardians of retailer sovereignty, they too are jostling to become just another data hub.

The promise of composable commerce is that each service can do one thing brilliantly. But when every service is building its own customer profile, tracking its own event history, storing its own partial view of behavior and identity, you don’t get brilliance — you get chaos. You get twenty systems with twenty slightly different definitions of who the customer is, none of them fully accurate, and none of them governed under a coherent structure. Integration, in this scenario, isn’t solving the problem — it’s multiplying it.

API Access Isn’t Consent: Reclaiming Data Sovereignty

What’s missing is a common language for data in the MACH ecosystem, a shared protocol for communication, and above all, a deliberate architectural decision that the retailer’s data lake must be the backbone, not any one vendor or SaaS tool. This backbone doesn’t mean centralizing every byte of data in one place, but it does mean enforcing a layer of intelligence, curation, and governance that gives the retailer complete ownership and control over who accesses what, when, and why.

APIs, as they’re currently implemented, often make things worse. They open the floodgates, offering full access instead of filtered insight. That’s the red flag. Any vendor that offers wide-open, unaudited access to sensitive customer data under the guise of convenience is not respecting data sovereignty — they’re exploiting it. The future lies in agentic APIs: interfaces that provide just enough access to serve a specific function, but no more. Intelligent gates, not open doors.

Composable systems need guardrails. And those guardrails must be built by — and for — retailers. It’s not enough to trust that your tech stack will act in your best interest. You need to be able to enforce it. You need to start from the principle that all data belongs to you, and all access must be provisioned through a layer of governance that you define. This isn’t just about compliance. It’s about maintaining competitive advantage. Because in the age of predictive models and AI-driven insights, who controls the data controls the future.

AI will only amplify this problem if we don’t address it now. Machine learning models depend on clean, contextual, and coherent data. If your AI is trained on fractured, redundant, or poorly-governed datasets, you’ll not only get bad outputs — you’ll train your organization to trust the wrong signals. Worse, you’ll become dependent on vendors whose models were trained with your data, but whose benefits you can never fully internalize. You’ll be handing over competitive intelligence in exchange for service convenience.

The platforms and vendors won’t fix this for you — not unless you make it a condition of the relationship. And the truth is, there are already some MACH-native data lake and orchestration services that can act as neutral territory. The missing piece is intention. Retailers need to open the conversation, assess their stack, ask for guarantees, and start governing their own data ecosystem like it’s the core of their brand — because it is.

The Retailer Must Rise

The most important role in a composable architecture doesn’t belong to the checkout service, the CMS, the personalization engine, or even the AI layer. It belongs to the entity that interfaces with the customer. The retailer. The one with the brand, the loyalty, and the customer trust. The only one that should — and must — own the relationship and everything that flows from it.

And that brings us to the real question that every retail CTO, head of digital, and transformation leader needs to be asking right now: Do you have complete ownership of all the data being generated in your MACH stack? And do you control how it propagates across your ecosystem?

If the answer is no — or worse, if you’re unsure — then you have a silent crisis on your hands. Composable commerce doesn’t work without data orchestration. APIs don’t mean trust. And no matter how modern your stack is, without sovereignty, it’s not truly yours.

The retailer must rise. Not as another endpoint in a MACH map. But as the architect of their own future.

At Kore Business, we’re actively exploring this very challenge — and engaging with solutions that respect and reinforce retailer sovereignty. Many of our clients across industry and retail are already raising these concerns, and we believe this is not a side issue — it’s central. That’s why the data layer is a core focus within our Ecosystem Architecture service. We don’t see it as a technical afterthought, but as a foundational question of trust, autonomy, and long-term strategy.

If this conversation resonates with where your organization is heading — or where it needs to go — we invite you to go deeper. The time to rethink your data architecture isn’t when it breaks. It’s now, before someone else decides what you get to own.

Rafael Esberard is a Digital Ecosystem Architect and Strategic Consultant specializing in retail digital transformation. As the founder of KORE Business, he helps companies design, govern, and evolve their digital ecosystems through a pragmatic, business-driven approach to composable commerce, MACH architecture, and AI integration. Rafael works alongside retailers and industry leaders to guide the selection, validation, and orchestration of best-fit solutions across complex multi-vendor landscapes, ensuring scalability, agility, and long-term ecosystem health. His expertise spans omnichannel strategies, AI-driven ecosystem optimization, and accelerating time-to-value and time-to-market across digital transformation projects. By bridging technology evolution with real-world business needs, Rafael enables clients to transform ambition into sustainable competitive advantage.

#ComposableCommerce #MACHArchitecture #DigitalTransformation #RetailInnovation #EcosystemArchitecture #FutureOfCommerce #MACHAlliance #DataSovereignty #RetailTech #HeadlessCommerce #APIGovernance #CustomerData #EcommerceStrategy #DigitalRetail #DataBackbone #BusinessEcosystems #RetailTechnology #DataOwnership

Leave a Reply

Your email address will not be published. Required fields are marked *

April 30, 2025
The MACH Illusion: Blaming the Mirror for the Mess
The MACH Illusion: Blaming the Mirror for the Mess
Introduction: When Freedom Feels Like Free Fall Composable commerce arrived with bold promises: to unshackle businesses from outdated monoliths and give them the freedom to build […]