The question every employer asks and few say aloud. Why the obvious answer backfires, the three things worth measuring, and when monitoring is right.
12 September 2026 · 10 min read
Every employer with people working from home has asked this. Very few say it out loud, because it sounds like distrust, and the software industry has been happy to sell an answer that confirms the fear rather than resolving it.
It is worth taking the question seriously, because underneath it is usually a reasonable concern that has been badly framed.
When an employer says "I don't know if they're working", they almost never mean they suspect fraud. Push on it and the real worry is one of these:
Those are three different problems. Only the third has anything to do with watching a screen, and even then a screenshot is a poor way to answer it.
The instinct is to install monitoring, see activity, and feel better. It does work, in the narrow sense that you will see activity.
What you also get:
The measurement changes the behaviour, and not usefully. People who know activity is scored produce activity. Mouse movement is trivially easy to generate and impossible to distinguish from thinking. You end up with a number that is high for your least valuable staff and low for the person who spent the afternoon on a hard problem.
You learn things you did not want to know. A screenshot captures whatever was on the screen. Their banking. A medical appointment. A conversation about a sick parent. You now hold that, and you have to store it, secure it, and eventually explain it.
It costs you the people you least want to lose. The staff who mind most are usually the ones with options. Nobody good leaves over a clock-in. Plenty leave over a screenshot gallery.
None of that means monitoring is never appropriate. It means reaching for it first, as a general answer to a vague unease, tends to produce a worse outcome than the unease.
Are the hours real? This one is legitimate and easily answered. A clock record from wherever they are, against the roster you published. If someone is rostered and has not started, you want to know at 9:15am, not at payroll. This is the part where you are entitled to certainty, because you are paying for it.
Is the work happening? Tasks moving, work shipped, questions answered, customers handled. For most roles this is the entire answer, and it needs no monitoring at all. If you cannot tell whether a job is progressing without watching someone's screen, the gap is usually in how the work is defined, not in your visibility of it.
Is anyone struggling? This is the signal that genuinely matters remotely and the one nobody markets, because it does not sound like control. Someone working every evening. Someone who stopped clocking off. Someone who has gone quiet in a channel they used to lead. Remote teams lose people slowly and invisibly, and by the time it is obvious they have already decided.
The third one is where the money is. A remote employee doing 60% of what you hoped is a management conversation. A remote employee quietly burning out is a resignation, a recruitment cost, and six months of lost knowledge.
There are cases where it is genuinely the answer, and it is worth naming them rather than pretending otherwise:
What those have in common is a concrete purpose. "I want to see what people are doing" is not one. The narrower the purpose, the more defensible the monitoring, and the easier it is to explain to the person it applies to.
If you do monitor, in Australia: disclose it, in writing, before it starts. Requirements differ by state — NSW requires 14 days notice for computer surveillance — and covert monitoring is the thing that is actually prohibited. We have written a longer piece on the law.
Before switching anything on, ask: could I describe exactly what this captures to the person it captures, in plain terms, without softening it?
If the answer is no, the problem is not how to word the announcement. It is the arrangement.
That test has a useful second property. Monitoring you would happily describe tends to be monitoring with a real purpose, proportionate to it — which is also the monitoring that survives scrutiny.
WorkAndGo does have monitoring. It is off by default, enabled per person rather than across a workforce, disclosed before anything is captured, and on a personal device it only runs between clock-in and clock-out. Owners, managers and contractors are never monitored regardless of the setting.
But the parts we think matter more for a remote team are the ones that are not monitoring at all: clock records you can trust from anywhere, rosters people can see without watching a chat thread, hours worked against hours rostered, and a watch-list that surfaces the person working every night before they hand in their notice.
The employee sees their own productivity view too. That is not a courtesy feature. A measure only the employer can see is a measure the employee cannot argue with, and any number about a person that they cannot inspect will eventually be wrong about someone in a way nobody catches.
If you cannot tell whether remote work is happening, more surveillance is rarely the fix. Clear expectations, honest hours and a way to spot someone struggling will tell you more than a screenshot gallery, cost you less, and keep the people you would rather not replace.
Monitor when you have a reason you could say out loud. Otherwise, measure the work.