Sponsored by
The Work Split: Redesigning One Role Before Anyone Asks You To
CPO Path #30 CW 42 2026
Hi {{first_name}}
This issue opens The Redesign, a six-part series on leading procurement’s AI transition from the director’s chair: the work split, the proof finance accepts, the data minimum, agent authority, the new scorecard and the team after. Each part delivers one instrument you can apply to your own area.
Enjoy
EXECUTIVE PREMISE
Procurement’s AI productivity challenge is becoming an organisational design question.
Gartner’s 2026 survey of 101 CPOs found that individual productivity gains from generative AI decline significantly at team level. Only 36% were very confident in their ability to redesign the procurement function for AI (Gartner, 2026).
A procurement director need not wait for a function-wide mandate to address this. One thoughtfully redesigned role, supported by measurable outcomes, can establish a credible starting point.
WHY THIS MATTERS
The resource equation is tightening. The Hackett Group projects an 8% increase in procurement workload during 2026, against a 0.9% reduction in staffing and a 0.4% reduction in operating budgets (The Hackett Group, 2026).
The organisational question is therefore not simply whether AI improves individual productivity. It is who converts those improvements into additional capacity, better decisions and business outcomes.
A buyer saving several hours preparing quotations may experience a lighter administrative burden. Yet the wider organisation may see no difference if that capacity disappears into other routine activities.
For an ambitious procurement director, the gap creates a leadership opportunity: demonstrating an ability to redesign work, rather than merely introducing technology.
THE REAL DYNAMIC
The evidence cautions against treating an entire job as equally suitable for AI.
Dell’Acqua and colleagues studied 758 consultants and found significant performance improvements on tasks within AI’s capability frontier. On a task selected to lie outside that frontier, AI-assisted participants were 19 percentage points less likely to reach the correct solution (Dell’Acqua et al., 2026).
Brynjolfsson, Li and Raymond studied 5,172 customer-support agents and found average productivity improvements of 15%, with substantially greater benefits among less-experienced workers (Brynjolfsson et al., 2025).
Neither study measures strategic procurement directly. Together, however, they support a useful distinction: AI deployment is better evaluated task by task than by assuming uniform benefits across a role.
Consider a strategic buyer responsible for electronic components. The role combines commercial negotiations, supplier risk assessment, RFQ preparation, quotation comparison, market intelligence and administrative coordination. These activities differ in their dependence on judgment, information quality, repeatability and commercial accountability.
A broad instruction to “use AI more” leaves the operating model largely unchanged.
A more useful approach separates the work into three categories: human-owned decisions, AI-assisted analysis and controlled AI-native execution.
The distinction also changes the internal conversation. Rather than presenting technology as a headcount argument, the director can show how existing resources might address commercial priorities that have so far been neglected.
WHAT STRONG LEADERS DO DIFFERENTLY
They start with one role, not an organisation chart.
The following illustrative task map allocates a strategic buyer’s working week across 13 activities. Percentages represent assumed current effort, not measured savings.
The Work Split: Strategic Buyer, Electronics
Task | Time | Classification |
|---|---|---|
Supplier negotiations | 15% | Human |
Commercial risk decisions | 10% | Human |
Stakeholder alignment | 10% | Human |
Supplier development | 10% | Human |
RFQ preparation | 10% | AI-assisted |
Quotation comparison | 10% | AI-assisted |
Market intelligence | 8% | AI-assisted |
Contract review preparation | 7% | AI-assisted |
Supplier performance analysis | 5% | AI-assisted |
Cost-driver analysis | 5% | AI-assisted |
Reporting compilation | 4% | AI-native* |
Data classification | 3% | AI-native* |
Routine status updates | 3% | AI-native* |
Total | 100% |
*AI-native describes a candidate for automation after validation, with defined exception handling and human oversight.
Three design criteria determine the split.
Commercial accountability: negotiations, supplier commitments and consequential risk decisions remain human-owned.
Analytical augmentation: AI prepares comparisons, identifies anomalies and structures information; the buyer validates assumptions and conclusions.
