In large part, 2025 supply chains have been defined by tariffs. More than just an accounting problem, they have become an ongoing point of contention and stress for manufacturers struggling to protect margins and maintain product sustainability. Amidst this volatility, success increasingly depends on one critical factor: how well manufacturers understand and act on the data that underpins product quality and margin performance.

As tariffs continue to shift the manufacturing landscape, whether it be raw materials like lumber or more finished products like furniture, each policy shift demands that leaders re-evaluate costs, suppliers, and schedules.  Historically this would have been a herculean task, today’s technology can help, but to do so it needs the right inputs, the right data. 

This isn’t new, nor is it rocket science to know that when data is fragmented, errors creep in and decisions lag. Teams rely too heavily on legacy understanding or gut feelings that don’t have measurement or governance in place. As processes strain under new untested variables like cost spikes, lead-time changes, and material substitutions; risk to product quality can arise. In fact, many manufacturers are finding that the real challenges aren’t the tariffs per se; but rather that their data systems can’t model the financial or operational impact quickly enough to respond with confidence.

Quality under pressure

In these instances, data blind spots aren’t just slowing operations; they’re also eroding trust as fragmented systems run the risk of compromising the one area manufacturers can’t afford: quality. At present, there’s a widespread belief among executives that intelligent manufacturing systems will be the engine behind America’s next industrial advantage. However, while there is a lot of optimism around this technology, for manufacturers to succeed it’s vital that they have the right data infrastructure, quality, and governance processes in place to support their ambitions. To achieve these goals a clear roadmap is needed, but all too often, the ability to develop these roadmaps is clouded by disjointed data, making it extremely difficult to turn strategic ambition into measurable outcomes, and even harder to trust the results that follow. 

In fact, Avanade’s 2025 AI Value Report found that while 96% of surveyed decision makers are using AI to help make those decisions, only 26% fully trust the results. This tension between trust and decision making is especially interesting in the context of data, where concern about data quality is called out as organizations’ top reason for limiting employee usage or access to generative AI tools.

This lack of trust often stems from a more fundamental issue: data fragmentation. When data from sourcing, cost management, and quality control that’s needed to inform wider decision making across the business is derived in silos, it becomes increasingly difficult to see the full picture of how a tariff or raw-material cost swing will affect product performance or profitability. Especially when disparate data is lost to different spreadsheets, pieces of paper, or memory. Without a centralized, trusted repository, manufacturers risk making the wrong decisions and adjustments. Worse yet, they risk not knowing what went wrong and with macroeconomic consequences raised, even a single misstep could cause a costly ripple through a complex global supply chain. 

For manufacturers to set themselves up for success in times of change, they must prioritize modernizing legacy ERP systems, integrating quality and production data, and establishing governance standards that ensure data accuracy from the shop floor to the boardroom. Once that foundation exists, AI can begin to reveal patterns that were previously invisible, like how a supplier’s late shipment will affect production, or the financial impact tariff-driven material substitution could lead to greater variation in part dimensions.

This is where modern data architecture comes into play. A modern data architecture can turn isolated quality metrics into a living system of continuous feedback. Instead of reacting to issues after the fact, organizations can simulate scenarios in real time and adjust before a problem reaches the line or the customer. Turning that vision into reality requires a clear, practical roadmap for transformation.

A practical roadmap for resilient quality

Building a modern data foundation is only the first step. The next is translating it into practical actions. These three actions consistently improve resilience when trade or market conditions shift. 

• Unify data around the process. Integrate production, supplier, and financial data into a single, governed environment. This creates the visibility needed to understand how external pressures like tariffs or transportation costs affect product quality and throughput.
• Target high-impact use cases. Start with problems that directly influence yield or customer satisfaction, such as optimizing inspection schedules or predicting when supplier variability will cause defects. Connecting these use cases to cost and schedule data demonstrates measurable ROI fast.
• Measure results in weeks, not quarters. Short validation cycles prove the link between data modernization and operational outcomes. When quality and finance teams can show improvement together (lower scrap, shorter lead times, reduced cost of quality), the business case for scaling AI becomes self-evident.

Real-time visibility in action

While these three actions form a solid foundation for resilient quality, some manufacturers are taking this further by leveraging external tools which bring together financial, supply, and operational data to model the impact of tariff changes in real time. By simulating how new import duties ripple through material costs and supplier networks, teams can make informed choices before disruptions compromise production or quality.

What’s more, this value and type of visibility isn’t limited to tariffs. The same foundation supports predictive maintenance, energy optimization, and defect prevention. The common denominator being trustworthy data. When every function works from the same version of truth, organizations can finally move from reactive to proactive.

A catalyst for smarter manufacturing

Tariffs may be the current stress test of process integrity, but they reveal a broader truth: fragmented data limits agility, while intelligent systems unlock it. By investing in data readiness and intelligent systems today, manufacturers gain not only the agility to navigate tariffs but also the foundation for continuous improvement in product quality, yield, and sustainability.

Tom Nall, Manufacturing Industry Director, Avanade