In procurement teams, generative AI and automation tools are now part of the daily routine: supplier emails are summarized, specification drafts are prepared quickly, and quotes are compared in less time. Even so, many managers keep asking the same question: "How do we automate procurement processes, and how should roles be redesigned?" The reason is that individual employees getting faster does not always show up in the department's overall results.
A survey Gartner published on May 19, 2026 puts a number on this gap. According to the study, conducted in January–February 2026 with 101 chief procurement officers (CPOs), only 36% of CPOs are very confident in their ability to redesign roles and processes around artificial intelligence. The same survey shows that individual productivity gains have not yet translated into results at the team or organization level.
In this article we look at why this gap exists, a task-based approach to automation in procurement processes (requisition, ordering, invoice matching, supplier management), and the points to watch in role design.
The productivity paradox: the gain stays with the individual
According to Gartner, current uses of generative AI in procurement raise individual productivity in areas such as time savings, volume of work produced and quality; however, these results lag significantly at the level of team output. Gartner calls this disconnect the "AI productivity paradox."
Gartner's explanation is simple: procurement operating models have not changed at the speed needed to reflect generative AI's impact on existing roles. When part of a role is automated, the employee can become more productive; but unless the work is redesigned across the function, the organization does not capture the same benefit on its own.
In practice it looks like this: a specialist finishes a quote comparison in half the time, but because the approval chain, the order-creation step and the invoice check stay as they were, the time it takes to fulfill the requisition does not change. The time saved is lost waiting at another point in the flow.
The three changes Gartner recommends
Gartner makes three recommendations to procurement leaders for getting more return from their AI investments:
- Redesign roles around AI: assess existing roles, identify common tasks, and separate the work into tasks that are inherently human and tasks suited to AI.
- Tie gains to financial outcomes: rather than relying only on efficiency metrics, focus on cost optimization and revenue growth.
- Update productivity metrics: broaden performance measurement to also cover the innovation, complexity and new outputs that emerge from AI-enabled work.
Gartner also stresses that traditional measures focused on output per unit of time are increasingly out of step with an AI-enabled procurement function. In other words, asking "how many orders per person" on its own does not give the right picture.
The distribution of work is being redrawn
Gartner's August 2026 assessment adds another dimension to the topic. As AI assistants and autonomous agents become part of daily operations, procurement leaders are taking on a larger role in determining how capabilities are distributed among employees, suppliers and digital workers. Gartner's recommendations also include measuring AI investments by their business outcomes and by their impact on the workforce.
This shows that role design is not just a matter of "which work will the bot do?" The real question is clarifying which decision belongs to whom within the same process: the human, the software robot or the supplier.
The picture in Türkiye
According to the October 2, 2026 bulletin of TÜİK (the Turkish Statistical Institute), the share of enterprises in Türkiye using at least one AI technology is 14.0% in 2026; among enterprises with 250 or more employees it is 37.1%. The barrier most often cited by enterprises considering using AI, at 72.3%, is a lack of relevant expertise within the organization. This picture underscores why preparing the team for the new division of work matters before choosing tools.
A task-based approach to procurement processes
The breakdown below is a general example of applying Gartner's "human task / AI-suited task" recommendation to the procurement flow; each organization needs to adapt it to its own authority structure and regulatory obligations.
- Requisition: Completing the requisition form in full, checking the budget code and routing to the right approver are rule-based tasks and suit automation. Deciding whether the need is genuinely necessary, and which alternative solutions to pursue, should remain with people.
- Ordering: Creating an order from an approved requisition, checking it against framework agreement terms and sending it to the supplier can be automated. Off-contract purchases and exceptional terms should be routed to human approval.
- Invoice matching: Comparing order, goods-receipt and invoice data is repetitive, high-volume work. Software moves matching records forward; investigating the cause of variances outside tolerance and resolving them with the supplier is human work. For automation on the payment and accounting side, see our article on automation in finance processes.
- Supplier management: Collecting and monitoring data such as delivery times, quality notifications and contract dates suits automation. Negotiating with suppliers, managing relationships and assessing risk, on the other hand, require judgment.
Getting started: a practical checklist
- Build a task inventory. List the tasks in procurement roles one by one; mark which are rule-based and repetitive and which require judgment.
- Measure the whole flow. Alongside individual processing time, also record the total time from requisition to order and from order to payment. The bottleneck is more often inside the hand-off points than with the individual.
- Redesign the hand-off points. So that the time gained through automation is not lost in the approval queue, review approval thresholds, exception rules and owners together.
- Define human approval points in writing. For steps such as out-of-tolerance invoice variances, off-contract purchases and new supplier registration, make clear who holds the decision.
- Update roles. For employees whose routine work is shrinking, add more complex tasks such as negotiation, supplier risk and category analysis to their job descriptions.
- Tie metrics to financial outcomes. Alongside the number of transactions per person, track cost impact, exception rate and process time.
Conclusion
Automation tools in procurement raise individual productivity; Gartner's data shows that for this gain to turn into a team result, roles and processes need to be redesigned deliberately. Separating the requisition, ordering, invoice-matching and supplier-management steps at the task level, reorganizing the hand-off points and tying metrics to financial outcomes are the cornerstones of this transformation. Clarifying the division of work before deciding on technology makes it easier for the investment to pay off at the department level.
Why logistics experience matters in procurement automation. Procurement does not end at the purchase order. Goods receipts are recorded where the goods arrive, and delivery times and quality notifications typically originate in receiving and inbound operations, so invoice matching and supplier management rest on data created by the physical flow of goods. Tracing a single order from requisition through goods receipt to the paid invoice is therefore a practical way to find the hand-off points described above. One of our consultants brings more than 30 years in operations, logistics and supply chain: at a technology distribution company of roughly one billion dollars in scale, he ran logistics, warehousing, technical service and administrative operations end-to-end, building and leading teams of 100+ people and delivering lasting gains through KPI-driven programs. He has also managed nationwide warehouse and transport operations. An industrial engineer, he now brings that supply-chain discipline to SMEs through AI and automation. Meet our consultants.
#interimmanagement #forwarddeployedengineering