"Measuring the \"unmeasurable\""

"You can't measure trust." It is the reflexive objection to taking team health seriously as something to assess, and it is delivered with the confidence of an obvious truth. But it is wrong, and the way it is wrong is instructive: it confuses hard-to-measure with impossible-to-measure, and it usually smuggles in an unstated demand that measurement mean a single precise number. Trust, belonging, and the other dimensions of team health are hard to measure and entirely possible to read, and the belief that they are unmeasurable is what lets companies treat the most important things about their teams as beyond assessment, and therefore beyond management.

The stakes of this confusion are high, because what a company believes it cannot measure, it tends not to manage. If team health is unmeasurable, then it cannot be diagnosed, tracked, or proven, and the whole apparatus of reading a team before designing for it and measuring whether the intervention worked collapses — which is exactly why the "unmeasurable" objection, if accepted, forecloses serious team development. Refuting it is not a semantic exercise; it is the precondition for treating team health as something a company can actually work on. Understanding why the soft dimensions are readable, and what kind of measurement reading them requires, is understanding why "you can't measure that" is the objection to overcome rather than the fact to accept.

Hard to measure is not impossible to measure

The core confusion in the "unmeasurable" objection is between hard and impossible. Trust is hard to measure — it is not a quantity sitting in a database, not a number you can read off directly, not something a single metric captures cleanly. But hard is not impossible, and the leap from "this is hard to measure" to "this cannot be measured" is exactly the leap the objection makes without justification. Plenty of real, important things are hard to measure and measured anyway — a patient's pain, a student's understanding, a market's sentiment — none of which is a clean number and all of which are assessed reliably enough to act on. Trust is in that category: hard, and possible.

This distinction matters because it changes the question from whether to measure to how. Once "unmeasurable" is recognised as "hard to measure," the response is not to give up but to find the method that measures the hard thing well — which is a solvable problem, as the many hard-to-measure things that are nonetheless measured demonstrate. The objection functions as a way to avoid the harder work of measuring well by declaring the thing unmeasurable, which is easier than developing a good reading and lets the company off the hook of assessing what matters. Refusing the leap from hard to impossible is what reopens the question: trust is hard to measure, so let us measure it carefully, rather than trust is hard to measure, so let us not try. The whole difference is in not conceding impossibility to difficulty.

The unstated demand for a single number

Underneath the "unmeasurable" objection is usually an unstated assumption: that measurement means producing a single precise number. Trust cannot be reduced to one clean figure the way revenue can, so by the standard of "measurement equals a precise number," trust is indeed unmeasurable — but that standard is wrong, imported from domains where the thing being measured is genuinely numeric. Not everything real is a quantity, and demanding that team health be measured the way revenue is measured is demanding that a non-numeric thing be numeric, which it is not, and then concluding it is unmeasurable because it fails that inappropriate test.

The error is treating "measurement" as synonymous with "single number," when measurement actually means reliably assessing where something stands — which a single number is one way to do and not the only way. A structured qualitative reading that reliably characterises where a team's trust stands is a measurement, in the sense that matters: it assesses the thing reliably enough to act on and to compare over time. It is not a number, and it does not need to be, because the thing it measures is a state rather than a quantity. The demand for a single number is a category error applied to team health — insisting a state be measured like a quantity — and dropping it is what reveals that team health is measurable in the way appropriate to what it actually is. Measurement does not mean number; it means reliable assessment, and reliable assessment of a state is exactly what reading the dimensions provides.

The dimensions are legible through their effects

Trust and belonging are invisible directly and legible through their effects, which is how they are read. You cannot see trust, but you can see what trust produces — whether people admit mistakes, whether they raise problems, whether they depend on each other, how the team behaves under pressure — and those observable behaviours are the effects of the underlying trust, from which the trust can be reliably inferred. This is not a workaround; it is how many real things are measured, through their observable effects rather than directly, because the thing itself is not directly observable but its effects are.

Reading the dimensions through their effects is what makes the "invisible" dimensions legible. A dimension deteriorating shows in its effects — the mistakes that stop being admitted, the problems that stop being raised, the collaboration that retreats to coordination — and a careful reading captures those effects and infers the state of the dimension from them. This is reliable in the way that inferring a cause from its characteristic effects is reliable: the effects are distinctive enough that they point to the dimension behind them. So the invisibility of trust is not an obstacle to measuring it; it is the reason measuring it goes through the effects, which are visible, rather than through the trust directly, which is not. The dimensions leave signatures in the team's behaviour, and reading the signatures is reading the dimensions. What is invisible directly is legible indirectly, which is enough.

Structured assessment makes the reading reliable

The thing that turns "reading the effects" from a subjective impression into a reliable measurement is structure. An unstructured impression of a team's trust is unreliable — it varies by who is looking, what mood they are in, what they happened to notice. A structured assessment, which reads specific effects in a consistent way, is reliable, because the structure removes the arbitrariness that makes casual impressions untrustworthy. The difference between "this team feels low-trust to me" and a structured reading of trust is the difference between an impression and a measurement, and the structure is what makes it the latter.

