“Technology doesn’t transform organizations. People do — with the right technology.”
The Day I Realized Something Had Changed
It happened during a digital transformation project at a large industrial company. We were in the scoping phase of a Procurement solution, facing a team of seasoned buyers. These were professionals with fifteen, twenty years of experience. People who knew their suppliers by first name, who could smell a pricing anomaly before even opening a contract.
And yet, throughout the workshops, something kept coming up in their testimonials:
“We spend 60% of our time on tasks we shouldn’t be doing anymore.”
Chasing a supplier. Consolidating data in Excel. Manually comparing bids. Writing RFP summaries. Checking invoices line by line.
That’s not what these professionals were trained for. And it’s certainly not where their value lies.
That’s when I understood that the real question around AI in procurement wasn’t technological. It was human.
What the Numbers Reveal: An Unprecedented Transformation
Recent data is eloquent and worth pausing on.
According to an AI at Wharton study, weekly use of generative AI in Procurement functions jumped from 50% to 94% between 2023 and 2024 a gain of 44 percentage points in one year, the steepest rise recorded across all enterprise functions. In other words, procurement has become the most active function in adopting generative AI in business.
On the decision-maker side, the ProcureCon CPO Annual Report 2025 indicates that 80% of Chief Procurement Officers consider AI investment a priority for the next 12 months, with 66% making it a high priority.
The Hackett Group reveals that 49% of Procurement teams piloted generative AI in 2024 more than double the previous year (23%). Some reported productivity gains of up to 25%.
As for the global AI in Procurement market, it is estimated at $3.1 billion in 2025, with a projection of $22.6 billion in 2033 representing 28% annual growth (source: AI in Procurement Market Report, 2025).
These figures are not anecdotal. They signal a structural shift.
What AI Actually Does in Procurement Today
As a Business Analyst and Product Owner on Procurement projects, here is what I observe in the field not in commercial brochures.
Spend Analysis is probably the most mature use case. Tools like Coupa, Ivalua or SAP Ariba can now consolidate, categorize and analyze millions of transaction lines in minutes what used to take an entire team weeks.
Contract management is another fertile ground. AI engines can compare contractual clauses, detect anomalies or legal risks, and generate RFP summaries in seconds. According to Procurement Magazine, 56% of professionals identify source-to-contract as the primary area where AI can be effective.
Supplier risk prediction is emerging rapidly. By crossing internal data (performance history) with external data (news, market fluctuations, geopolitical risks), AI can flag a disruption risk before it materializes.
The automation of transactional processes — purchase orders, invoice reconciliation, supplier onboarding — frees up considerable time. McKinsey estimates that next-generation “agentic” AI systems could make Procurement operations 25 to 40% more efficient.
A striking example: in a pharmaceutical company cited by McKinsey, an AI audit recovered $10 million in uncaptured value in less than a month.
But Here’s What Nobody Really Tells You
The figures are impressive. The use cases are real. And yet, the reality on the ground is far more nuanced.
MIT Sloan Management Review delivered an unsparing assessment: despite $30 to $40 billion invested in generative AI by enterprises, 95% of pilot projects produce no measurable ROI. Only 5% of pilots reach a mature production stage.
Why? Three main reasons consistently emerge:
- The data problem. AI cannot work on incomplete, unstructured or poorly governed data. In procurement, this is often the case. The ProcureAbility 2026 report reveals that 36% of companies cite inadequate data governance policies as the main barrier to AI adoption.
- The adoption problem. An unused AI is a wasted budget. Yet 54% of companies do not collaborate on AI governance between IT and Procurement teams (ProcureAbility, 2026). The result? Tools deployed, but not integrated into real practices.
- The skills problem. 26% of organizations cite a lack of internal skills to manage and analyze data as a limiting factor. Buyers were not trained to be data analysts. And IT departments don’t always have the business knowledge to understand the specific challenges of Procurement.
The Key Role of the Business Analyst and Product Owner
This is precisely where the role of the Business Analyst (BA) and Product Owner (PO) becomes strategic, well beyond simply writing user stories.
