In the first article of this series, I argued that measuring quality does not necessarily mean knowing quality. In the second, I examined the chain that connects measurement with management and showed that information must pass through several transitions before it can influence a decision.
This leads to another question:
If organizations collect more data than ever before, why do they still struggle to understand their own quality?
One possible answer is that data and knowledge are often treated as though they were interchangeable. They are not. Data provides a representation of what has been observed, while knowledge emerges only when those observations are interpreted, connected, tested, and placed within a meaningful context.
Every measurement begins with an observation
When a quality indicator changes, something has been observed. A defect may have been recorded, a complaint received, a delivery delayed, or a machine found to be operating outside its normal vibration range. Each observation provides useful information about the condition of a process.
However, an observation is not yet an explanation.
It can tell us that something happened, but it does not necessarily explain why it happened, how it relates to other events, whether it represents a temporary deviation or a systemic problem, or what action should follow. Measurement therefore improves visibility, but visibility alone does not produce understanding.
This distinction is important because organizations often assume that once an event has been captured and displayed, the underlying quality problem has already become known. In practice, the organization may only know that an indicator has changed.
Data does not contain its own meaning
Data is often discussed as though it carries answers within itself. Yet most data contains signals rather than conclusions. Meaning appears only when those signals are interpreted.
Consider a rising defect rate. It may indicate equipment deterioration, variation in incoming materials, insufficient operator training, an unstable production method, or a planned change that has temporarily affected process performance. The indicator can reveal that a deviation exists, but it cannot independently determine which explanation is correct.
Someone must connect the number with the reality of the process.
That connection requires technical knowledge, operational experience, historical comparison, an understanding of current conditions, and awareness of the assumptions built into the measurement itself. Without these elements, even accurate data can support an incomplete or misleading interpretation.
Context gives data its meaning
A measurement never exists in isolation. Every quality indicator is part of a much larger system of relationships. A defect rate, a process capability index, customer complaints, equipment downtime, delivery delays, or audit findings all describe particular aspects of a process. Yet none of these indicators explains the process by itself.
A defect rate of 3%, for example, has no inherent meaning.
- Is it high or low?
- Has it increased or decreased?
- Is it statistically significant?
- Does it affect one production line or the entire facility?
- Is it connected with a supplier change, a new operator, seasonal demand, or a modification in process parameters?
Without answers to these questions, the number remains only a measurement.
Its meaning emerges only when it is interpreted within the context of the process that produced it. For this reason, context should not be viewed as additional information surrounding data. Context is what allows data to become knowledge in the first place.
The same indicator may therefore lead to very different conclusions not because the measurement changes, but because the surrounding understanding changes.
Organizational knowledge is distributed
Another difficulty is that the knowledge required to understand quality rarely exists in one place.
Operators may understand the daily behaviour of equipment and notice small changes that never appear in formal reports. Engineers may understand the technical relationships between process parameters. Quality specialists may recognize patterns in statistical variation. Managers may understand commercial priorities, resource constraints, and wider organizational risks. Customers may experience consequences that remain invisible inside the company.
Each group sees a different part of the same system.
The challenge is not simply to ensure that every participant has information. It is to connect these different perspectives into a coherent explanation of what is happening. An organization may possess all the necessary fragments of knowledge and still fail to assemble them into shared understanding.
In this sense, knowledge can exist within an organization without fully becoming organizational knowledge.
More data does not guarantee more knowledge
Modern organizations have become remarkably effective at collecting information. Sensors continuously monitor equipment, ERP systems record transactions, MES platforms capture production events, dashboards update indicators in real time, and artificial intelligence can identify patterns across millions of records.
As a result, the technical cost of producing and processing data has fallen dramatically. Organizations can now observe processes at a level of detail that would have been impossible only a few decades ago.
Yet the creation of shared understanding has not become equally simple.
Understanding still depends on professional experience, communication, interpretation, collaboration, and the willingness to question accepted explanations. It requires people to compare perspectives, identify contradictions, test assumptions, and distinguish meaningful signals from background variation.
Technology can support these activities. It can make patterns more visible, identify anomalies, retrieve relevant information, and accelerate analysis. But it cannot guarantee that the organization will interpret the result correctly or reach a shared conclusion.
The ability to collect more data and the ability to understand more are related, but they are not the same capability.
The hidden imbalance
This creates an important imbalance in many quality systems.
Organizations invest heavily in improving the collection, storage, visualization, and analysis of data. They purchase new software, introduce additional indicators, automate reporting, and increase the frequency of monitoring. These investments are visible, measurable, and relatively easy to justify.
Far less attention is often given to the mechanisms through which knowledge is created, challenged, shared, and used. Who interprets the data? Which perspectives are included? How are disagreements resolved? How does operational knowledge reach decision-makers? How is an explanation tested before it becomes the basis for action?
When these questions remain unanswered, the volume of available information may grow much faster than the organization’s ability to understand it.
The result can be a data-rich but knowledge-poor quality system: one that describes many aspects of performance but still struggles to explain what is happening or determine what should be done.
From data availability to quality knowledge
This suggests that the next stage of quality management may not be defined simply by the ability to measure more. It may be defined by the ability to transform measurement into reliable organizational knowledge.
That transformation requires more than data collection. It requires context, interpretation, communication, integration of expertise, and a clear connection between what is observed and what the organization believes it understands.
The relevant question is therefore not only:
How much quality data do we have?
It is also:
How much do we actually know about the quality we measure?
The answers may be very different.
In the next article, I will explore a related paradox: Can an organization have excellent dashboards and still misunderstand its own quality?