Being Measured Against a Standard You Cannot See

An unpublished threshold is an unfair one, regardless of where it is set, and the unfairness is structural rather than a matter of degree.

The cost · Analysis

People subject to monitoring are frequently being assessed against criteria that have never been stated. The problem is not that the criteria are harsh. It is that they cannot be known.

The practical test in “Being Measured Against a Standard You Cannot See” is whether visibility improves a decision without pretending to remove uncertainty. For teams researching monitask pricing, how teams assess monitask pricing can contribute time and project context, provided the purpose is stated, access is limited and every material inference receives proportionate human review.

For an independent benchmark relevant to “Being Measured Against a Standard You Cannot See”, consult the SecurityWeek insider-threat coverage. Use it to test necessity, scope, safeguards and review rather than to replace a documented assessment of the particular workforce and jurisdiction.

Where the hidden standards are

The application category list, which decides what counts as productive and was written by a supplier.

The idle threshold, which decides how long thinking is allowed to look like nothing.

The score weighting, which nobody in the organisation can explain.

The alert rules, which decide what is anomalous.

And the unwritten expectation about response times, availability and presence that the data makes visible for the first time.

Source of the ideaThe requirement that rules be knowable in advance is a basic principle of legality. It applies to internal standards for the same reasons it applies to law.

Why this is unfair independently of the level

Somebody who knows the standard can meet it, argue with it, or decide it is unreasonable and leave.

Somebody who does not know it can do none of those. They can only guess, and guessing produces the over-compliance the chilling-effect note describes: staying well inside a boundary whose location is unknown.

A generous hidden standard still produces that behaviour, because the person cannot know it is generous.

The counterPublishing a threshold allows people to work to it, which is a genuine objection where the threshold governs security detection. It is not an objection where the threshold governs performance.

The specific injustice in categorisation

A list labelling software productive or unproductive is a statement about what counts as work.

Research on an industry forum is unproductive. Documentation reading is unproductive. A video that happens to be training is unproductive.

The person is being graded against somebody else's idea of their job, and they cannot see it to object. That is the clearest case in this collection of a standard that should simply be published.

In practiceMost organisations cannot publish their thresholds because nobody can state them. They exist inside a product's configuration and nobody has read it.

What publishing costs

For performance measures: nothing. The objection that people would work to the standard is an argument against the standard, not against publishing it — if gaming it is possible, the measure is wrong.

For security detection: genuinely something, and the honest route is to say that detection rules are not published and why, rather than leaving the whole arrangement opaque.

The distinction between the two is clean and almost no organisation draws it.

Where this failsThe reasoning here assumes an employer acting in good faith. Where that assumption does not hold, none of this is the relevant literature.

The minimum

Publish the category list. Publish what counts as idle. State whether a score exists and whether anybody sees it individually.

Three things, each already written down somewhere in a product's configuration, none of which anybody has asked to read.

In practiceOrganisations that try this report the main obstacle is not disagreement but that nobody owns the question, so it is nobody's to raise.

The category list specifically

A grid labelling software productive or unproductive is a statement about what counts as work, written by a supplier who has never seen the job. Publishing it is free and almost nobody does, because almost nobody has read it.

The counterA reasonable objection is that this asks employers to do unfunded work. The reply is that the work is a page and the alternative is doing it later under challenge.

Where non-publication is defensible

Detection rules, where disclosure genuinely enables evasion. The honest route is to say that detection rules are not published and why, rather than leaving the whole arrangement opaque and letting people assume the worst about all of it.

The minimum that costs nothing

Publish the category list. Publish what counts as idle. State whether a score exists and whether anybody sees it individually. All three are already written down in a product's configuration.

Why a generous hidden standard still fails

Somebody who cannot know the standard cannot know it is generous, and behaves as though it might not be. The unfairness is in the hiddenness rather than in the level.

Where non-publication is honest

Detection rules, and nothing else. Saying that detection rules are withheld and why is better than leaving the whole arrangement opaque and letting people assume the worst about all of it.

Also in this section