You Can’t Measure What You Can’t Define – Or Can You? (part 3 Of 5)

Rethinking Performance – Part Three

In Part 3 of Rethinking Performance, written by Camille Rabier, Consultant at 21st Century, the series shifts from diagnosing the gaps in performance systems to exploring how they can be redesigned. The article examines how behavioural expectations can be translated into structured, usable criteria that make performance evaluation more consistent, transparent and defensible.

Making behaviour evaluable

Performance is shaped by both what is delivered and how it is achieved, yet only outcomes have been consistently integrated into performance systems. The question, therefore, is how behaviour can be translated into something usable within performance systems so that performance can be represented more accurately and fairly.

From behaviour to criteria: the translation problem

We introduced earlier a model that enables the translation of values - which define how work is expected to be done through behaviour:

Value =  Behaviour principle =  Observable behaviour = Everyday action

This is not just a model for clarity, but a design mechanism for evaluation. It operationalises values into behaviour principles, which are then translated into observable behaviours and everyday actions that can be identified in practice. It improves the visibility of behaviour by making it explicit and observable. Although it does not eliminate subjectivity, it provides a structure within which judgement can be applied more consistently.

What makes behaviour evaluable

The translation model makes behaviour more visible, but visibility alone is not enough for evaluation. For behaviour to be used within performance systems, it must be expressed in a way that can be recognised in practice and applied consistently across individuals. This requires a clear distinction between behavioural intent and observable expression.

Behaviour principles define what is expected, while observable behaviours define what that looks like in practice. This means moving away from broad behavioural labels such as 'collaboration' or 'ownership', which operate at the level of principles rather than observable behaviour, and defining what those behaviours actually look like in day-to-day work. For example, rather than assessing 'collaboration', evaluation requires clarity on what can be observed, such as how individuals share information, involve others in decisions, or respond to input.

These observable behaviours form the basis for evaluation, but they are not yet sufficient on their own.

From behaviours to indicators

If behaviour can be expressed in a way that is recognisable in practice, the next question is how it can be evaluated more consistently. Behaviour cannot be evaluated directly in its abstract form. Instead, it is inferred through observable indicators that reflect how work is carried out in practice. Indicators do not describe behaviour itself, but they provide a structured way of assessing how consistently and effectively it is demonstrated.

This introduces an additional layer within the model:

Value   à

Behaviour principle   à

Observable behaviour       à  Behavioural indicators

·       Everyday action

Within this structure, indicators now sit directly beneath observable behaviours, as they define how these behaviours can be recognised and assessed. Alongside them, everyday actions illustrate how the same behaviour may be expressed in different roles and contexts. As such, rather than assessing collaboration as a single 'observable behaviour', evaluation should focus on its indicators, such as how consistently individuals share information, how early they involve others in decisions, or how effectively they respond to input. These indicators make behaviour more tangible. They provide reference points that can be observed, discussed, and compared across individuals, reducing ambiguity without removing interpretation.

However, making behaviour more assessable through indicators does not make it fully measurable, nor does it remove subjectivity. What changes is not the presence of judgement, but how it is applied.

The limits of indicators

Indicators do help make behaviour more assessable by providing observable signals of how it is expressed in practice and across individuals, and by translating abstract behavioural expectations into tangible patterns. However, they also introduce a set of limitations, particularly when over-relied on:

• False sense of precision: when indicators are reduced to checklists, they can create the impression that behaviour is assessed objectively, stripping it of context and reducing it to isolated actions in which frequency is mistaken for effectiveness.
• Simplification of behaviour: indicators are simplifications and cannot fully capture nuance, intent, or situational constraints. The same indicator may reflect sound judgement in one context and poor judgement in another.
• Bias towards visibility: indicators tend to privilege behaviours that can be directly observed and articulated, while overlooking less visible aspects such as judgement, intent, or informal contributions.
• Risk of performative behaviour: indicators can encourage individuals to demonstrate visible alignment rather than underlying effectiveness, turning behaviour into something that is signalled rather than exercised with sound judgement.

As such, indicators should not be treated as definitive measures of behaviour, but as reference points that guide more consistent and informed evaluation. They standardise how behaviour is described, but not how it is judged, leaving an essential role for managerial judgement.

Implications for practice

For organisations, this shifts the focus from simply defining behaviours to structuring how they are evaluated in practice. Indicators should be applied with careful attention to context, rather than used as scoring mechanisms. This places greater emphasis on managerial capability as a critical component of the system. The effectiveness of any behavioural framework depends not only on how clearly behaviours are defined, but on how consistently they are interpreted.

In practice, this means performance systems must be designed to support both structure and interpretation, providing clear behavioural anchors while still allowing space for context to be considered. Without the development of evaluative judgement, even well-defined frameworks risk reinforcing inconsistency rather than reducing it. This may require clearer guidance on how indicators are applied, as well as more deliberate calibration across managers.

Conclusion

As organisations seek to integrate behaviour into performance evaluation, the challenge is not only defining what good behaviour looks like but ensuring it can be assessed consistently and meaningfully. While indicators improve how behaviour is described and discussed, they remain only partial representations and require interpretation in context. The role of any operationalisation model is therefore not to eliminate judgement, but to anchor it.

Ultimately, the aim is not to make behaviour objective, but to make its assessment more disciplined, transparent, and comparable. It is in this balance between structure and judgement that the potential for more credible and effective performance systems lies.

This need for interpretation places managerial judgement at the centre of behavioural evaluation, shifting the focus from defining behaviour to ensuring it is applied consistently in practice. The next article explores how this can be achieved.

Total Words: 1088