Did you miss the recent webinar “From Pilot to Payoff: 5 Strategies to Turn AI into ROI,” featuring Andrew Bartolini, founder and chief research officer for Ardent Partners, and Rinus Strydom, chief revenue officer for Pactum AI?
The webinar highlighted how leading procurement organizations are breaking the pattern to build AI programs that actually pay off, including five proven strategies that separate the AI achievers from the AI wannabes.
Today marks the second of a three-part article series that brings forth the key points from the webcast, including a link to the event.
Artificial intelligence is reshaping procurement, but the path to measurable ROI is far from uniform. While leaders express optimism—nearly three-quarters of CPOs expect AI to have a transformational impact within two to three years—many procurement functions remain early in their maturity journey. The gap between enthusiasm and tangible outcomes highlights a foundational truth: AI only generates value when paired with strong execution, thoughtful prioritization, and the right conditions for adoption. Lessons from CPOs, pilots, and early adopters reveal how organizations can bridge this gap and unlock meaningful ROI.
The Growing Pressure to Deliver AI-Driven Results
Procurement leaders increasingly feel pressure to do more with less. Budgets remain constrained, workloads continue climbing, and talent shortages persist across the function. Against this backdrop, AI presents an appealing solution: an opportunity to automate manual work, uncover deeper insights, and accelerate sourcing outcomes. Yet, even as the technology advances rapidly, many organizations remain unsure how to translate AI investment into measurable cost reduction, productivity gains, or strategic influence.
The challenge is not simply that AI is new. It’s that procurement’s operating environment (i.e., fragmented data, inconsistent processes, decentralized buying behaviors), creates obstacles that technology alone cannot solve. CPOs acknowledge that AI can deliver value across strategic sourcing, contract management, supplier intelligence, and transactional workflows. But to achieve ROI, procurement must align its people, processes, and data structures with the capabilities AI offers.
Choosing the Right Projects: The First Critical Step
One of the most consistent lessons from early adopters is that project selection is everything. Not all AI opportunities deliver equal value. Some use cases create incremental efficiency; others fundamentally change workflows and unlock significant savings.
The highest-ROI projects often share three characteristics:
- They address a high-friction process.
- They unlock significant financial or time impact.
- They are feasible with existing data and systems.
For example, many teams begin with AI-assisted sourcing negotiations. These projects produce measurable outcomes (often 8–15% incremental savings), while requiring minimal integration. They also provide a visible, politically compelling win. In contrast, organizations that start with overly complex, multi-system integrations often struggle to build early momentum, losing stakeholder confidence and delaying value realization.
Matching Technology to the Problem (and Avoiding Overreach)
A second key insight is the importance of selecting tools that match both the project and the organization’s readiness. Procurement leaders increasingly recognize that AI solutions vary dramatically in maturity, particularly in generative and agentic capabilities. A solution that thrives in one environment may underperform in another due to data complexity, workflow differences, or unclear business rules.
To overcome this, leading organizations insist on:
- Pilots using their own real data, not vendor-supplied demo environments.
- Execution-focused evaluation, assessing not just insights but outcomes.
- A crawl–walk–run approach, expanding scope only once value is proven.
This disciplined approach prevents organizations from over-investing in tools that cannot deliver on their promises and ensures that expansion accompanies measurable success.
Start With the Data You Have — Not the Data You Wish You Had
Historically, procurement technology initiatives stalled due to poor data quality. AI changes this equation. Modern AI agents can ingest unstructured data, including GL exports, contract PDFs, and supplier communications, and interpret it with surprising accuracy. Instead of waiting years to cleanse and harmonize data, leading organizations adopt the philosophy: start with what you have and improve as you go.
This approach is particularly powerful in spend analytics, contract review, and sourcing recommendations. AI can now classify spend, extract contract terms, identify suppliers, and detect price variance with minimal human intervention. The result is faster time-to-value and a reduced burden on IT.
Don’t Abandon Traditional Technology Governance
Even as AI unlocks new capabilities, traditional technology success factors still matter. Early adopters emphasize that AI projects succeed when they incorporate:
- Clear executive sponsorship, particularly from finance.
- Well-defined policies for compliance and savings measurement.
- Iterative deployment, paired with real-time user feedback.
- Change management tailored to different roles.
Without this foundation, AI delivers insights but fails to change user behavior — a common reason ROI stalls.
Leverage Provider Expertise and Category Benchmarks
Procurement technology providers accumulate insights across thousands of sourcing events and categories. Leading CPOs tap into this knowledge to understand:
- Expected savings ranges by category.
- How many suppliers should be included for optimal outcomes.
- Which processes benefit most from AI intervention.
- Where agentic automation drives the fastest wins.
This benchmarking accelerates decision-making and reduces the risk of pursuing low-impact projects.
AI will become an increasingly central pillar of procurement strategy. But ROI will not come from technology alone. Instead, it will come from the discipline of selecting the right projects, validating the right tools, starting with available data, reinforcing governance, and leveraging provider expertise. Procurement teams that master these fundamentals will move beyond experimentation toward a scalable, repeatable, high-ROI AI operating model. Those that do not risk falling behind as agentic AI rapidly reshapes the competitive landscape.
