08/06/26 17:17

Why Manufacturers Don’t Have a Data Problem

Senior manufacturing leaders rarely complain about not having enough data.

They talk about not being able to see the whole picture, not trusting the numbers they do see, and struggling to move from reports to real decisions. That isn’t a manufacturing data problem. It’s a visibility, confidence and action problem.

From ‘data problem’ to ‘decision problem’ in modern manufacturing

Most manufacturers already sit on years of sensor readings, production logs, energy bills, MES events and supplier metrics. The real issue is that this information is fragmented, late, or hard to turn into decisions at the pace the business now requires.

A recent survey of more than 600 manufacturing leaders found that while 55% rely on automated machine data, 50% still depend on manual frontline input, creating a hybrid environment where critical insights are delayed or distorted. Only three in ten leaders say their data reflects the shop floor in real time, and just 21% find it very easy to access the information they need, according to research reported in Manufacturing Dive.

In other words, plants are not short of data; they are short of usable, timely, trusted intelligence. Leaders experience this as:

Endless dashboards but no clear narrative
Conflicting numbers from finance, operations and sustainability
Analysis cycles that take weeks, long after the window to act has closed

The result is a decision bottleneck. Strategy moves slower than disruption, even as data volumes grow.

Where the data actually lives: OT, energy and supply chain examples

If you map a typical manufacturer’s footprint, you quickly see that data is already embedded in almost every part of the value chain. The problem is not collection; it is coherence.

On the shop floor, PLCs, HMIs, industrial IoT sensors and quality systems stream second‑by‑second information on throughput, scrap, downtime causes and environmental conditions. One automation specialist notes that process data is collected within every control device and meter, yet plants often remain “data rich and information poor,” as highlighted by Automation World.

Energy data is similarly abundant. Smart meters log half‑hourly consumption; utilities publish tariffs; building‑management systems track temperature, compressed‑air load and HVAC cycling. Many plants already have sub‑metering on major assets, but those readings sit in vendor portals, spreadsheets or PDF bills that no one consolidates.

Supply chain and logistics teams, meanwhile, manage detailed order histories, shipment tracking events, lead‑time variability, supplier scorecards and even Scope 3 indicators. Research from Capgemini shows that 70% of executives identify new‑generation, AI‑enabled supply chains as a top tech trend, with 76% focusing on risk management and 68% on clear supply chain objectives.

Individually, each of these datasets is rich. The challenge is stitching them together so a plant manager can answer a simple question such as: “Which suppliers, product lines and shifts are driving both cost and carbon this quarter?”

Why insight breaks down: silos, quality issues and trust gaps

If the data is there, why does it fail leaders at the moment of decision? The root causes tend to sit in four areas: silos, quality, context and ownership.

First, systems were never designed to talk to each other. Years of uncoordinated technology investment have left manufacturers with MES, ERP, maintenance and energy platforms that each hold a partial truth. Grant Thornton describes this as a lack of “data foundation” that leaves AI and analytics running on disconnected, low‑quality inputs, undermining reliability and trust, as noted in Grant Thornton’s analysis.

Second, quality and timeliness are inconsistent. L2L’s research, covered in their 2026 data report, found that over 65% of frontline supervisors waste up to four hours per shift on manual data entry and reconciliation. When figures are stitched together by hand at the end of a shift or month, errors are inevitable.

Third, data often lacks business context. A beautifully tagged historian is still just a list of tags if no one has translated them into language that finance, procurement or sustainability teams can use.

Finally, ownership is fuzzy. Without clear governance, no one is accountable for definition drift (“What exactly counts as ‘energy for production’?”) or for resolving conflicts between sources. That is how “data chaos” emerges, even in technically advanced plants.

Turning raw signals into shared visibility for leaders and teams

Fixing this is less about buying another platform and more about designing how information flows from the line to the boardroom. The aim is shared visibility: the same numbers, seen through different lenses, by operations, finance, sustainability and procurement.

A practical starting point is a single, joined‑up data model for a limited scope: for example, one plant, or one major product family. Rather than trying to integrate every tag and every system, leading manufacturers focus on a handful of critical questions, such as:

What is our true, all‑in cost per unit by line and product?
Where do we incur avoidable energy and material losses?
Which suppliers and lanes expose us to the most risk and emissions?

They then work backwards to identify the minimum data needed, the systems that hold it, and the rules for cleansing and joining it. This approach mirrors what Capgemini describes as building a “data/AI foundation” to unlock siloed sources and support smart forecasting across the supply chain.