Controlled execution: repeatable administrative activities become automation candidates where data integrity, permissions and exception rules permit.
The resulting capacity requires separate measurement.
Assume a 40-hour working week. AI-assisted and AI-native activities currently represent 22 hours. If a controlled trial reduces their combined effort by 25%, gross capacity released is 5.5 hours. After 1.5 hours of additional validation and exception handling, net capacity is four hours. These are illustrative assumptions, not research findings.
Redeployment commitment: the four verified hours go to supplier capacity-risk reviews and to negotiations on expiring component quotations.
The director can now discuss a specific allocation of capacity, and its intended commercial contribution, with the Head of Supply Chain or the CPO. That is a materially different management proposition from reporting how many hours an AI tool has saved.
CPO PATH DIAGNOSTIC
Statement | Fully true | Partly true | Not true |
|---|---|---|---|
We know where one buyer’s working week goes. | ☐ | ☐ | ☐ |
We distinguish judgment from preparation and execution. | ☐ | ☐ | ☐ |
AI boundaries and human decision rights are defined. | ☐ | ☐ | ☐ |
We measure net capacity after verification effort. | ☐ | ☐ | ☐ |
Recovered capacity is assigned to specific commercial work. | ☐ | ☐ | ☐ |
ONE-LINE VERDICT
The executive contribution begins when productivity gains become deliberate choices about where procurement directs its capacity.
YOUR TURN
Which role in your team would you split first? Reply with one line, I read every answer.
QUICK WINS
This week: ask one buyer to list their recurring tasks and estimate the share of time each takes. Thirty minutes is enough for a first version.
Next week: classify each task as human, AI-assisted or AI-native, and write one sentence on why for each.
Before any pilot: name the commercial work the released hours will go to, and agree that allocation with your manager in advance.
{{first_name}} - Did this episode add value to you?
I read every single message, and past posts already shaped this space. Reach out to me and tell me what you’re up to and the path you’ve chosen or post online your thoughts about this episode. I read every email and linked post.
Talk soon,
Pascal
Did this resonate with you?
Tools I use to improve efficiency and ensure stable processes:
Unify Your Teams and Tech Stack With HubSpot
Connect your customer data, teams, and tools without the hassle of complex integrations or lengthy setups. One easy platform gives marketing, sales, and service teams a unified customer view and the tools to turn it into growth.
Why HubSpot and what's new
• Generate leads and automate marketing with Marketing Hub
• Build your pipeline and close more deals with Sales Hub
• Scale customer support and drive retention with Service Hub
• Keep customer data clean, connected, and actionable with one, unified platform
Join 306,000+ in over 135 countries using HubSpot to grow their businesses.
See what a more connected approach to growth can do for you and your team. Get setup quickly and start checking off your hardest tasks.
SOURCES
Brynjolfsson, E., Li, D., & Raymond, L. (2025). Generative AI at work. The Quarterly Journal of Economics, 140(2), 889–942. https://doi.org/10.1093/qje/qjae044
Dell’Acqua, F., McFowland, E., III, Mollick, E., Lifshitz-Assaf, H., Kellogg, K. C., Rajendran, S., Krayer, L., Candelon, F., & Lakhani, K. R. (2026). Navigating the jagged technological frontier: Field experimental evidence of the effects of artificial intelligence on knowledge worker productivity and quality. Organization Science. Advance online publication. https://doi.org/10.1287/orsc.2025.21838
Gartner. (2026, May 19). Gartner survey shows just 36% of chief procurement officers are very confident in ability to redesign function for AI [Press release]. https://www.gartner.com/en/newsroom/press-releases/2026-05-19-gartner-survey-shows-just-36-perecent-of-chief-procurement-officers-are-very-confident-in-ability-to-redesign-function-for-ai
The Hackett Group. (2026). 2026 procurement agenda and key issues study. https://www.jaggaer.com/wp-content/uploads/Hackett-2026-Procurement-Agenda-and-Key-Issues-Study-Results-JAGGAER.pdf
Thank you for your time and trust,
Pascal Hecker | Editor-In-Chief, CPO Path.