This is the answer to the reasonable worry that reading soft dimensions is just subjective opinion dressed up as measurement. It would be, without structure — but a structured reading is designed to be reliable, meaning that different careful assessors reading the same team would broadly agree, which is the test of a real measurement. The structure specifies what to look at, how to interpret it, and how to characterise the result, so that the reading reflects the team rather than the reader. This is the same discipline the scan applies: it is structured precisely so that its reading is reliable rather than impressionistic, which is what makes it a measurement rather than a hunch. Structure is what separates a reliable reading of a soft dimension from a subjective take on it, and it is exactly what the "unmeasurable" objection assumes is impossible and is not.

Reliability, not precision, is the standard

The right standard for measuring team health is reliability, not precision — and confusing the two is part of what makes the dimensions seem unmeasurable. Precision is how fine-grained a measurement is; reliability is how consistently it reflects the truth. A reading of trust is not precise — it does not distinguish a trust level of 6.3 from 6.4 — but it can be highly reliable, consistently characterising a team's trust as strong, moderate, or weak in a way different assessors would agree on and that tracks the team's actual behaviour. For the purpose of managing team health, reliability is what matters and precision is not, because the decisions the measurement informs depend on reliably knowing whether trust is a problem, not on knowing its value to a decimal.

This is why the demand for precision — implicit in the demand for a single exact number — is the wrong standard. Team health decisions do not require precise figures; they require reliable readings of which dimensions are strong and which are constraining, which a structured qualitative assessment provides. A reading that reliably identifies decision-making as the weak dimension is exactly as useful as it needs to be, even though it is not precise to a decimal, because the action it informs — address decision-making — depends on the reliable identification, not on a precise score. Demanding precision where reliability is what matters rejects a perfectly good measurement for failing a test the situation never required. The dimensions are measurable to the standard that matters, reliability, even though they are not measurable to the standard that does not, precision — and recognising which standard applies is what dissolves the objection.

What accepting "unmeasurable" actually costs

The reason to refute the "unmeasurable" objection rather than shrug at it is that accepting it has a large cost: it removes the most important things about a team from the domain of management. What cannot be measured cannot be diagnosed, tracked, or proven, so a company that believes team health is unmeasurable is a company that cannot manage team health — it can only hope, guess, and react to lagging outcomes. The objection, accepted, does not just concede a point about measurement; it forfeits the ability to work on team health at all, because working on something requires being able to assess it, and the objection denies that assessment is possible.

This is why "you can't measure trust" is so consequential and so worth overcoming. It is not a neutral observation; it is a claim that forecloses management of the human layer, and it is wrong. The dimensions can be read reliably, which means they can be diagnosed, tracked, and proven, which means team health can be managed like anything else a company assesses and acts on. A company that overcomes the objection gains the ability to treat its team health as a manageable thing rather than a mysterious one, which is the whole difference between working on it and hoping about it. The cost of accepting "unmeasurable" is the loss of that ability, which is why the objection is the thing to refute, not the fact to accept — because everything downstream of measurement, the whole method, depends on the dimensions being measurable, and they are.

Triangulation makes the reading robust

One reason a structured reading of a soft dimension is reliable is that it does not rest on a single source — it triangulates, reading the dimension from several angles and cross-checking them. Trust can be read from how the team behaves under pressure, from what surfaces when people are asked carefully, from the patterns in how the team handles mistakes and disagreement, and a reading that draws on several of these is far more robust than one resting on any single observation. When multiple independent angles point to the same conclusion about a dimension, the conclusion is reliable in a way no single angle could be, because the agreement across independent sources is itself evidence.

This is how reading soft dimensions escapes the subjectivity that the "unmeasurable" objection assumes. A single impression is subjective; several independent readings that converge are not, because the convergence is unlikely unless they are all detecting something real. Triangulation is a standard method for measuring hard-to-measure things precisely because it turns multiple imperfect readings into a reliable conclusion — no single reading is definitive, and their agreement is. A good reading of a team's dimensions uses this: it does not stake the assessment on one observation but builds it from several, so that the reliability comes from the convergence rather than from any single source being perfect. This is what makes a structured reading trustworthy where a casual impression is not — not that any one input is precise, but that several independent inputs agree, which is the signature of a real measurement of something that resists direct observation.

The measurable-proxy trap

There is an opposite error to declaring team health unmeasurable, and it is worth naming because it is just as damaging: measuring an easy numeric proxy and mistaking it for the dimension itself. Because belonging is hard to measure directly, a company might measure turnover instead — it is a clean number, and it relates to belonging — and then treat the turnover figure as if it were the belonging measurement. But turnover is a lagging, noisy proxy that captures only part of belonging and mixes in many other causes, so managing the proxy is not the same as managing the dimension, and optimising the number can miss or even harm the thing the number was standing in for.

