10
Loading video...
By Brad DeMent
Partner
0
Share
Newsweek is a Trust Project member
See more of our trusted coverage when you search.Prefer Newsweek on Googleto see more of our trusted coverage when you search.
Global organizations are making a dangerous bet. Companies are restructuring their workforces around AI productivity gains while many have yet to build the operating models needed to deliver them.
The gap between technological ambition and operational readiness is substantial. According to SSON Research & Analytics’ 2026 research, digitalization is a leading Global Business Services (GBS) priority, yet fewer than one in five organizations have automated the majority of their processes.
Legacy technology remains the leading barrier to transformation. The message for executives is uncomfortable: the technology race is advancing faster than the organizational changes required to make it pay off.
Read More on Leadership Insights
I believe this is one of the most consequential management challenges of the AI era. Companies are investing in technology capable of coordinating work across entire business processes while retaining structures built around departmental expertise.
They expect AI to eliminate inefficiencies that their own operating models continue to create. More concerningly, some are beginning to make workforce decisions based on productivity gains that have yet to materialize.
Consider order-to-cash, the process that takes a customer from initial setup to completed payment. Sales takes the order, supply chain fulfills it, and finance issues the invoice and collects the money.
Each department may meet its performance targets, yet the customer can still experience delays, errors, and frustration. Every handoff introduces another opportunity for the process to break down.
This is how an enterprise can have green dashboards across its departments and a failing business process overall. The fundamental problem is that companies are organized around expertise rather than outcomes.
The problem is not a lack of awareness. Companies have appointed Global Process Owners (GPOs) for years, recognizing the need to coordinate work across traditional functional boundaries.
Yet these leaders are frequently expected to influence decisions without possessing the authority to make them.
An adviser can recommend a better process. An accountable owner can deliver one.
A Critical Distinction
SSON Research & Analytics found that 90 percent of organizations with GPOs reported a positive impact on process performance.
That is an encouraging signal, but the same research identifies a critical distinction between having a process owner on the organizational chart and giving that person the authority to change how the work gets done.
A title cannot resolve competing departmental priorities. Decision rights can.
AI makes this structural weakness more consequential. Who determines which decisions an AI agent can make? Who resolves conflicting data standards? Who decides which systems are redundant?
If authority is fragmented, every answer becomes a negotiation. The technology encounters the same organizational bottlenecks that have slowed transformation for years.
That gap becomes dangerous when projected AI efficiencies begin driving workforce decisions. Companies may reduce headcount before the processes and capabilities needed to absorb that work are actually in place.
Employees can lose their jobs before the technology is ready to assume their responsibilities, leaving the remaining workforce to manage the resulting failures and backlogs.
The Consequences
The consequences extend into everyday life. Customers wait longer for services, employees inherit poorly designed workflows, and businesses pay for technology that underdelivers.
Properly implemented AI within an evolved operating model could free people from repetitive administration and give them more time for problem-solving and customer engagement.
Implementing AI without thinking through the operating model could leave organizations understaffed, operationally fragile, and unable to meet the expectations they have created.
There is a practical way forward, and it does not require reorganizing the entire enterprise overnight.
For one global technology company, improving procure-to-pay performance began with understanding where problems started. The GPO analyzed operational data and traced many accounts payable issues back to the purchase order process.
Those findings led the team to prioritize purchase order improvements rather than automate inefficient downstream work or pursue a sweeping AI transformation.
This would have been a challenging task if the GPO was buried in the organization without proper authority to make decisions.
Following these changes, roughly 70 percent of purchase orders required no buyer intervention. The organization then worked to utilize AI to extend touchless processing into accounts payable.
Organizational Design
This is the kind of incremental progress executives should demand.
The company did not need to resolve every enterprise-wide technology problem before improving a specific process. It identified the operational barriers, addressed them, and measured the outcome.
I would apply the same discipline to organizational design. C-level leadership should encourage the selection of one high-volume process, such as purchase-to-pay, and appoint a leader, such as a GPO, accountable for its performance from beginning to end.
That leader needs the authority and breadth of functional experts reporting to them to resolve disputes across procurement, finance, supply chain, and technology, along with measurable targets for cost, cycle time, and reliability.
The company should then introduce AI and other process improvements where they can produce demonstrable improvements, retain human oversight where necessary, and use the results to guide wider adoption.
This also addresses a familiar objection that poor data must be fixed across the entire organization before AI can work. A focused process narrows the problem to the information and systems required for a defined outcome.
The First Move
Companies can improve those foundations incrementally, without waiting for a perfect enterprise-wide data environment.
The responsibility for making this change belongs to C-level leadership.
Boards should demand evidence that processes have been redesigned and productivity gains demonstrated before approving workforce reductions justified by projected AI efficiencies.
Leaders should also examine whether their process owners have actual decision-making authority or merely the responsibility to recommend change.
AI will not automatically make a company more productive. The technology can accelerate work, but only changes to the operating model will deliver real value.
Companies that fail to confront this reality risk paying twice, first for technology that underdelivers, and again for the operational damage caused by acting as if it had succeeded.
The first move is not another deployment. It is giving someone the authority to make the entire process work.
Brad DeMent is a Partner at ScottMadden, which advises organizations on Global Business Services, process design and automation; all opinions are the author’s own.
View 0 Comments
Submit Correction
View Editorial & AI Guidelines