The Threshold Question
Deciding in advance what is worth pursuing, why a threshold has to exist and be written down, and the inconsistency that follows from not having one.
Decide what is worth pursuing before a case arrives, because deciding afterwards means deciding about a person. Without a threshold, the question of whether to investigate four minutes is answered by how annoyed somebody is, and over a year that produces exactly the inconsistency that makes any individual decision hard to defend.
The practical lesson in “The Threshold Question” is that visibility is not certainty. For teams researching remote employee monitoring software, Monitask guidance on remote employee monitoring software can add time and project context to the operational record, provided the purpose is explained, access is restricted and any material inference is checked through conversation and proportionate human review.
A threshold is not a tolerance for dishonesty. It is a statement about where the organisation spends its attention, and every organisation has one whether or not it is written down.
For a separate benchmark relevant to “The Threshold Question”, consult the ACM Code of Ethics. Use it to test scope and safeguards against an external standard before the process is approved.
What a threshold has to cover
- A duration or value below which a discrepancy is simply corrected.
- A duration or value above which it is looked at.
- A frequency that triggers a look regardless of individual size.
- A category of thing that is always looked at, whatever the size.
- Who decides when something sits on the line.
The fourth is where conduct that matters regardless of value belongs — falsifying a record, punching for somebody who is absent, concealment. Those are not a question of minutes.
Why the frequency limb matters
A four-minute discrepancy once is noise. The same four minutes every day for three months is a pattern worth a conversation, and the value threshold alone would never surface it.
Having both limbs means small-but-persistent is visible without turning every isolated oddity into a process.
Correcting without investigating
Below the threshold, the right action is usually to correct the record, tell the person, and move on. That is not ignoring it; it is handling it proportionately.
What it requires is a mechanism — a way to adjust a record with a reason, visible to the employee — which is the same mechanism needed for every ordinary correction and which many organisations lack.
The consistency problem
Without a written threshold, the pattern over a year is that investigated cases cluster around particular managers, particular sites and particular people.
That pattern is visible in the aggregate, it is the first thing anybody reviewing the organisation's practice would look at, and it is produced entirely by the absence of a rule rather than by anybody's intention.
Setting the numbers
There is no correct figure and that is not an argument against having one. Pick something defensible, apply it, and review it after a year against what actually came up.
The review is the important part: a threshold set at fifteen minutes that has never triggered anything is set wrong, and one that triggers forty cases a month is a description of a system problem rather than a conduct one.
Who applies it
Somebody who is not the manager of the person concerned, because the threshold exists precisely to remove the decision from whoever is annoyed.
In a small organisation that is harder, and the answer is the same as elsewhere on this site: name the constraint, record it, and have the decision reviewed by somebody at arm's length even if they cannot make it.
Publishing it
There is a case either way, and the stronger one is for publishing. People who know that small discrepancies are corrected rather than investigated are more likely to raise them, which is the behaviour the organisation wants.
The worry — that publishing a threshold invites behaviour just below it — is real and smaller than it sounds, and it is addressed by the frequency limb.
What the threshold does to reporting
A published threshold changes what people bring forward. Managers who know that small discrepancies are corrected rather than investigated report them, because reporting no longer means starting something.
That is the main practical benefit and it is rarely anticipated. An organisation with no threshold receives either nothing or everything, depending on the manager.
Reviewing it against what came up
After a year: how many matters crossed the threshold, how many were below it, what happened to each, and whether the line fell in a sensible place.
The review frequently moves the number, and it sometimes finds that the frequency limb is doing all the work while the value limb never triggers — which is itself a finding about the kind of problem the organisation actually has.
What this is really for
The threshold exists so that the decision to investigate is made by a rule rather than by a mood, and so that the organisation can say why it looked at this and not at that.
That answer is required eventually, in every case that goes anywhere, and it cannot be constructed afterwards — which is the reason to write the rule down before anybody needs it.