Stop Calling it Soft Skills
The important things at work can be measured.
In business, it’s a common but lazy term.
If something is financial, operational, or technical, we call it hard.
If it involves people, things like confidence, trust, readiness, leadership, we call it soft. Soft becomes shorthand for nice but not measurable.
That’s an expensive mistake.
Organizations spend millions optimizing workflows, systems, and technology while treating judgment, coordination, and readiness as side effects instead of performance drivers. Then they are surprised when productivity plateaus, safety incidents repeat, the culture doesn’t support the mission, or leadership pipelines run dry.
The forces that most reliably shape performance, retention, safety, strategy, and execution are in how people think, decide, coordinate, and adapt when the work gets complex. Calling those capabilities soft makes them easier to ignore.
Subjective Doesn’t Mean Lacking Credibility
Some things are objective and directly observable, such as revenue, cycle time, and defect rates.
Some things are real but not directly observable, like confidence, engagement, judgment, readiness, and trust.
In measurement science, those second-category factors are called latent constructs: real variables that cannot be observed directly but can be inferred from consistent patterns of behavior, self-reports, and performance indicators (Borsboom, 2005; DeVellis & Thorpe, 2021).
This is standard practice in behavioral science and organizational psychology. We measure subjective phenomena indirectly, using structured methods designed to be:
Reliable — producing consistent results across time and conditions
Valid — reflecting the construct we claim to be measuring (Kaplan & Saccuzzo, 2017)
This is the domain of psychometrics and scale development. Subjective experiences can be measured objectively enough to support serious decisions using quantitative data derived from structured indicators rather than informal impressions or anecdotes.
How You Measure the Subjective Stuff at Work
When organizations do this well, they translate abstract concepts into observable indicators. For example:
Confidence
willingness to take initiative, persistence after setbacks, self-ratings of task capability
Readiness
performance in simulations, time to independent execution, error-recovery behavior
Trust
information sharing, escalation behavior, willingness to admit uncertainty
Engagement
discretionary effort, participation, retention risk indicators
Then, using scientific methods, translate the theory-defined constructs into measurable variables. Common measurement approaches include:
Structured surveys and rating scales • Behavioral observation checklists • Manager and peer assessments
Consistent interview protocols • Performance-based scenarios and simulations
These tools use consistent scoring scales, so responses are converted into numerical data that can be tracked over time, compared across teams, and linked to outcomes such as performance, safety, retention, and more.
These methods are structured, repeatable, and designed to reduce random judgment by converting experience and behavior into consistent, analyzable data. They are used in high-stakes domains such as personnel selection, safety-critical training, and clinical assessment, where decisions carry real consequences (Kaplan & Saccuzzo, 2017).
Sure, they are imperfect, but so are financial forecasts and operational dashboards. We still use those because they are useful approximations, not because they are flawless.
Decades of research in naturalistic decision making, safety science, and human capital show system outcomes hinge on adaptive human performance, not just procedural compliance.
– Klein, 1998; Hollnagel et al., 2006; Ployhart & Moliterno, 2011
Why Measuring the Subjective Stuff Matters in Complex Work
In stable, tightly controlled environments, performance often improves when we standardize procedures, clarify rules, and optimize workflows and processes. But most professional work today is not fully specifiable or stable.
In complex, high-stakes environments such as healthcare, energy, operations, leadership, and large-scale change, performance depends largely on:
How people interpret evolving situations
How they coordinate under pressure
How they adapt when plans break down
How they make judgments with incomplete and imperfect information
It is common wisdom in safety science that people are not the problem in complex systems. They are the heroes, as the source of adaptation that keeps work functioning when conditions change (Cook, 2000). Decades of research in naturalistic decision making, safety science, and human capital show system outcomes hinge on adaptive human performance, not just procedural compliance (Klein, 1998; Hollnagel et al., 2006; Ployhart & Moliterno, 2011).
These are cognitive and social capabilities. They cannot be managed well if we don’t measure them. When organizations avoid measuring these factors because they are considered soft, they default to what is easiest to count instead of what actually drives outcomes.
The Real Risk Is Not Measuring People
There’s reluctance tied to the soft-skills label and an assumption that measuring human capability is too fuzzy to be fair, or too political to be useful. But the alternative is worse.
When we don’t measure:
We train without knowing if the performance issue is truly a knowledge and skills problem
We promote because some experienced people said “we know readiness when we see it”
We intervene without certainty that the performance gap can be fixed by our current programming
we redesign systems without understanding how people are experiencing them
In practice, this leads to pet solutions like coaching when the problem is structural, training when the problem is cultural, policy when the problem is capability. This is a classic sociotechnical failure: treating human behavior as the root cause when the real constraints are embedded in system design (Dekker, 2014).
If the only tool you have is a hammer, everything looks like a nail. Measurement helps you figure out what kind of problem you’re actually dealing with before you swing.
Good Measurement Is a Thinking Discipline
Objective measurement of subjective constructs is about asking better diagnostic questions:
What exactly are we trying to change?
How would improvement show up in real behavior?
When would we expect to see movement?
What else could be influencing the results, and how would we separate that out?
That kind of measurement forces clearer thinking about causality and correlation. It connects interventions to outcomes instead of treating development like a black box. Done well, it turns informed intuition into evidence that can guide quality decisions.
Let’s Reframe Soft Skills
We call engineering hard because it has methods, models, standards and most importantly it seems - numbers. The same is true for human capability. Leadership, judgment, resilience, and collaboration are complex, harder to measure directly, and central to performance. These things can be measured.
References
Borsboom, D. (2005). Measuring the mind: Conceptual issues in contemporary psychometrics. Cambridge University Press.
https://doi.org/10.1017/CBO9780511490026
Cook, R. (2000). How complex systems fail. Cognitive Technologies Laboratory, University of Chicago.
(Reprinted in: Cook, R. I. (2002). Cognitive Technologies Laboratory Technical Report.)
Available via patient safety and resilience engineering collections.
Dekker, S. (2014). The field guide to understanding human error (3rd ed.). Ashgate.
DeVellis, R. F., & Thorpe, C. T. (2021). Scale development: Theory and applications (5th ed.). Sage.
https://doi.org/10.4135/9781506341559
Hollnagel, E., Woods, D. D., & Leveson, N. (2006). Resilience engineering: Concepts and precepts. Ashgate.
Kaplan, R. M., & Saccuzzo, D. P. (2017). Psychological testing: Principles, applications, and issues (9th ed.). Cengage Learning.
Klein, G. (1998). Sources of power: How people make decisions. MIT Press.
Ployhart, R. E., & Moliterno, T. P. (2011). Emergence of the human capital resource: A multilevel model. Academy of Management Review, 36(1), 127–150.
https://doi.org/10.5465/amr.2009.0318