On a typical morning, thousands of customers place orders through the Starbucks app before they even reach the store. Within seconds, payment is processed, loyalty points are updated, the order is routed, and inventory adjusts for ingredients and items. A barista sees the order. The customer gets a pickup time. It feels simple. It is not. And here is the part most executives are underestimating. The complexity behind that transaction is now one of the largest sources of hidden revenue loss in retail. This is no longer about e-commerce versus in-store. Every transaction, regardless of where it originates, now depends on the same underlying system.
As Arjun Wadwalkar, Senior Manager, Product Management with over 10 years of experience, puts it, “What looks like a seamless customer experience is often powered by deeply fragmented systems behind the scenes.”
The Hidden Cost of Fragmentation
Retailers have spent the last decade investing in digital channels. What they have not done is rebuild the infrastructure underneath those channels. The result is what I would call a fragmentation tax. It shows up in places most companies track separately, but rarely connect. Payment authorization failures, inventory inaccuracies, order cancellations and substitutions, pricing and promotion mismatches, and manual reconciliation and refunds. Individually, these look like operational inefficiencies. Together, they form a systemic problem.
“Most retailers don’t realize they’re paying a fragmentation tax because these failures are tracked in isolation, not as a system-level problem,” says Arjun Wadwalkar.
Consider payments alone. Across e-commerce, authorization failure rates can range from 5 percent to 15 percent, depending on issuer behavior, routing logic, and retry strategies. That is not fraud. That is recoverable revenue that never converts. Inventory tells a similar story. Industry studies have consistently shown store-level inventory accuracy hovering between 60 percent and 70 percent. That means roughly one out of three inventory records is wrong. Now layer in digital ordering. When a system says an item is available and it is not, the result is not just a bad experience. It is a canceled order, a refund, or a lost customer. These are not edge cases. They are structural leaks.
The Cracks Are Already Visible
In recent years, McDonald’s has expanded across kiosks, mobile apps, delivery platforms, and drive-thru, creating multiple entry points into the same store operations. But when pricing, promotions, and order flows are not perfectly synchronized across these channels, customers can experience inconsistencies depending on how they order.
“What we’re seeing isn’t system failure, it’s coordination failure across channels that were never designed to work as one,” explains Arjun Wadwalkar.
Similar challenges are playing out in retail. Companies like Target and Walmart have invested heavily in buy-online-pickup-in-store and same-day fulfillment, but inventory accuracy and real-time synchronization remain persistent issues, leading to canceled orders or substitutions when items are not actually available. Even Starbucks has faced challenges with mobile ordering at scale, where centralized order intake does not always reflect real-time store capacity, resulting in long wait times and order stacking. These are not outages. They are coordination failures that emerge when distributed demand meets fragmented systems.
Retail Was Built for a Different World
Most retail systems still reflect a legacy assumption. Transactions start at the register. So the architecture evolved into separate systems. POS runs the store, e-commerce runs online, delivery platforms integrate separately, and inventory syncs across systems. This is not a unified system. It is a collection of systems connected over time. That worked when digital was incremental. It breaks when digital becomes primary.
Omnichannel Solved the Experience. Not the System
For years, omnichannel was the strategy. But omnichannel largely meant stitching together systems that were never designed to behave as one. That created multiple sources of truth, latency between systems, fragile integrations, and continuous reconciliation. It improved the experience customers see. It did not fix the infrastructure that powers it.
The Shift That Is Actually Happening
Retail is now undergoing the same transformation enterprise software went through. From monolithic systems and batch updates to distributed, API-driven, event-based architectures. At the center of this shift is a new model. The commerce engine. Companies like Shopify, Square, and Toast are already building toward it.
“The future isn’t about adding more channels. It’s about building a single system where every transaction resolves consistently,” says Arjun Wadwalkar.
The principle is simple. Every transaction resolves through a single system of record. Orders, payments, inventory updates, and returns all flow through one core platform, regardless of origin. That is what eliminates drift.
The Store Is No Longer the Center
In this model, the register is no longer the brain of the system. It is one interface among many. Mobile apps, kiosks, delivery platforms, APIs, and even physical sensors all connect to the same transaction layer. Amazon provides a clear view of where this is going.
There is no checkout. Every action generates a digital event. The system maintains a live transaction. Payment happens automatically when the customer exits. The store is no longer separate from the system. It is the system.
AI Will Expose Every Weak Link
The next disruption is not another channel. It is a different type of customer. AI agents. Today, they assist with decisions. Tomorrow, they will execute transactions. That changes the model. Human to agent to API to commerce engine. And it introduces a hard constraint. Humans tolerate friction. Machines do not.
“AI agents won’t tolerate the inconsistencies humans ignore. They’ll simply move to the next best option,” notes Arjun Wadwalkar.
If your system returns incorrect inventory, inconsistent pricing, or failed payment responses, the agent will not retry. It will select another vendor instantly.
Infrastructure Becomes the Differentiator
As transactions become increasingly automated, reliability becomes a competitive advantage. Commerce systems must answer, in real time. Is the item actually available? What is the final price after all the adjustments? Can the order be fulfilled within a defined window. If the system answers correctly, the transaction proceeds. If it does not, the transaction disappears. Often without the retailer ever knowing it existed.
The Bottom Line
Retail is not being disrupted by better apps or better storefronts. It is being redefined by infrastructure. Right now, most companies are trying to scale distributed demand on top of fragmented systems. That approach is already leaking revenue through failed payments, inaccurate inventory, and broken order flows. And those leaks will only grow as commerce becomes more real time and more automated.
“In the next phase of retail, infrastructure reliability will matter more than front-end experience,” concludes Arjun Wadwalkar.
Because in the next phase of retail, the companies that win will not be the ones with the best customer experience. They will be the ones whose systems do not break when everything connects to them.
