Over the past year, many of the most substantive conversations I’ve had with CPOs, CIOs, and leaders at start-ups and growth-stage technology companies have revolved around a common, often unstated question: What happens to enterprise software when users no longer interact with it the way it was designed to be used?
For procurement leaders, this question surfaces around control, trust, and accountability.
For technology providers, particularly those building and scaling solutions, it surfaces around differentiation, relevance, and long-term value creation.
That convergence is why I view AI agents as one of the most consequential trends in enterprise B2B technology today. Not because of the novelty of large language models, but because of what they represent: a structural shift in how intent, data, and decision-making come together. For procurement and supply chain leaders, this shift is no longer theoretical. For technology providers, it is becoming a defining strategic test.
A Lesson From an Earlier Inflection Point
Early in my career, I remember working with Ariba Operating Resource Management System (ORMS) – (yes this is what it was called 😀 ) at a time when workflow was emerging as a true competitive differentiator. What set it apart was not simply functionality, but visibility. The workflow designer made approval paths tangible — POs, requisitions and any electronic or Eform could be modeled, demonstrated, and understood visually in term of approval state.
At the time, this was a sharp contrast to solutions like SAP SRM, which often struggled to show even basic approval processes in a way business users could clearly grasp. That lack of visibility mattered. Buyers hesitated when they could not see how work actually moved through the system. This workflow visibility combined with reporting became a game changer. It shifted buying decisions because it reduced ambiguity and increased trust. Users did not just assume the system worked—they could see how it worked.
That moment is worth remembering, because the pattern is repeating.
Why This Moment Matters
Enterprise software has historically evolved by layering on more capability — more workflows, more configuration, more dashboards. When adoption lagged, the response was better UI/UX, better design and not a fundamental rethink of interaction of humans and computers.
Generative AI and AI agents change that equation. Instead of learning systems, users increasingly express intent: Where am I exposed to supplier risk? What should I renegotiate next? How do I protect margin without disrupting supply?
What makes this shift unavoidable is convergence. Mature LLMs, accessible agent frameworks, enterprise-grade security tooling, and years of accumulated structured and unstructured data have all matured at the same time. The result is a new interaction model that feels as significant as the move from command lines to graphical interfaces—and arguably more disruptive.
From User Experience To Outcome Experience
Traditional procurement platforms were designed around processes. Users navigated workflows, followed steps, and consumed outputs through reports and dashboards.
Agent-driven interaction reverses that logic. The user starts with the outcome. The agent interprets intent, reasons across structured and unstructured data, invokes workflows across systems, and returns a recommendation—or executes it.
For CPOs, this lowers friction but raises stakes. Decisions happen faster, but they also risk becoming opaque if not governed correctly. Just as workflow visibility once built trust, explainability will now define it.
Four Questions Every CPO Should Be Asking Now
As this shift accelerates, CPOs should anchor their technology strategy around four critical questions:
1. Where does decision authority sit—human or agent?
As agents recommend suppliers, flag risks, or trigger actions, CPOs must define where automation is acceptable and where human oversight is mandatory. This is not a configuration issue; it is a governance decision.
2. Can the system explain its recommendations in business terms?
Trust will determine adoption. If an agent cannot clearly articulate why a recommendation was made—what data it used, what assumptions it applied, and what trade-offs it considered—CPOs will hesitate to rely on it for material decisions.
3. How effectively does the platform reason across fragmented data?
Procurement decisions increasingly depend on unstructured inputs—contracts, supplier communications, market intelligence, ESG disclosures. Platforms optimized only for structured ERP or S2P Suite data will struggle as agents become the primary interface.
4. What happens to the procurement operating model?
As agents automate analysis and execution, procurement roles shift toward exception management, supplier strategy, and value orchestration. Skills, roles, and accountability models must evolve accordingly.
What This Shift Means For Technology Providers
The Ariba ORMS example is instructive for today’s technology providers. At that time, workflow visibility—not just workflow capability—became the differentiator. Today, AI agents face a similar test. The foundational components of agent-based systems are rapidly commoditizing. LLMs, orchestration frameworks, vector databases, and enterprise AI tooling are broadly accessible.
Differentiation will not come from having an agent, but from what sits beneath it:
- Embedded domain intelligence, not generic automation
- Decision governance and explainability, not black-box outputs
- Outcome reliability, not surface-level AI features
Providers that cannot clearly show how decisions are made will face the same skepticism once directed at opaque workflow engines.
Suites Vs. Best of Breed in an Agent-Driven World
The long-standing suite versus best-of-breed debate does not disappear — it evolves.
Suites benefit from unified data models and end-to-end process visibility, enabling agents to reason across sourcing, contracting, planning, and execution. This supports broader orchestration but may dilute depth.
Best-of-breed solutions retain an advantage in specialization and analytical rigor. However, without a compelling agent narrative, they risk becoming invisible components under a higher-level orchestration layer.
For CPOs, the right question is no longer “Which solution is better?” It’s “Which ecosystem enables agents to deliver trusted, explainable outcomes across the wider source to pay value chain?“
Will The Prompt Be the Differentiator?
Prompts and prompt design will become the dominant interaction surface—but they will not be the competitive advantage.
Two platforms can receive the same prompt and deliver very different outcomes based on data architecture, reasoning logic, governance rules, and embedded expertise. So as interfaces fade, decision quality becomes the new battleground.
- For CPOs, the mandate is clear: demand transparency, accountability, and control.
- For technology providers, the challenge is sharper: prove that your platform still matters when the screen no longer does.
The interface may disappear. Strategic relevance cannot.
How Liberis Consulting Helps Navigate This Shift
For start-ups and growth-stage technology providers, the move toward agent-based interaction creates both opportunity and confusion. The market is crowded with vendors leveraging similar AI tooling, making differentiation increasingly difficult. Liberis Consulting works with technology companies that recognize this moment as a strategic inflection point. We help providers:
- Define where intelligence truly lives within their platform
- Design agent strategies that prioritize explainability, trust, and control
- Align product direction with how CPOs evaluate value and risk
- Position effectively against suites and adjacent competitors
- Translate technical capability into clear, credible market narratives
Just as workflow once separated leaders from laggards, agent intelligence will now determine who shapes the next generation of procurement technology. Liberis Consulting helps ensure your platform is on the right side of that divide.
Learn more at Liberis Consulting.
