Enterprise software vendors sell complexity. That’s the business model. Build something powerful, make it configurable enough to handle every edge case the largest customer can dream up, and then charge everyone else for the privilege of carrying that complexity whether they need it or not. The top 1% of customers drive the roadmap, and the other 99% subsidize features they never asked for, will never use, and that actively make the product harder to implement, more expensive to operate, and slower to evolve.
This isn’t cynicism: it’s Clayton Christensen’s “Innovator’s Dilemma”, and it has played out in every layer of the enterprise technology stack for decades. Incumbents don’t fail because they stop innovating. They fail because they keep innovating in the wrong direction, building for their most demanding customers while the broader market goes looking for something simpler. A new entrant arrives, delivers 80% of the value in a radically simpler and cheaper package, and methodically moves upmarket until the incumbent’s complexity advantage becomes a complexity liability.
This pattern played out in enterprise Java. It played out in CRM. It’s playing out right now on commerce platforms. And it’s about to play out in order management.
A Pattern as Old as Enterprise Software
Before getting into commerce, it’s worth looking at where this pattern has already run its course, because the parallels are striking.
In the early 2000s, the enterprise Java application server was the center of gravity for building serious web applications. BEA WebLogic and IBM WebSphere were the dominant platforms – powerful, feature-rich, expensive, and complex. Deploying a Java EE application meant licensing a platform that cost hundreds of thousands of dollars, staffing a team of specialists to configure and tune it, and accepting deployment cycles measured in weeks or months. Having led the WebLogic Enterprise Java platform group at Oracle following the BEA acquisition, the view from inside the cathedral was clear: the platform did everything, and the market was starting to realize it didn’t need everything.
The Innovator’s Dilemma was playing out in real time. WebLogic and WebSphere kept adding capabilities (clustering, advanced transaction management, sophisticated security frameworks, elaborate management consoles) because their largest enterprise customers demanded them. Each release became more powerful and more complex, requiring more specialized expertise to operate. The platforms were being optimized for the requirements of banks, airlines, and telcos while becoming increasingly overkill for the vast, vast majority of applications that didn’t need that level of sophistication.
What happened next was textbook. JBoss arrived as an open source alternative that delivered the core capabilities at a fraction of the cost. Spring Framework stripped away even more complexity, giving developers a dramatically simpler programming model that didn’t require the heavyweight application server at all. Then cloud infrastructure delivered the final blow: application developers realized they needed a few primitives (containers, queues, object storage, a load balancer) and the rest of the application server’s feature set was unnecessary weight. The vast, vast majority of what WebLogic and WebSphere sold turned out to be complexity that developers were happy to leave behind.
The same pattern played out in CRM. Siebel Systems built the most comprehensive, feature-rich CRM platform the enterprise world had ever seen. It had modules for everything (pharmaceutical sales tracking, insurance claims workflows, telecom provisioning) because Siebel’s largest customers in those verticals demanded them. The platform became enormously expensive to implement, painful to customize, and slow to evolve, with implementations routinely running into the millions and taking over a year. Salesforce didn’t beat Siebel by building a better Siebel. It beat Siebel by delivering the 80% that mattered (contact management, pipeline tracking, reporting) in a cloud-native model that was simpler, cheaper, and faster to deploy. Siebel’s feature depth, the very thing its enterprise customers had demanded, became a liability rather than an asset once the broader market decided it didn’t need all that depth.
These aren’t isolated examples, but the rule. Incumbents sell complexity because their biggest customers demand it, and then the rest of the market pays the price. It happens in every layer of the stack, in every era of enterprise software.
The Commerce Platform: Same Story, Different Decade
Fifteen years ago, choosing a commerce platform was one of the most consequential technology decisions a brand could make. It defined the architecture, dictated the roadmap, and if we’re honest, locked teams into multi-year relationships that often felt more like hostage situations than partnerships.
The market was sprawling. Demandware. Hybris. ATG (Oracle). Magento. IBM WebSphere Commerce. Volusion. Miva Merchant. osCommerce. X-Cart. 3dcart. BigCommerce. Shopify. WooCommerce. PrestaShop. Zen Cart. NetSuite SuiteCommerce. Salesforce Commerce Cloud (née Demandware). SAP Commerce Cloud (née Hybris). Every year, a new entrant arrived promising to be the platform that finally “got it right,” and every year the complexity grew.
The dominant business model for the biggest players was the GMV take rate: a percentage of every dollar flowing through the platform. The logic was seductive: align the platform’s success with the merchant’s success. In practice, however, it created a perverse dynamic where platform costs scaled dramatically alongside revenue, often outpacing the incremental value the platform was actually delivering. A brand doing $50M in GMV was paying millions to a platform for largely the same functionality it had at $5M.
