Yesterday’s disruptor is today’s legacy

Legacy software used to be easy to spot. It came with dated interfaces, long implementations, armies of consultants, and training manuals thick enough to qualify as office furniture.

‍

The new version looks better. It arrives with a clean front door, a more recent founding date, a multibillion-dollar valuation, an expanding product portfolio, a transformation office, dozens of AI agents, and a management layer for those agents.

‍

Intake-and-orchestration platforms deserve credit for identifying a painful problem. Traditional procurement software, like Coupa, gave employees a rough starting point. Requests arrived through email, Slack, spreadsheets, and whichever side channel seemed least likely to summon a procurement form. A single entry point made the experience cleaner. It helped procurement coordinate finance, legal, IT, security, and purchasing around the request.

‍

But scale has a funny way of aging software. The architecture gets stretched. The product starts collecting baggage. Services move in. Packaging multiplies. The AI story has to work harder to make the whole thing feel coherent. That’s usually when legacy stops being something a company competes against and starts becoming something it resembles.

‍

Legacy has nothing to do with age

The SaaS industry loves treating “legacy” as a synonym for “old.” That definition lets every younger vendor declare itself modern by producing a recent incorporation date and hiring a designer who knows what rounded corners are.

‍

Legacy is better understood as a pattern of behavior. The platform expands faster than the architecture beneath it. More workflows become more products. More products create more packages. More packages need more services and SOWs. Every major market shift gets another layer, another name, and another reason to reopen the contract.

‍

A company can be young and still carry an old operating model. It can have a beautiful interface while depending on increasingly complex packaging, specialist support, custom engineering, and a growing ecosystem around the product. Legacy doesn’t need decades. Sometimes hypergrowth gets it there faster.

‍

Orchestration was built for the last procurement revolution

The original insight behind procurement orchestration was simple and valuable: capture the request, route it through the right teams, connect the systems already doing the work, and give employees one clean place to begin. For the procurement market of 2020, that was a revolutionary idea. Then generative AI reset the enterprise software market.

‍

The AI era asks for something deeper than a front door and a workflow map. Agents need complete procurement context: suppliers, contracts, approvals, purchase orders, invoices, budgets, payments, policies, and the financial record underneath them.

‍

That creates a problem for orchestration platforms built primarily to sit across existing systems. They can make fragmented tools feel connected while the underlying records, exceptions, supplier context, and financial truth still live elsewhere. AI can only act on the portion of that reality the platform can reliably see. The cleaner the orchestration layer becomes, the easier it is to overlook what still sits underneath. And AI has very little patience for fragile architecture.

‍

The agent story is becoming a symptom of the platform story

The first wave of procurement AI brought task agents. Then came dozens of specialized agents. Soon, those agents needed role-based managers to coordinate them.

‍

Apparently, even artificial employees need middle management. We’ve already unpacked the cape phase of procurement AI in detail. 

‍

The broader issue is what this escalation says about orchestration platforms. They were built to route work. AI now requires them to act on that work across policies, budgets, suppliers, contracts, invoices, approvals, and financial records. That’s a significant architectural leap. The naming just makes it sound much tidier.

‍

Legacy platforms often respond to a new architecture cycle by placing another layer above the original one. The new layer gets the launch event. Customers get the archaeology project.

‍

Hypergrowth leaves a mark

Some of the fastest-growing orchestration platforms have expanded from intake into sourcing, supplier management, contracts, purchase orders, invoicing, payments, budgets, risk, apps, AI agents, governance, and custom automation in only a few years.

‍

That growth is impressive. It also creates pressure in a market already drowning in it. Enterprise customers want broader functionality. Investors want a larger market. Sales teams need more expansion paths. Product teams move into adjacent workflows. Services organizations appear to help customers manage everything the platform has become. Each addition can make perfect sense on its own. That’s how the pile gets built. Legacy software rarely begins bloated. It earns the bloat one sensible roadmap decision at a time.

‍

Eventually, the platforms that made procurement simpler will need increasingly elaborate explanations of their products, packages, services, AI capabilities, and implementation options. The front door stays clean. Everything behind it starts resembling a storage unit.

‍

One front door. A lot more ways to pay.

Orchestration platforms entered the market with a simple proposition: give employees one clean place to start. Their commercial expansion has been considerably less singular.

‍

New capabilities create new expansion paths. Those paths become add-ons, upgrades, service packages, consumption models, variable pricing, and strategic reasons to revisit the contract. The user still gets one front door. The invoice has discovered the open floor plan.

‍

There’s nothing unusual about software companies monetizing a broader platform. The pattern becomes worth watching when every response to the market creates another product, package, service, or upgrade path. The interface can remain simple while the buying relationship becomes increasingly elaborate.

‍

“Enterprise-grade” has a long history of becoming the polite phrase for “please contact your account team to find out which part costs extra.”

‍

The system around the platform keeps getting bigger

Orchestration platforms increasingly arrive surrounded by transformation advisors, custom engineers, Silicon Valley AI engineers, governance packages, integration specialists, and strategic services.

‍

There can be value in that support. Procurement is complex, and enterprises need help redesigning processes, driving adoption, and deploying AI responsibly. The question is what the surrounding ecosystem says about the platform.

‍

Legacy software has always been good at turning complexity into proof of enterprise readiness. The implementation partner, the specialist, the transformation office, and the custom development team become evidence that the platform can handle a sophisticated organization.

‍

Sometimes, though, they’re just evidence that the organization has to do a lot of work to make the platform sophisticated enough. A platform can remain visually modern while developing a very traditional need for people around it. Rounded corners don’t make the operating model any lighter.

‍

Legacy doesn’t wait for wrinkles

The orchestration era didn’t age into legacy. It scaled into it. Intake became a platform. The platform became a portfolio. The portfolio became an ecosystem. Then AI arrived on top and started handing out promotions.

‍

That’s how legacy takes shape. The platform expands, services become structural, packaging gets harder to follow, and the interface stays clean enough to hide the weight. Eventually, all that weight lands on the foundation.

‍

And the next era of procurement won’t be won at the entrance. It’ll be won by the architecture underneath.