%20(1).png)
I recently read an interesting article by Marijn Overvest, founder of Procurement Tactics. The statistics in the report describe a procurement function in genuine acceleration and a real opportunity for the teams that move deliberately.
$9.5B projected for procurement software by 2028. 74% of organizations increasing AI budgets this year. 90% moving toward AI agents. 70% of intake expected to be AI-assisted by 2027. Underneath all of it: 58% of teams carrying more workload, with only 13% seeing headcount growth to match.
Read together, these figures don't describe a technology trend. They describe a structural shift in what procurement is expected to deliver, and why AI has moved from interesting to necessary for the teams that want to stay ahead of it. Manual procurement and manual processes are increasingly seen as inefficient, error-prone, and unsustainable, while automated procurement and procurement process automation offer significant benefits - reducing manual errors, increasing efficiency, and enabling organizations to scale procurement activities without proportional increases in headcount.
The investment figures are the surface story. The workload figures are what's driving them.
Procurement is being asked to do more, across more categories and supplier relationships, with more strategic weight, without a corresponding increase in team size. That's not a temporary pressure. It's the new shape of the function. And the 90% move toward AI agents is the clearest signal of how organizations are responding. Automated procurement products are enabling organizations to automate and streamline processes - from handling purchase requests and approval processes to invoice management and payment -helping teams manage higher transaction volumes and complex supply chains efficiently.
The shift toward agents isn't a continuation of the copilot era, where teams used AI to move faster on tasks they were already doing. It's a structural realignment. The question is no longer "how does AI help my team work?" It's "how does my team manage AI that works?" That's a different operating model, and the organizations building toward it now will have a meaningful lead when it becomes standard.
The 70% intake figure makes this concrete. Gartner's projection that 70% of purchase requisitions will be AI-assisted by 2027 isn't describing a future state. It's describing what leading teams are already building toward. Intake is the front door: high frequency, partially structured, deeply manual in most organizations. When AI runs that process with human oversight rather than assisting a human through it, the operating leverage of the entire procurement function changes.
That's the shift the budget figures are chasing. The teams that get there first will be running a structurally different operation than the ones still building toward it.
The 90% figure covers a wide range of maturity. Some organizations are running early pilots. Others have AI operating across multiple workflows at scale. The gap between them isn't primarily budget. It's how they think about what AI is for.
The teams pulling ahead have made a specific shift: AI isn't a tool the team uses. It's capacity the team manages. That reframe changes which workflows they start with, how they measure success, and what they expect from their people.
They start at the execution layer. Intake processing, supplier follow-up, PO exception handling, not spend dashboards or risk reports. These aren't the most technically sophisticated use cases. They're the right ones because they change what the team has to do today, not just what leadership can see.
They redesign the workflow alongside the deployment. AI performs best when the process it's running is well-defined. For intake, that means routing logic, completion criteria, and policy guardrails. For supplier coordination, it means clean data, clear ownership, and defined escalation paths. The teams moving fastest treat this design work as part of the deployment, not a prerequisite that precedes it.
They bring structure to supplier communication without changing how suppliers work. Fragmented supplier interactions are one of the highest-overhead areas in operational procurement: responses across email threads, no single source of truth, significant time spent on manual reconciliation. The teams reducing that overhead fastest aren't forcing suppliers into new portals or workflows. They're building AI that operates natively in the channels suppliers already use, extracting the relevant data, and keeping the procurement team's view consolidated and actionable.
They measure throughput, not activity. Not tools deployed or processes touched, but how much the team can handle and how that changes over time. That's the metric that builds the internal case for further investment and tells you whether the operating model is actually shifting.
Automating repetitive tasks and procurement tasks using AI-driven automation tools allows procurement teams to focus on higher-value business processes, such as strategic sourcing and supply chain optimization, rather than being bogged down by manual processes.
The combination of these statistics points to something specific: a window right now where moving deliberately on AI creates compounding operational advantage.
The market is early enough that execution quality still differentiates. In 12 months, AI-augmented procurement will be table stakes. The teams running it operationally now, managing agents, refining workflows, building institutional knowledge, will have a lead that's hard to close.
The 58% workload figure is the most honest signal in the data. It describes teams already at capacity, already being asked to scale without growing headcount, and already looking for a structural answer. AI is that answer, deployed at the execution layer with the right sequencing, available now.
The teams moving through this window deliberately are the ones building the operational foundation that will define procurement performance for the next several years. The window won't stay open indefinitely.
