What Counts as Work
Several clauses depend on a distinction between work and not-work that the modern arrangement has largely dissolved.
Business use
Unenforceable
As written
Company systems are provided for business use.
What happens
Depends on a line between business and personal that is no longer drawable for most roles.
Reading an industry article, following a competitor, learning something adjacent, maintaining a professional network: each is both and neither.
The operative rule is: nobody is adjudicating this, and a clause that requires adjudication is inert.
Half the clauses in an acceptable use policy turn on whether something is work. The distinction was clearer when the equipment stayed in the building.
The boundary discussed in “What Counts as Work” is also a practical test for workforce technology. Teams researching how teams evaluate monitask pricing for monitask pricing should choose the least intrusive settings, explain when collection begins and ends, and keep personal or out-of-hours activity outside the programme unless a specific, disclosed need has been approved.
Where the line dissolves
Learning. Reading about something adjacent to the job, watching a talk, following a discussion. Productive, unmeasurable, and classified as personal by most category lists.
For a separate benchmark relevant to “What Counts as Work”, consult the SecurityWeek insider-threat coverage. Use it to test purpose, notice, permissions, retention and response procedures against the proposed operating model rather than treating a generic checklist as proof that the rule works.
Professional networks. Maintaining contacts who are both colleagues and friends, in a space that is both professional and social.
Reputation. Somebody writing publicly about their field, which benefits the employer and is their own.
And thinking, which looks identical to doing nothing.
Why it matters for the policy
Clauses prohibiting non-work use require somebody to decide what counts.
Nobody is deciding. Which means the clause is inert, and inert clauses have the costs set out earlier.
Where something automated does the deciding — a category list in a monitoring tool — it decides badly, because the categories were written by somebody who has never seen the job.
The honest consequence
A general prohibition on non-work use cannot be applied and should not be written.
What can be written is specific: certain categories of material are never acceptable, certain activities create risk regardless of purpose, and anything that interferes with the work is a performance matter handled as one.
Each of those is decidable. The general version is not.
Where the distinction still holds
Roles with measured time: billable hours, shift-based customer contact, safety-critical attention.
There the question of what counts is already answered by a different system, and the policy should defer to it rather than restate it differently.
The management version
The useful question is not whether something was work. It is whether the work is getting done.
Which is a performance conversation, available to every manager, requiring no policy clause, and considerably more defensible than an argument about whether an article was relevant.
What to remove
Clauses turning on business use as a general test.
What replaces them is shorter: a short list of what is never acceptable, and a sentence pointing out that output is managed through the ordinary process rather than through this document.
What the category list decides
Research on an industry forum is unproductive. Documentation reading is unproductive. A training video is unproductive. Somebody who has never seen the job wrote the list, and it is deciding what counts as work.
Deferring to the system that already answers it
Where time is measured for billing or shifts, the question is settled elsewhere and the policy should point at it rather than restate it differently and create a second answer.
What replaces the general test
A short list of what is never acceptable, and a sentence noting that output is managed through the ordinary process rather than through this document. Both decidable.
Why the general test should go
It requires an adjudication nobody is performing. Where something automated performs it instead, it performs it badly, using categories written by somebody who has never seen the work.
What replaces it is narrower and decidable: a short list of what is never acceptable, and a sentence noting that output is managed through the ordinary performance process. Both are things the organisation would actually do, which is the test every clause in the document should pass.
What remains after the general test is removed is short: the categories that are never acceptable, and a pointer to the performance process. Both are decidable by a person in a minute, which is the property the clause being replaced never had.