How are leading manufacturers navigating the complexity of modern operations? With labor shortages, supply chain volatility, and rising performance expectations, many are turning to analytics and AI—not just to gather data, but to unlock smarter, faster decisions across the factory floor.
What Sets Top Performers Apart
In a recent global research study conducted by Tech-Clarity and MESA International, and sponsored by Arch Systems, more than 420 manufacturers shared how they’re using analytics and AI to improve their operations. Among the participants, a distinct cohort stood out: 21% of respondents consistently outperform their peers across key operational metrics like OEE, throughput, yield, and capacity utilization. These top-performing companies are not just investing in digital tools, they’re translating those investments into measurable gains in both factory performance and business outcomes.
So, what are they doing differently?
Compared to their peers, top performers are:
- More than twice as likely to have tightly integrated IT and OT organizations, enabling seamless data flow and faster collaboration
- Far more likely to use AI and machine learning for monitoring and detection (75% vs. 56%), providing real-time visibility and early warnings to prevent issues.
- Three times more likely to hit cost targets at least 96% of the time, a strong indicator of financial discipline and strategic alignment.
They also outperform across business metrics like operating margin, cost of quality, and perfect order rates, showing that operational excellence translates into competitive advantage.
“We now need a team of three engineers for digitalization, including one on the shop floor. It’s worth the investment because the results are there—even if you don’t see them up front.” ––David Batet, CIO, DigiProcess
What’s especially telling is that Top Performers don’t just use technology—they structure their organizations and strategies around it. They prioritize use cases based on business value, foster cross-functional collaboration, and invest in systems that scale. Many also maintain a bias toward action, starting with small, focused pilots and expanding them systematically.
In short, Top Performers are proving what’s possible when manufacturers align data, tools, and people behind a shared goal: intelligent, continuous improvement.
As these companies advance, they don’t stop at dashboards. Instead, they evolve their analytics strategies—moving from visibility to foresight to action.
Descriptive → Predictive → Generative: The Analytics Maturity Path
As manufacturers evolve their digital strategies, they typically move through three distinct, but connected phases of analytics maturity. Each step builds on the last, advancing from visibility to foresight to real-time, AI-supported action. The journey isn’t linear for everyone, but understanding this progression is essential for unlocking the full value of your data and systems.
1. Descriptive Analytics: Seeing Clearly
This first phase focuses on understanding what is happening—or has happened—on the shop floor. It includes real-time dashboards, KPIs, and historical data analysis.
85% of Top Performers use interactive dashboards, which offer plant-specific, role-specific visibility.
The goal is to enable continuous improvement (CI), spot anomalies quickly, and create a shared language across roles.
“We are creating safety and auditing tools. The more you can automate, the more you can control errors. The fewer keystrokes someone has to do, the better.”
— Jason Bassett, IT Manager, Madsen’s Custom Cabinets
But descriptive analytics alone can’t drive proactive operations or close the knowledge gap between experienced engineers and newer frontline staff.
2. Predictive Analytics: Acting in Advance
Predictive analytics takes you a step further, using historical and real-time data to forecast what will happen next. This often involves machine learning models trained to identify patterns in process drift, equipment failure, or quality issues.
Nearly all Top Performers in the study use predictive AI, and 76% use Digital Twins to simulate and optimize performance.
Predictive tools help shift from reactive firefighting to proactive control, saving time, reducing costs, and preserving quality.
“Our predictive analytics solution pulls together data from various sources into a comprehensive dataset… we constantly refine these models to predict the desired outcomes accurately.”
— Will Spears, Sr. Product Owner, Smart Manufacturing, Sonoco
Still, scaling predictive analytics across a diverse factory ecosystem often requires strong data governance and cross-functional buy-in.
3. Generative AI: Guiding the Workforce
Generative AI represents a transformational leap, especially in how it supports people. By understanding language, context, and structured/unstructured data, GenAI enables frontline workers, supervisors, and executives to ask questions and receive intelligent guidance in natural language.
Use cases include delivering SOPs, answering ERP or MES queries, translating complex data into actionable insights, and automating documentation.
Top Performers use GenAI to close the workforce knowledge gap, reduce training time, and increase consistency across shifts and sites.
“We got early access to Epicor’s Knowledge Assistant… users can ask it questions like ‘How many open POs do I have?’ It saves them time—and it saves me time.”
— Jason Bassett, IT Manager, Madsen’s Custom Cabinets
As manufacturers adopt GenAI, they’re not just accelerating decisions, they’re reshaping how expertise is scaled and embedded into day-to-day work.
Yes, There are Still Hurdles
Despite the momentum, digital transformation in manufacturing is far from turnkey. The study reveals that while nearly all manufacturers are investing in analytics and AI, persistent barriers continue to slow progress, especially when it comes to scaling solutions across plants, teams, and legacy infrastructure.
Top Challenges Include:
- Poor data quality and governance (60%) – Without clean, contextualized data, even the most advanced algorithms can’t deliver meaningful insights.
- Lack of real-time data access (37%) – Time-sensitive decisions are only possible when operators and engineers have immediate, trustworthy data at their fingertips.
- Inability to scale from pilot to broader rollout (31%) – Many manufacturers see early success with proof-of-concept projects but struggle to replicate results across lines or facilities.
- Limited internal data science skills (50%) – The talent shortage extends beyond operations to include AI-savvy analysts and developers.
Interestingly, cultural resistance and lack of trust in AI tools—often seen as major barriers—were less of a factor among companies that had already started using predictive or generative AI. Successful pilots, it seems, build internal credibility and help overcome skepticism.
“When we started, we didn’t know how difficult it was going to be—mainly because of the cultural shift. But when people see the benefit, it’s exponential. It creates a singularity that makes things go very fast. […] When the culture and processes begin to change, you see a new paradigm. That’s the nice thing. Once we had the data, we could implement new ways to collect it—and do more magic. We now see it because we’ve fulfilled the other layers.”
— David Batet, CIO, DigiProcess (quote lightly edited for clarity and continuity)
The findings point to a critical insight: technical hurdles are real, but they can be overcome with focused execution, internal upskilling, and the right partners. As more vendors embed predictive and generative capabilities into their platforms, the need for specialized AI talent may diminish—but data quality and integration will remain essential.