Reshaping operations for a new era of agentic AI transformation: 4 leadership priorities for COOs

Priority No. 1: Aligning implementations with business outcomes

COOs should start by collaborating with C-suite leaders to define the business outcomes the organization would like to achieve. Once those outcomes are clear and clearly communicated, they can then determine how agents can be leveraged to support this change.⁵ “You can spend a lot of money on agentic AI without achieving real, valuable outcomes,” says Sameer Shetty, group head, digital business and transformation and strategic programs at Axis Bank, one of India’s largest private sector banks. “You need the end-to-end view, not point solutions: What’s going to deliver real ROI? And how is your organization going to evolve to enable it?”

In an agentic AI-enabled model, the focus should shift to value realization: identifying where agents can improve business outcomes, measuring their impact, and enabling employees to work effectively alongside AI agents.⁶ This requires a change in mindset not just from frontline employees, but from COOs as well.

Priority No. 2: Orchestrating shifts in work and human-agent responsibilities

Unlike many previous technology-driven implementations that focused on specific functions or processes, agentic AI can deliver the most value when organizations rework end-to-end workflows, not simply automate existing tasks.

Deploying AI agents into existing roles, workflows, and processes without redesigning the work itself is unlikely to yield the benefits COOs and their organizations are seeking.⁷ The transition requires a shift from fixed task completion to fluid human-agent collaboration.⁸

To unlock value across the organization, COOs should approach process transformation as a redesign challenge rather than an optimization exercise. This involves rebuilding operating models so that humans and agents can each play to their strengths, with clear accountability measures built into every step.

Getting that right requires a holistic view of the organization. Alessio Marras, cofounder and head of organization for AideXa, an Italian digital bank serving small- and medium-sized businesses, says: “We’re developing a broader strategy. It’s based on a clear vision for where and how we want to use AI agents, then implementing them progressively. The challenge is coordination–moving from one use case to multiple orchestrated agents requires an end-to-end view. Agents that are very efficient in single use cases can become much more difficult to coordinate and govern when they operate as part of an orchestrated system. That’s why we believe in continuously monitoring outcomes and being ready to adjust the balance between automation and human oversight as processes evolve.”

AI can help reduce the time employees spend on execution, analysis, and routine tasks, allowing them to focus more on strategic and value-creating priorities and responsibilities. But in the near term, organizations may face challenges in developing the internal agentic AI expertise required for scaling.⁹ Employees who have worked alongside external teams may need to be reskilled to operate effectively in an agent-based environment. That transition will require COOs and other senior leaders to carefully manage accountability, risk, and performance. They will need to establish guardrails that account for multi-agent systems and the dynamic relationship between humans and machines.¹⁰

As agent networks grow, the complexity of coordinating them—across business units, systems, and decision points—could increase rapidly. COOs may need to manage that complexity by defining how work should flow between humans and agents; where escalation points should be placed; how to monitor performance across the full system rather than within individual use cases; and how to maintain visibility and control as the network scales. Getting these orchestration layers right is what may likely set organizations that build coherent, enterprisewide operational capabilities apart from those that are only able to implement a collection of AI experiments.

Priority No. 3: Leveraging flexibility into workflows and processes

A structural shift is also emerging in how organizational leaders think about capability, ownership, and delivery.¹¹ For example, over the years, many COOs have guided a deliberate push toward external service providers—engaging best-in-class collaborators so their organizations could stay focused on the core business. But some COOs expect agentic AI to push the pendulum back, from externalization to internalization.

This shift in ownership may reflect a broader transformation in how work is structured and delivered. Beyond organizational and skills changes, companies should rethink delivery models and adopt more cross-functional co-creation approaches.

One of agentic AI’s most important—and underrated—characteristics is its modularity. Unlike large-scale technology investments that require a significant upfront commitment, agentic AI can start small and be scaled incrementally. COOs can take advantage of this flexibility by building systems that can be continuously reconfigured. They can, for example, add agents or adjust workflows, adapting how the technology is used as conditions evolve and the organization learns.

Priority No. 4: Tracking value delivery against real success criteria

Many organizations may successfully pilot an agent, but few can deploy hundreds of agents across operations with confidence. This requires a much stronger process foundation than traditional automation.

Many enterprise performance systems measure what employees do—tasks completed, volume delivered. As a result, organizations have optimized around activity rather than outcomes. To leverage agents at scale, COOs need to prioritize deployments that align with enterprise strategy and track value delivery against success criteria.¹² As agentic AI takes on more tasks, employees should have more time to focus on ensuring the right business outcomes are being achieved.

Axis Bank’s Shetty draws a clear distinction between traditional AI and its agentic successor. “Classic AI is still relatively contained,” he says. “You enter data, the model makes a recommendation, and frontline users don’t need to change much in terms of their processes. But with agentic AI, the goal is to have even more users pursue open-ended goals, asking even more questions. It has a broader focus, [it’s] less constrained, and can have a much bigger impact on outcomes—all of which requires being open, ready, and able to change processes.”

It’s time to make agentic aspirations happen

Agentic AI could fundamentally change how organizations work. Organizations that are pulling ahead aren't waiting for perfect clarity or complete buy-in. They’re making decisions now using the information they have and building the internal capacity to make course corrections as they go.

For COOs, the question that remains is whether they’re shaping that change—or inheriting whatever shape it takes. That’s the real pressure COOs are feeling.

But COOs, by virtue of their role, are positioned and have the skills to play a key role in broad-scale, end-to-end transformations. Few other executives have the cross-functional perspective, the process authority, and the organizational relationships required to make these goals a reality across the enterprise.

Ultimately, how well an organization implements agentic AI isn’t likely to center on how many agents are deployed or the number of efficiencies they yield, but on how well these implementations can deliver sustainable advantages and enduring business value.¹³