At Shared Services & Outsourcing Week (SSOW) in San Diego, leaders gathered to examine what it really takes to achieve measurable results with digital transformation. Across Global Business Services (GBS), investment in automation – and more recently, agentic AI – has surged as organizations pursue greater productivity, speed, and scale. Yet many teams are still struggling to achieve the efficiency gains AI promises.
The obstacle is not the technology but attempting to automate fragmented or poorly understood processes. As Shillpa Pol, VP, Finance Business Shared Services at Schindler Elevator Corporation, warned: "If you apply AI to a broken process, you are just going to create more trouble."
So, before organizations scale AI across GBS, which process foundations must they put in place? Here's why enabling a successful GBS transformation can mean going back to basics.
The AI-Readiness Gap is a Visibility Gap
Many organizations have invested heavily in automation and generative AI, only to find that technology cannot overcome weak operational foundations, including:
- Fragmented processes
- Inconsistent or poor-quality data
- Unclear process ownership
- Limited process visibility
Visibility into processes remains a challenge, even for relatively mature organizations. Research presented at the event by ScottMadden found that 30% to 50% of process work takes place outside formal systems, making it harder to understand costs, identify bottlenecks, and pinpoint where automation can create meaningful value.
Without a complete picture of how work is performed, AI is more likely to amplify inefficiency than eliminate it.
As Sameer Andi, VP, People Digital Services & Analytics at Cushman & Wakefield, explained: "AI readiness is less about a tool and more about trust." Building that trust begins with understanding:
- How work flows through the organization
- Who is involved at each stage
- Where decisions are made
- Which activities create value
Only with that visibility can leaders decide where AI will add value – and where the workflow must be redesigned first.
Why Process Standardization is Back in Focus
Process standardization is often treated as an administrative task to complete during an ERP implementation, before attention shifts to more strategic priorities. However, in an AI-enabled operating model, standardization is key to enabling scale.
Throughout the week, transformation leaders repeatedly stressed the need to simplify and standardize processes before introducing new technology. The process should inform the technology, not the other way around.
Standardizing P2P Before Intelligent Invoice Capture
In a session on procure-to-pay (P2P) transformation, Bo Senu-Hodo, Global Source-to-Pay Transformation Leader at Astellas Pharma, and Brian Nichols, Director, Global Shared Services at Sazerac Company, showed why process discipline must come before technology deployment.
Before launching intelligent invoice capture or AI-enabled workflows, Senu-Hodo advised organizations to reduce process variation by:
- Consolidating supplier communications
- Standardizing invoice coding
- Increasing purchase order adoption
- Centralizing transactional work
The principle extends well beyond procurement: reduce variation before introducing automation. Without that foundation, technology simply inherits the complexity that human teams have spent years navigating manually.
The Hidden Cost of Skipping Process Foundations
The pressure to move quickly is universal. Competitive demands, executive expectations, and a fear of falling behind on AI opportunities have all compressed transformation timelines.
But speed without structure creates significant risk. As Matt Lorey, Director, Record-to-Report, and Nichole Franklin, HR Business Partner, both at CRH, noted, organizations can fall into a "stabilization gap."
In this cycle, organizations repeatedly treat symptoms instead of resolving root causes. Teams then build workarounds for persistent problems until those temporary fixes become embedded in the operating model.
The cost is substantial. Employees spend valuable time validating outputs, correcting errors, and managing work that should have been eliminated earlier in the transformation. Layering AI onto this environment just increases complexity further.
The result may be a flurry of implementation activity without measurable business outcomes.
Change Management is the Missing Piece
Throughout the conference, speakers discussed to the role of human behavior in determining transformation success. As Delara McManus, Director, IT Finance at Keurig Dr Pepper, noted: "This was human behavior... how do we make them feel safe?"
Many organizations still approach change management as a supporting activity when it should be a core workstream. The most successful transformations treat communication, adoption, and stakeholder alignment as strategic priorities.
This is particularly important in the age of AI. Employees are less concerned about the technology itself than they are about what it means for their roles and day-to-day processes. If teams are not incentivized to fix inefficient ways of working, they may have little reason to support tools that expose those weaknesses. In some cases, they may even be relieved when AI fails to take hold.
Without transparency about why processes are changing, how employees can help improve them, and what success means for their work, resistance can emerge even when solutions deliver measurable improvements.
As transformation veteran Tom Sheahan, Director, Cash and AR Management at Google reminded attendees, borrowing from Simon Sinek: "People don't buy what you do; they buy why you do it."
Process redesign and technology implementation must be accompanied by deliberate efforts to build trust, engagement, and understanding.
Final Thoughts: A Practical Formula for AI-Enabled Transformation
Across the discussions at SSOW, a practical formula for AI-enabled transformation emerged in four simple actions:
- Understand and simplify the process.
- Standardize and strengthen the data.
- Build ownership and trust among stakeholders.
- Only then apply technology where it can create measurable value.