On the ground, shared visibility looks like a control‑tower style view that blends:

OT data (throughput, downtime, scrap, rework)
Energy data (kWh, fuel, steam, compressed air)
Commercial data (price, margin, customer priority)
Sustainability data (Scope 1 and 2, and where possible, Scope 3)

When teams see the same reality, conversations shift from arguing about the numbers to arguing about the options.

From dashboards to decisions: building confidence to act

Even when visibility improves, many organisations still struggle to act. Leaders are understandably cautious about shutting a line, switching suppliers or redesigning a product based on analytics alone. Confidence to act requires traceability and clear rules.

First, every number on an executive dashboard should be drillable. A sustainability or operations lead needs to be able to click from a site‑level carbon intensity metric down to the specific assets, shifts or suppliers that drive it. This lineage builds trust: people can see where figures come from and how they are constructed.

Second, thresholds and playbooks matter. Research from L2L highlights how hybrid data collection leaves teams to interpret metrics on the fly. In contrast, high‑performing plants codify responses. For example:

“If unplanned downtime exceeds X minutes per shift for three days, we trigger a cross‑functional root‑cause review.”
“If energy intensity for a product moves more than Y% from baseline, we investigate changeovers, idle time and compressed‑air leaks.”

Third, leadership behaviours signal that data‑driven decisions will be supported. When a plant team takes action based on a new energy‑and‑yield view, and the outcome is mixed, the response should be to refine the model, not retreat to intuition.

Confidence does not come from perfect data; it comes from transparent data, clear decision rules and a culture that learns publicly.

Practical use cases: cost, resilience and sustainability wins

Concrete examples help make the opportunity real. Across manufacturing, the same underlying datasets are already being used to drive commercial and sustainability value when they are connected.

Cost efficiency: Capgemini reports that 61% of executives say they have effectively reduced supply chain operational cost, with 73% citing automation and process improvements as key levers. Combining production, maintenance and energy data allows plants to identify “high‑cost hours” where overtime, peak tariffs and scrap all spike—and then redesign staffing and schedules to avoid them.

Resilience: 76% of executives focus on risk management across the supply chain, and 64% use supplier‑diversification ratios as a resilience KPI. When supplier performance data is joined with logistics events and plant downtime causes, procurement can see which suppliers create the most disruption per unit purchased, not just the lowest nominal price.

Sustainability: Three‑quarters of organisations in Capgemini’s research agree that sustainable practices drive cost efficiencies, while 71% believe the business value of sustainability initiatives outweighs the cost. By aligning energy meters, production data and emissions factors, manufacturers can pinpoint which product‑equipment‑supplier combinations drive the highest emissions per unit—and then trial lower‑carbon alternatives with a clear business case.

These are not speculative pilots. They are examples of what becomes possible when existing data is cleaned, connected and pointed at specific business questions.

Laying the foundations for AI, digital twins and agentic supply chains

Many manufacturers are exploring digital twins, GenAI and agentic AI in the supply chain. Yet without a solid data foundation, these initiatives will over‑promise and under‑deliver.

Capgemini’s 2025 research shows that 51% of organisations have already implemented more than five GenAI use cases in supply chain, and 27% have established dedicated agentic‑AI teams. At the same time, Grant Thornton warns that disconnected, low‑quality data is “dragging down” AI, limiting impact on visibility and decision‑making.

The implication is clear: the groundwork for advanced AI is largely organisational, not just technical. It includes:

Common definitions for core metrics across operations, finance and sustainability
Governance that assigns ownership for data quality and lineage
Integration patterns that allow OT, IT and external data (for example, tariffs and emissions factors) to coexist

As TXI points out in its discussion of manufacturing intelligence, the differentiator is not how much data a plant has, but how it uses that data to make faster, better decisions, as outlined in TXI’s insight. AI amplifies whatever foundation it sits on—clarity or confusion.

From conversation to action: themes for the Data & Reporting lunch

This raises the question we will explore in detail during our Data & Reporting working lunch: what should senior manufacturing, operations, sustainability, procurement and digital leaders do next if they accept that they do not have a data problem, but a decision problem?

Three themes will frame the discussion:

From systems to questions. How to start with a small set of cross‑functional questions—cost, risk, carbon—and work backwards to the minimum viable data foundation that answers them.
From owning data to sharing visibility. Practical ways to break down silos between OT, energy, supply chain and finance without replacing every system you already have.
From pilots to operating rhythm. How to embed trusted metrics and playbooks into monthly S&OP, capital‑allocation and sustainability‑governance cycles so that insight reliably turns into action.

Manufacturers do not need more data dashboards. They need clearer line of sight from the factory floor to the boardroom, so that every terabyte they already collect can support faster, more confident decisions in an increasingly volatile world.

Supply Chain & Operations