The backlash was inevitable. Brands started demanding (and getting) flat SaaS pricing models with predictable costs and no penalty for success. The shift from GMV-based to SaaS-based pricing was the market telling platform vendors that they aren’t as special as they think they are, because that’s fundamentally what commoditization looks like. When the market decides that 80% of what a vendor does is table stakes and the remaining 20% isn’t worth the premium, the pricing model is the first thing to break.
Shopify: The Proof That 80/20 Always Wins
Just as Spring and cloud primitives displaced the enterprise Java application server, and just as Salesforce displaced Siebel, Shopify displaced the complex commerce monoliths by delivering the 80% that matters in a radically simpler package.
While legacy platforms were busy building increasingly elaborate feature sets (trying to be all things to all merchants, chasing the requirements of their most complex enterprise customers) Shopify made a bet that the market didn’t want more features. It wanted fewer decisions, the ability to go live in days instead of months, and the freedom to stop paying consultants to configure things that should just work out of the box. The market responded emphatically.
Shopify didn’t win by being the most feature-rich platform. It won by being radically simpler, dramatically faster to deploy, and meaningfully cheaper to operate. It delivered the 80% that matters (product catalog, checkout, payments, basic order management, storefront) in a way that was accessible to everyone from a single-person DTC brand to a publicly traded enterprise. This is the Innovator’s Dilemma in action: enter at the low end with a “good enough” product, then relentlessly improve until the incumbents’ complexity advantage becomes a complexity liability.
The legacy platforms scoffed. “Shopify can’t handle complex B2B.” “Shopify doesn’t support multi-currency natively.” “Shopify isn’t enterprise-grade.” Then, one by one, Shopify closed those gaps not by building bloated monoliths, but by creating an ecosystem of apps, APIs, and partners that let merchants assemble exactly what they needed without carrying the weight of what they didn’t. The dismissal from incumbents is always the most reliable signal that disruption is underway.
Choosing Shopify isn’t just a cost play: it’s an agility play. Shopify ships product at a pace that legacy platforms can’t match. Shopify Editions, Shop Pay, Hydrogen, the constant stream of checkout innovations: when a platform vendor is iterating that fast, brands aren’t just saving money. They’re gaining capabilities that would have taken their old platform years and millions of dollars in custom development to deliver. Incumbents weighed down by two decades of accumulated complexity simply cannot move at that speed.
The New Frontier: Social Selling and Agentic Commerce
The commoditization of the commerce platform is only accelerating, because the definition of “where commerce happens” is fundamentally changing.
TikTok Shop has grown from a curiosity to a legitimate commerce channel doing billions in GMV, and it’s just one of a myriad of social selling channels reshaping how consumers discover and buy products. The traditional assumption that a brand’s .com is the center of the commerce universe is increasingly outdated. Transactions are happening on social platforms, in marketplaces, through live-stream shopping, and in places that didn’t exist five years ago.
Then there’s agentic commerce: the emerging reality where AI agents act on behalf of consumers to research, compare, and purchase products. Standards like MCP (Model Context Protocol) and onX (Order Network eXchange) are laying the foundation for a world where a significant share of transactions happen without a human ever visiting a product page. The traditional commerce platform, built around the assumption of a human browsing a website, wasn’t designed for a world where an AI agent is making the purchase decision on behalf of the consumer.
This has massive implications for the entire commerce stack. When transactions originate from a dozen different social channels and AI agents rather than a single brand-owned storefront, the commerce platform layer becomes even more commoditized, and the order processing layer beneath it becomes even more critical. The ability to ingest orders from any source, route them intelligently, and orchestrate fulfillment across a complex network of warehouses, 3PLs, and drop-ship vendors: that’s the capability that actually matters in a world of fragmented, multi-channel, agent-driven commerce.
Now Look at the Middle Layer
The commerce platform is one layer. Below it sits the order management and processing layer: the system that actually orchestrates what happens after the “Buy” button gets clicked (or after an AI agent places the order). Routing orders to the right warehouse, managing inventory across channels, handling splits, backorders, returns, and the thousand edge cases that define real-world commerce operations.
In this layer, the same commoditization pattern is playing out, just ten years behind.
The incumbents here are names like Sterling Commerce (now IBM Sterling) and Manhattan Associates: systems that are 25+ years old, built for a world of linear supply chains and single-channel retail. They’re powerful, no doubt. They’re also enormously complex, expensive to implement, slow to change, and designed for a commerce reality that no longer exists. This is WebLogic in 2008. This is Siebel in 2004.
And the Innovator’s Dilemma is playing out here with textbook precision. These platforms have spent decades adding features to satisfy the requirements of their largest, most complex customers: retailers with thousands of stores, manufacturers with byzantine supply chains, logistics companies with edge cases that apply to almost nobody else. Each feature addition made the platform more powerful for those specific customers and simultaneously more expensive, harder to implement, and more bewildering for everyone else. The implementation timelines tell the story: six to twelve months, seven-figure budgets, armies of consultants. All to route an order to the right warehouse.
