How do you measure remote productivity?

Most companies that install tracking software are trying to answer "is this remote team working?" The data they get back answers a much narrower question, and the gap between the two is where the damage happens.

What does the activity percentage actually measure?

The share of time intervals that contained a keypress or a mouse movement. That is it.

Work that scores high: data entry, email, chat, writing code at speed. Work that scores low: reading a specification, a long client call, thinking through an architecture, reviewing someone else's work, whiteboarding, debugging by staring at a log.

The second list contains most of the senior work in any company. Rank a team by activity percentage and you will systematically rank your most experienced people last. That is not a calibration problem to be tuned away; it is what the metric measures.

What happens when you use it as a score?

People optimise for it, because people always optimise for the measure they are judged on. Within a month you get mouse-jiggler software, documents typed into and deleted, and a team that has learned the company cannot tell the difference between work and the appearance of work.

The measurement has then destroyed the thing it was trying to observe, and you cannot undo it by explaining that you did not mean it that way.

So what should you measure?

The last row is the important one. There is no dashboard that evaluates professional judgement, and vendors who imply otherwise are selling comfort.

How should tracked hours be used, then?

• As a denominator. Output per hour is interesting; hours alone are not.
• In aggregate, for planning. "This kind of integration takes 40 hours, not 20" is the most valuable thing time data produces.
• For pricing. Knowing what a retainer actually consumes turns renewal from a negotiation into arithmetic.
• As a welfare signal. Consistent 11-hour days are worth a conversation before they become a resignation.
• Never as a ranking. The moment hours become a league table, they stop being true.

What about teams where output is hard to see?

Support, operations and administrative work often lack a countable deliverable. Two things help, and neither is activity percentage:

• Daily work reports. A few lines on what was done, written against a project. Low friction, and they accumulate into a record of the month that no metric captures.
• Service-level outcomes. Queue depth, response time, resolution rate — measures of the function rather than the person.

Both are in Employee Desk; see workforce management.

A measurement set that works for most remote teams

1. Hours per project, weeklyCaught early, an over-serviced retainer is a conversation. Caught in month five, it is a loss.
2. Task completion against plan, per sprint or monthThe honest signal on whether commitments are realistic.
3. Daily work reports, read not filedThe cheapest insight available, and the first thing to stop working if nobody reads them.
4. Utilisation per person, monthlyFor capacity and for spotting overload. Not for comparison between people.
5. A conversation, monthlyNothing above replaces a manager who knows what good work looks like in this role.

Related