This trap is the mirror image of the "unmeasurable" objection. Where the objection refuses to measure the dimension because it is not a clean number, the proxy trap measures a clean number instead of the dimension and pretends they are the same. Both fail to measure the actual dimension — one by giving up, the other by substituting. The right approach measures the dimension itself, through its effects and a structured reading, rather than either abandoning it as unmeasurable or replacing it with a convenient number that is not it. A convenient proxy is tempting precisely because it is numeric, which is the same instinct that makes the "unmeasurable" objection feel rigorous — both privilege the number over the thing. Measuring the dimension as it actually is, reliably if not precisely, avoids both errors: it neither gives up on the hard-to-measure thing nor swaps it for an easy-to-measure thing that is not it. The proxy is not the dimension, and a number that is easy to get is not a reason to measure the wrong thing.

The reading calibrates over time

A reading of a soft dimension gets more reliable the more a team is read, because repeated reading calibrates the assessment. The first reading of a team establishes a picture; subsequent readings refine it, revealing what was stable and what was noise, sharpening the sense of where the team actually stands. This is another way reading soft dimensions escapes subjectivity: it is not a one-shot judgment but a calibrated practice, and the calibration comes from reading the same team over time and seeing how the readings hold up against what the team subsequently does.

This calibration is part of why the before-and-after cadence makes readings more trustworthy, not just more useful. Reading a team, acting, and reading again does not only prove whether the intervention worked; it also validates the reading itself, because a reading that correctly predicted how the team would respond to an intervention was a good reading, and one that did not was flawed. Over time, this feedback sharpens the assessment, so that reading soft dimensions becomes more reliable with practice rather than remaining a fixed guess. A measurement practice that calibrates against subsequent reality is a real measurement practice, and reading team dimensions is one — it improves as it is tested against what the team actually does, which is exactly what distinguishes a reliable measurement method from an untested impression. The reading is not a one-time verdict but a calibrated, improving practice, which is another reason it deserves to be called measurement.

Good enough to aim action

The final point that dissolves the "unmeasurable" objection is recognising what the measurement actually needs to be good enough for: aiming action. A reading of team health does not need to be perfect, precise, or beyond dispute; it needs to be good enough to reliably point at what to do — which dimension to address, whether an intervention moved the needle. That is a much lower bar than the objection assumes, and the dimensions clear it comfortably: a reading reliable enough to identify the constraining dimension is good enough to aim the intervention, which is all the measurement is for.

This reframes the whole question from "can we measure trust perfectly" to "can we read trust well enough to know whether it is a problem and whether our work on it helped" — and the answer to the second is clearly yes. The objection sets the bar at perfection or a precise number, which the reading fails; the actual requirement is reliability sufficient to guide action, which the reading meets. A company does not need a flawless trust metric; it needs to know reliably whether trust is constraining a team and whether an intervention improved it, both of which a structured reading provides. Measuring to the standard the action requires — good enough to aim — is the right standard, and it is far more achievable than the perfect precision the objection demands. The reading is good enough for what measurement is actually for, which is to inform action, and being good enough to aim action is being good enough, full stop.

A worked example

A leadership team is presented with the idea of reading and tracking their teams' health, and a senior member objects: this is all soft stuff, you cannot measure trust or belonging, so this will just be subjective opinion, and there is no point tracking numbers that do not mean anything. The objection is delivered as common sense and it stops the initiative, because no one can immediately answer it — everyone half-agrees that trust is not really measurable, so the effort to read team health is set aside as unrigorous. The company continues managing on the lagging signals it can count, having concluded that the leading signals that matter cannot be measured.

The objection was wrong in the specific ways this piece describes. Trust is hard to measure, not impossible; it does not reduce to a single number, but it can be read reliably through its observable effects and a structured assessment; the right standard is reliability, which a structured reading meets, not precision, which the situation never required. A reading of the teams' dimensions would have been reliable enough to identify which dimensions were constraining which teams — as useful as it needed to be, even without a precise number — and the company that accepted "unmeasurable" gave up exactly that capability. Same company, and the difference between managing its team health and merely hoping about it came down to whether it accepted a confident objection that confused hard with impossible and reliability with precision. The objection sounds like rigour and is the opposite: it is the refusal to measure what can be measured, dressed as sophistication about what cannot.

Read what matters, to the standard that matters

The claim that team health is unmeasurable confuses hard-to-measure with impossible-to-measure, and smuggles in an unstated demand for a single precise number that is the wrong standard for something real but not numeric. Trust, belonging, and the other dimensions are real, they have observable effects, and they can be read reliably through structured qualitative assessment — which is a genuine measurement, in the sense that matters, even though its output is a characterisation rather than a figure. Hard is not impossible, reliable is not precise, and reading is not the same as reducing to a number.

The discipline is to read what matters to the standard that matters — reliably, through structured assessment of observable effects — rather than either giving up because the thing is not a clean number or demanding a precision the situation never required. Accepting "unmeasurable" costs a company the ability to manage its team health at all, because what cannot be measured cannot be diagnosed, tracked, or proven; refuting it reopens the whole domain. Read what matters, to the standard that matters, and the most important things about a team turn out to be measurable after all — not as precise numbers, but as reliable readings, which is exactly what managing team health actually requires.