To make matters worse, the legacy approach forces brands to cobble together separate systems for connectivity and order management. One platform to integrate the selling channels, another to integrate the fulfillment systems, yet another to handle the order routing logic, and a separate middleware layer to connect everything to the ERP. Each system has its own implementation timeline, its own license cost, its own support contract, and its own team of consultants. The complexity doesn’t just come from any single product: it comes from the architecture itself, where five or six products have to be stitched together just to move an order from point A to point B.
Today’s brands aren’t running single-channel supply chains. They’re selling on Shopify, Amazon, TikTok Shop, wholesale, and B2B simultaneously, fulfilling from their own warehouses, 3PLs, drop-ship vendors, and retail locations. They’re processing tens of thousands of orders a day across dozens of integration points, and they’re doing it with a fraction of the IT resources that were available a decade ago. The Sterlings and Manhattans of the world were never designed for this. They were designed for a world where a retailer had a few warehouses, a single channel, and an IT department with 50 people dedicated to managing the OMS. That world is gone.
The Shopify of Order Operations
This is exactly the problem Pipe17 was built to solve.
What Shopify did for the commerce platform layer, Pipe17 does for the order processing layer. What Spring and cloud primitives did for the enterprise Java stack (strip away the unnecessary complexity and give practitioners just the capabilities they actually need) Pipe17 does for order operations.
The critical difference is that Pipe17 combines connectivity and order management into a single platform. There is no separate integration tool, no middleware layer, no third-party connector sitting between the order management logic and the systems it talks to. Selling channels, fulfillment providers, 3PLs, ERPs: they all connect directly into the same platform that handles order orchestration, inventory synchronization, and fulfillment routing. One platform, not five. That architectural simplification alone eliminates an enormous amount of the cost, complexity, and fragility that defines the legacy approach.
Connect the selling channels: whether that’s Shopify, Amazon, TikTok Shop, or an AI agent placing orders via onX. Connect the fulfillment network. Connect the ERP. Let orders flow. No six-month implementations, no consultants writing custom XML transforms, no seven-figure license fees to process a sales order.
The point isn’t to replace the ERP. Nobody’s ripping out SAP or NetSuite, and frankly they shouldn’t have to. The system of record is the system of record. The layer between the commerce platform and the ERP, however (the layer that handles order orchestration, inventory synchronization, and fulfillment routing) is ripe for exactly the same disruption that has already reshaped every other tier of the enterprise stack.
Total cost of ownership drops dramatically, not marginally, but orders of magnitude lower than systems that cost hundreds of thousands or millions per year. Time to value compresses from months to weeks: connect, configure, go live. Agility increases fundamentally, because when commerce is fragmenting across social channels and AI agents at the pace it is today, the order processing layer needs to adapt just as quickly, and legacy OMS simply can’t. They weren’t built for it. They were built for the 1%, and the 99% is done paying for it.
And then there’s something that’s chronically underrated in enterprise software: the value of simplicity as a feature. When a team can actually manage order operations without a dedicated systems integrator on retainer, because connectivity and order management live in one place rather than scattered across a half-dozen products, that’s organizational freedom.
The Uncomfortable Truth
Here’s what the legacy vendors don’t want brands to hear: they sell complexity, and the vast, vast majority of that complexity is something nobody needs.
The same way developers discovered they didn’t need WebLogic to run a web application, the same way sales teams discovered they didn’t need Siebel to manage a pipeline, and the same way brands discovered they didn’t need a $2M Hybris implementation to sell products online, most brands are going to discover they don’t need a $1M Sterling implementation and three middleware products to route orders to warehouses. The 80% that matters (connectivity, order orchestration, inventory sync, fulfillment routing, returns processing) can be delivered in a single platform, faster, simpler, and cheaper.
This isn’t a knock on what Sterling or Manhattan built. Those systems were engineering marvels for their time, and their time was a different era of commerce. The Innovator’s Dilemma doesn’t say incumbents are stupid or lazy: it says they’re rational actors trapped by the economics of serving their most demanding customers. They built exactly what their biggest clients asked for, and in doing so they priced and complexified themselves out of relevance for the rest of the market. Christensen saw this play out in disk drives. The enterprise software industry has seen it play out a dozen times since.
The brands winning today are the ones with the most adaptable stacks. Any brand that has already made the move to Shopify, or is considering it, has already accepted the premise that simpler, faster, and cheaper beats complex, slow, and expensive at the platform layer. Now it’s time to apply that same logic one layer down.
The commerce platform shouldn’t be a 15-year legacy system. The order management layer shouldn’t be either. The commoditization wave doesn’t stop at the storefront: it’s coming for the entire stack, and the brands that ride it instead of fighting it are the ones that will win.
