Where Transformation Actually Breaks

Organizations are not failing at transformation because of weak technology.

The Process Excellence Network has published research and expert perspectives across multiple formats this year, and the pattern is consistent. Transformation stalls at the same place, in the same way, for the same reason.

The operations were not documented before the initiative began.

The Numbers Do Not Point to Technology

According to the PEX Report 2025/26, based on a global survey of more than 220 transformation, operations, and business leaders, 56 percent of organizations are deploying enterprise-wide transformation strategies. Fifty-eight percent consider transformation a mission-critical driver of growth.

Yet process inconsistency ranks among the five most common integration challenges slowing those strategies down. The PEX Report finds that many organizations still operate with fragmented and poorly documented workflows, making it difficult to integrate systems effectively or scale improvements across the business.

The investment is real. The intent is present. The foundation is not.

AI Does Not Tolerate What Manual Work Could Hide

For years, businesses operated on tacit knowledge. Workflows existed in conversations and institutional memory. Exceptions were handled by people who knew the workarounds. Inconsistencies were invisible because experienced team members compensated in real time.

AI changes that condition entirely.

As Process Excellence Network contributor Alex Vakulov observes, AI adoption forces a level of process formalization that many organizations had previously avoided. Manual workflows could sustain themselves on undocumented exceptions and inconsistent approvals because people absorbed the friction. Once AI enters those workflows, the same gaps that people quietly managed become operational liabilities. The system has no mechanism to compensate.

This is not a technology limitation. It is an operational exposure. AI did not create the problem. It removed the workaround.

Most Processes Are Not Processes

This distinction matters: a documented workflow and a verbal habit are not the same thing, even when they produce similar outputs.

Doug Stephen, CEO of CGS, addressed this directly at the PEX Network’s All Access: AI in Process Excellence webinar series, where thought leaders from Google, SAP, Camunda, and other organizations gathered to examine why AI initiatives fail to move from pilot to production. The figure he referenced is stark: 95 percent of AI pilots do not reach production.

Stephen’s position is that most organizational processes are not actually processes. They are disconnected workflows performed differently by each team member. There is no standard. There is no defined ownership. There is no documented version that could be handed to a new employee, an AI system, or an auditor.

You cannot automate what has not been defined. You cannot scale what has never been standardized.

The Foundation That Gets Consistently Skipped

The consensus across the PEX webinar series, from practitioners at Google, SAP, and Camunda, is clear: successful AI adoption requires foundational work in data quality, process excellence, and organizational readiness before any technology is deployed.

That sequence matters. Organizations that skip it end up with what Camunda describes as supercharged fragmentation, adding AI point solutions to disconnected systems and creating more complexity, not less.

The PEX Report reinforces the stakes. Less than half of the organizations surveyed currently have an AI governance policy. Without documented workflows defining ownership, decision authority, and accountability, there is nothing for governance to govern.

Process formalization is not a bureaucratic exercise. It is the condition that determines whether transformation has anywhere to land.

What the Operational Foundation Actually Requires

The integration challenges the PEX Report identifies are not independent problems. They share an underlying condition.

Legacy systems become problematic when there is no documented understanding of what the business requires from them. Fragmented data becomes unmanageable when there are no documented standards governing how data is created, owned, and used. Process inconsistency persists when workflows live in people’s heads instead of maintained documentation. Misaligned tools fail to connect because the operations they are supposed to support were never defined to a shared standard.

Transformation does not break at technology. It breaks at the absence of documented operational infrastructure that technology could build on.

Bottom Line

The transformation gap is a documentation gap.

This is the argument three separate bodies of reporting from the Process Excellence Network make in alignment. Organizations are investing in transformation at scale while operating on workflows that were never written down, standards that were never formalized, and processes that vary by the person performing them.

AI adoption does not fix that condition. It exposes it.

Businesses that want transformation to deliver on its investment need to document their operations before they automate them. Not as a compliance step. As the structural prerequisite that makes any of the rest of it possible.

Expert Tip

Before evaluating any AI tool or transformation initiative, run this test on the workflow you are considering.

Can you provide a description of that workflow to someone unfamiliar with your business and have them execute it correctly without asking a single question?

If the answer is no, incorporating additional tools or automation may create a new bottleneck.

Sources: Amelia Brand, “5 Integration Challenges Slowing Transformation and How to Overcome Them,” Process Excellence Network, March 20, 2026. Alex Vakulov, “How AI Is Changing Decision Authority in Data-Driven Organizations,” Process Excellence Network, March 27, 2026. Rose Morishita, “AI in Process Excellence: Laying the Foundation for Enterprise AI Success,” Process Excellence Network, March 30, 2026.