In an AI integration project in Procurement, the BA/PO is the bridge between three worlds that rarely communicate well: the business teams (buyers), the data and IT teams, and the solution vendors.
In practice, this means:
- Translating business needs into concrete use cases. “We want AI in procurement” is not a need. “I want to be automatically alerted when a supplier exceeds 15% of late deliveries over a rolling 30-day period” , that’s a use case.
- Identifying and prioritizing quick wins. There’s no need to transform the entire procurement cycle in 6 months. Start with a high-impact, measurable use case that teams are ready for. Confidence is built through results.
- Ensuring upstream data quality. An AI project that fails on data is often a project that didn’t have a BA involved early enough in the chain. Data flow mapping, silo detection, governance rule definition; this is functional analysis work before it’s technical work.
- Supporting change management. Adoption cannot be mandated. It must be built. Co-creation workshops, training, feedback loops, internal communication, the BA/PO who succeeds in AI projects is also a change management actor.
Defining success criteria. KPIs, adoption metrics, processing times, anomaly detection rates — without measurement, there’s no steering. And without steering, there’s no confidence in the solution.
AI Doesn’t Replace Buyers. It Redefines Their Value.
Gartner put it clearly: 65% of Procurement leaders are betting on AI to improve their productivity and decision-making. And 60% of leaders believe generative AI will create new roles and redefine existing ones (Ivalua Future of Work in Procurement Survey).
This is not a façade narrative. It’s a reality being built right now, before our eyes.
Repetitive, transactional, low-value tasks ; AI handles them. Better, faster, without fatigue.
What AI cannot do? Build a trust relationship with a strategic supplier. Negotiate in a geopolitically tense context. Understand the implicit dynamics of an organization. Exercise ethical judgment. Carry a vision.
These are the skills that tomorrow’s buyers will need to cultivate, alongside a growing mastery of data and AI tools.
What I Take Away from My Field Projects
After several years accompanying Procurement teams in their digital transformation, here are my convictions:
- Useful AI in Procurement is simple, explainable and business-oriented. Not a black box. Not a tool that impresses in demos but confounds in production.
- Technology is the last problem. The first problem is data. The second is organization. The third is people. Technology comes after.
- Projects that succeed have a strong business sponsor and a BA/PO involved end-to-end. Not an IT department delivering a turnkey tool, but a cross-functional team co-building a solution.
- The real ROI of AI in Procurement is the time returned to buyers for what they do best. Negotiate. Decide. Create value.
In Conclusion: A Window of Opportunity Not to Be Missed
The AI in Procurement market is structuring itself. Tools are maturing. Use cases are proving out. Organizations (that invest now intelligently, not massively) are building a lasting competitive advantage.
But the real differentiator won’t be the tool chosen. It will be the ability to integrate it, adopt it, and make it a decision-making lever; not just another gadget in an already complex stack.
For the Business Analysts, Product Owners and Procurement professionals reading this article: the coming years will be decisive. Not for those who undergo the transformation. For those who drive it.
Sources
- ProcureCon CPO Report 2025 — Annual survey of Chief Procurement Officers
- AI at Wharton (2024) — Generative AI Adoption in the Workplace
- The Hackett Group — Embracing the Future: How Gen AI Is Revolutionizing Procurement (2025)
- McKinsey & Company — Transforming Procurement Functions for an AI-Driven World (2025)
- MIT Sloan / MIT Research — State of Enterprise AI Adoption (2025)
- ProcureAbility / ProcureCon — CPO-CIO Report 2026
- Gartner — Supply Chain Practice, AI in Procurement (2025)
- Ivalua — Future of Work in Procurement Survey
- Procurement Magazine — AI Impact Report 2025
- Market Research — AI in Procurement Market Projections 2025–2033
You might also be interested in this article:
“Google’s Universal Commerce Protocol: The Foundation for AI-Driven Shopping”
Feel free to share this article on your social networks — this topic deserves serious discussion within your professional circles. If you have any questions or would like to discuss a project, contact us by email; we would be delighted to hear from you.
Follow our Pinterest page for visual inspiration and to discover new ideas around digital transformation and Product Management.
