Ask any CPO what their 2025 roadmap looks like, and “AI” is almost guaranteed to come up. Ask the same CPO how many of their AI pilots actually made it to full production, and the answer is usually a lot less impressive.
The gap between AI ambition and AI results isn’t unique to any one company, it shows up across industries, geographies, and technology budgets. So, we went looking for a pattern. In a recent podcast series, we spoke with procurement leaders from Micron, Novartis, Coca-Cola, Walmart, and other global enterprises to understand what actually separates the teams that see results from the teams still stuck in pilot purgatory.
Their answers converged on five levers, not a checklist, but a sequence. Get this right, and AI becomes a genuine force multiplier. Skip them, and it’s just another expensive tool nobody trusts.
The big takeaway: AI isn’t the silver bullet. It’s a force multiplier, but only when culture, data, and planning are solid.
Culture before code
“Don’t just buy transformation by buying software. Fix your culture first.”
— Zyad Khan, Associate Director of Procurement & Contracts at Dubai World Trade Centre, guiding Fortune 500 enterprises on procurement transformation.
Zyad frames it as Adaptability Quotient (AQ), the measure of how ready your people are to adopt new tools and processes. In AI projects, AQ beats technical readiness every time. Why? Because even the best-designed workflows fail if teams cling to old methods.
In practice:
- Audit your AQ: Do managers actively encourage experimentation? Do teams have permission to retire old processes?
- Sequence your rollout: Start with one module (e.g., eAuctions or onboarding), track adoption weekly, and build internal case studies before expanding.
- Build visible wins: Celebrate early successes to shift mindsets from “threat” to “tool.”
⚡Putting it into action: Start with a single module in your highest-friction area, track adoption in the platform’s built-in analytics, and expand once early wins are visible.
Data discipline or bust
“One big supplier named four different ways hides 10% of your spend.”
— Karthik Rama, Global Procurement Consultant known as “The Procurement Doctor,” specialising in supplier data governance and ERP transitions.
Karthik calls dirty data the “silent killer” of procurement ROI. Duplicate supplier records inflate vendor counts, dilute leverage, and distort spend analysis. If you skip the clean-up, AI will just automate bad decisions faster.
In practice:
- Golden supplier record: Legal name, tax IDs, parent-child hierarchy, bank details, ESG attributes — standardized.
- Fuzzy-match de-duplication: Catch “ABC Ltd” vs. “A.B.C. Limited” vs. “ABC Corporation.”
- Governance: Assign data stewards and make data health a tracked KPI.
⚡Putting it into action: Use AI cleansing to unify supplier data across ERPs, then lock in quality with validation rules at PR/PO stage to prevent new duplicates.
Break the firefighting cycle
“Procurement becomes a yes-department. Planning breaks the cycle.”
— Vera Rozanova, Procurement transformation leader with 15+ years at Coca-Cola HBC, Reckitt, and Philip Morris.
Vera’s fix is to embed procurement into the budget and S&OP cycles, not just turn up when a requisition lands. This shifts teams from tactical responders to strategic partners.
In practice:
- Publish a sourcing calendar tied to budget approvals.
- Run monthly pipeline reviews with finance and category leads.
- Enforce an “intake with business case” policy — exceptions only for genuine emergencies.
⚡Putting it into action: Use the intake assistant to capture requirements early, connect them to budget data, and automatically feed them into your sourcing pipeline so you’re planning ahead instead of chasing.
AI that works now
“Supplier onboarding and PO approvals are ripe for automation.”
— Megha Singh, Director – Procurement Transformations, Micron Technology
At Micron, Megha’s team didn’t start with grand AI plans — they began with high-friction workflows that drained time but carried low change risk. AI auto-drafted RFPs from historical templates, freeing her team to focus on refining requirements. In PO approvals, policy overrides were flagged before they reached her desk, dramatically reducing back-and-forth with requestors.
In practice:
- Identify bottlenecks with high manual effort and low change risk.
- Pilot AI in one upstream (e.g., onboarding) and one downstream (e.g., approvals) process.
- Keep a human-in-the-loop for the first 60 days to refine rules and build trust.
⚡Putting it into action: Deploy smart onboarding and PO approval modules in one pilot category, then measure impact on cycle times and compliance before scaling.
Modular procurement is the way ahead
“When teams start managing processes outside your platform via spreadsheets, emails, and side systems, you don’t have control. You have chaos dressed as standardization.”
— Pratik Thakore, COO at Powerweave
Big-bang deployments are expensive, risky, and slow to show results. In contrast, modular procurement, rolling out capabilities in phases aligned to business readiness, accelerates ROI and improves adoption. It also lets you pivot faster when priorities shift, whether due to market disruption, ESG pressures, or supply chain shocks.
In practice:
- Identify high-impact, low-dependency modules to deploy first.
- Use early rollouts to build proof points and executive confidence.
- Keep integration pathways open so modules can be scaled or swapped without rework.
⚡Putting it into action: Ewiz procure’s modular architecture lets you start with the workflows that matter most, from supplier onboarding to contract analytics, and expand without disrupting existing ERP or BI systems.
🎧 Listen to the podcast series → Beyond Procurement Podcast Series
📥 Download AI in Procurement Guide for enterprise-tested strategies and benchmarks
AI in Procurement: The strategic playbook for enterprise leaders
Beyond Procurement by Powerweave
We help enterprises rethink procurement and sustainability with purpose-built modular solutions like ewiz procure and Snowkap. Want to explore what this could mean for your team or where to start?
No pitch decks, just a real conversation.
