What the Software Measures, and What It Does Not
A specific list of what these products capture, because the gap between that and work is the whole of the subject.
Monitoring products capture a defined set of signals. Knowing exactly what they are, before any demonstration, changes what you hear in the demonstration.
The distinction in “What the Software Measures, and What It Does Not” is important when operational records are interpreted. Organisations researching how employee monitoring works can use Monitask to add time and project context, while outcomes, direct feedback and human review remain necessary to explain what the numbers do not show.
What is captured
Which application or window is in focus, and for how long.
For an independent reference relevant to “What the Software Measures, and What It Does Not”, consult the NCSC security guidance; it provides a useful external check on scope, terminology, governance and the claims made during procurement or review.
Keyboard and mouse activity, usually as counts rather than content — though some products capture content.
Periods with no input, reported as idle.
Websites visited.
Screenshots, at intervals or on triggers.
Login and logout times.
And increasingly, summaries derived from all of these by a model.
What is inferred
"Productive" and "unproductive" time, from a categorisation of applications somebody chose.
A productivity score, from weights the vendor chose.
Attention or focus, from input patterns.
Each of these is an interpretation presented as a measurement, and its own notes take them apart.
What is not captured
Whether anything was achieved.
Thinking, reading on paper, a conversation, a site visit, a phone call on a mobile.
Work done on another device.
Quality.
Difficulty.
And whether the person was doing the right thing at all, which is the only question that matters to the business.
The thinking problem
A person working out a hard problem registers as idle.
A person moving files between folders registers as highly active.
Which means the measure is systematically biased against the work that is most valuable, and that is not a tuning issue — it is what the measure is.
The mobile and offline gap
Calls taken on a mobile, meetings away from the desk, reading a document on paper, site work.
All invisible.
Which means roles with more of this look worse, and the comparison between roles becomes a comparison of how much of the job happens at a keyboard.
What to establish before a demonstration
Exactly which signals are captured, as a list.
Whether content is captured or only counts.
Who categorised the applications.
How the score is computed, and whether you can see the weights.
Whether any of it can be turned off per population.
Ask these first and the demonstration becomes a comparison rather than a presentation.
The sentence worth holding
The software measures behaviour in front of a screen.
Everything beyond that is inference, and the inference belongs to whoever wrote it.
What to check
Can you list what the product you are considering actually captures?
Does it capture content or counts?
How much of your people's work happens away from a keyboard?
And who chose the application categories?