Companies are using AI to speed up broken processes, not fix them
Businesses across Europe are wasting AI's potential by applying it to fragmented systems and outdated workflows instead of redesigning how they operate, according to Steve Crosson Smith, Head of Data Strategy, Governance and Architecture at SDG Group.
The problem is common: firms generate reports and dashboards faster, but leaders still lack clarity on where real problems lie or what action to take. Speed applied to a broken process just means broken results arrive quicker.
The siloed data trap
Most companies remain trapped by departmental thinking. Each team optimizes its own metrics while the bigger picture stays fragmented across separate data sources.
Crosson Smith said modern decision-making requires connecting information across departments and systems. "KPIs are still important, but you need a much more joined-up view of the company across all these data silos," he said.
When individual teams use AI to boost personal productivity, that gain often doesn't improve the overall process. A faster report from one department might still arrive too late for another team to act on it.
Rethink the process, then automate it
The real opportunity lies in stepping back first. Identify the bottlenecks and barriers slowing your operation. Then use AI to redesign how that process should work if those constraints didn't exist.
"Don't just try and make individual parts of your processes go quicker. Reimagine those processes as if those bottlenecks and barriers didn't exist, then redesign them and use AI to power that redesign," Crosson Smith said.
This approach moves beyond raw productivity gains. It asks whether the work being done faster actually matters to the business outcome.
Data quality comes first
AI can only work with reliable input. Data pulled from internal systems and external sources needs processing, validation and preparation before it supports real analysis.
Crosson Smith emphasized the need for "high-quality, trusted data" available quickly through a modern, scalable platform. Without this foundation, AI just amplifies existing problems.
AI can also process unstructured material - documents, videos, images - that was previously too difficult to analyze at scale. That opens new sources of insight.
From numbers to answers
The commercial value of AI lies in moving beyond metrics to interpretation. SDG Group's InsightGen platform generates automated reports and allows non-technical staff to ask questions of company data in plain language.
Crosson Smith said people don't want dashboards. They want answers: "Why did something happen? And most importantly, what can I do about it to mitigate a negative outcome or create a positive one?"
Treating AI as a consultant that works alongside humans, rather than a faster calculator, changes how operations teams approach decisions.
What comes next
Crosson Smith predicted businesses would move toward hyperautomation, where AI makes decisions and takes action within set guardrails - in procurement, approvals, and similar functions.
Self-healing software that can diagnose problems, write code and deploy fixes is also moving from theory to practice, he said. That capability will reshape how operations teams manage systems and respond to failures.
For operations professionals, the lesson is clear: before asking AI to speed up your work, ask whether the work itself needs redesigning.
AI Learning Path for Operations Managers covers process optimization and using AI to improve operational efficiency. Data Analysis resources address how to connect siloed information and improve decision-making.
Your membership also unlocks:
