What Else Could We Measure?

Why data-driven management often starts from the wrong end

"We now have a comprehensive people analytics solution. What else could we measure?"

This question is commonly heard in organizations that have invested in data infrastructure. It seems reasonable: when the capability to measure exists, why not utilize it more broadly?

However, the question reveals a problematic premise. It assumes that collecting data itself produces value. That a data pool is inherently worthwhile, with benefits to be determined later.

This is usually not the case. Merely hoarding data does not produce value in itself. Value emerges from measurement, where data connects to objectives and decision-making.

The paradox of data-driven management

Data-driven management has become an established organizational goal. The idea is appealing: when decisions are based on data, they are better than those based on intuition.

In practice, this often leads to a peculiar situation.

First, systems are acquired that enable data collection. Then data is collected because the systems make it possible. Finally, consideration is given to what might be done with the collected data.

This is data hoarding. It produces data pools that are assumed to represent reality. But a data pool without an objective is just numbers in a database.

When asking "what else could we measure", one is often already in the wrong place. The question is not what data can be collected. The question is what one wants to achieve.

Measurement is different from data collection

Measurement produces value when it includes:

• Identifying the phenomenon. What is this actually about? What is the thing we want to understand – and how does it manifest in practice? The phenomenon exists regardless of what we want to do about it. Operationalising the phenomenon means making it visible and measurable.

• Setting objectives. Once the phenomenon is identified, we can ask: where do we want to end up? The objective determines which aspect of the phenomenon is measured and in which direction change is sought. The phenomenon tells us what is happening. The objective tells us what we want to happen.

• Developing metrics. Which metric captures the phenomenon we want to understand? Is the metric valid and reliable?

• Measuring. Baseline, follow-up, comparison. Measurement that produces an interpretable result.

• Leveraging the survey effect. The mere act of asking about something directs attention and conversation. This effect can be utilized deliberately.

Data hoarding, by contrast, is collection without these elements. It produces data that answers no question, because no question was ever asked.

Metrics without objectives

Organizations typically have numerous metrics. HR tracks turnover, sick leave, eNPS, training days, time-to-hire. Each metric produces numbers. Numbers are reported. Reports are archived.

These basic metrics are important. They serve as wellbeing indicators and warning signs. They tell us there is smoke somewhere. But they do not tell us where the fire is or how to extinguish it. They are status metrics, not change metrics.

But what happens then?

If turnover is 12%, is that good or bad? What is it compared to? What should it be? And above all: why is turnover a relevant metric for this organization's objectives?

A metric without an objective easily remains just a number. It does not indicate whether action should be taken or not. It does not guide decision-making. It consumes time without producing value.

Yet metrics continue to be added. Because there must be data. Because others measure it too. Because the system makes it possible.

Wrong question, wrong order

"What else could we measure?" is often a misleading question because it starts from the tool rather than the need.

The correct order is:

1. What phenomenon are we dealing with? What do we want to understand?

2. What do we want to achieve? (If the phenomenon is already clear, this can also be the first question.)

3. How will we know we have succeeded?

4. What must we measure to obtain this information?

In this order, the metric is a tool, not an end in itself. It serves the objective, which emerges from the identified phenomenon. It produces information needed for decision-making.

When the order is reversed – first the metric, then perhaps the objective, the phenomenon never identified – data is generated that leads nowhere.

Example: Employee survey

Many organizations conduct an annual employee survey. It contains dozens of questions about wellbeing, leadership, engagement, and work environment.

The result is a report containing many numbers. The executive team reviews it. Departments receive their own results. Then everyone returns to daily work.

What happened? What decisions were made based on the results? What actions were taken?

Often the answer is: hardly any. The survey was conducted because it is part of the annual cycle. Data was collected because that is the custom. But because there was no objective, the data led nowhere.

Compare this to a situation where the organization has identified a phenomenon: collaboration between teams is not working. The phenomenon has been named and operationalized. The objective is to improve collaboration. A metric is built to measure precisely that. Results are monitored. Actions are taken. Measurement is repeated.

The same tool – a survey – but an entirely different outcome.

Data quantity does not replace data quality

The notion that more data is always better is common, but often mistaken.

Large data pools without focus produce noise. They obscure what is essential. They consume resources for collection, storage, and reporting without anyone doing anything with the data.

Furthermore, a data pool easily creates an illusion of comprehensiveness. When there is a lot of data, an impression forms that it represents the whole of reality. But if data is collected from one perspective or covers only part of the operating environment, it can systematically mislead.

Purposeful measurement produces less data but more value. It answers a question that is actually relevant. It ensures that the data covers the phenomenon from sufficiently diverse perspectives. It enables a decision.

The question is not how much data there is. The question is whether the data answers the question that needs to be resolved – and whether it covers the phenomenon from enough angles.

Measurement starts with the phenomenon

When measurement starts from identifying the phenomenon, the process looks different.

Step 1: Identify the phenomenon What is this about? What is the thing we want to understand or change? How does it manifest in daily work – what do we see, hear, experience? Describing the phenomenon in one's own words is the first step toward operationalising it.

Step 2: Define the objective Once the phenomenon is identified, we can ask: where do we want to end up? The objective determines which aspect of the phenomenon is examined and in which direction change is sought.

Step 3: Success criteria How will we know the objective has been achieved? What is the concrete change that should be visible?

Step 4: Select the metric Which metric captures this change? Does an existing metric exist, or must a new one be built?

Step 5: Baseline measurement What is the current situation? Without a baseline, change cannot be demonstrated.

Step 6: Actions and follow-up What will be done to achieve the objective? When will measurement be repeated?

In this model, the metric is not the first step, but serves the objective, which in turn emerges from the identified phenomenon.

What evaluoi.ai does differently

evaluoi.ai is built to be phenomenon-driven.

The process always begins with the question: what is this about? The system guides users to describe the phenomenon – what is happening – before creating a metric. This operationalization makes the invisible visible.

Once the phenomenon is identified, the objective is defined: where do we want to end up? The objective determines which aspect of the phenomenon is measured and in which direction change is sought.

The metric is built to serve the objective, which emerges from the identified phenomenon – not to collect data for its own sake.

This also means the system does not encourage measuring everything possible. It encourages measuring what is relevant.

When a metric is linked to an objective and the objective to an identified phenomenon, results are interpretable. They indicate whether progress has been made toward the objective or not. They enable a decision: continue in the same way or change direction.

Conclusion

"What else could we measure?" is a question worth examining critically.

A better question is: "What phenomenon are we dealing with, what do we want to achieve, and what do we need to know about whether it is succeeding?"

Data-driven management does not mean as much data as possible. It means the right data for the right question. Data hoarding is not measurement. Measurement is a process that begins with identifying the phenomenon and produces value even before results are known – because it forces us to make the invisible visible.

When this order is correct, measurement produces value. When the order is reversed, data pools are created that no one uses.

Book a demo or watch the intro video: evaluoi.ai