How Hubstaff really calculates your activity percentage

The 10-minute interval, second by second

Hubstaff splits your tracked time into 10-minute intervals and, within each one, checks for mouse or keyboard input. The activity percentage for that interval is the share of the sampled moments that contained any input at all. Every interval is scored the same way, and the daily figure you see on the dashboard is simply the average of all of them.

Two things follow from that design. First, the score is about coverage, not intensity: a second with one keystroke counts exactly the same as a second with thirty. Second, because the day is an average of independent blocks, one very low interval — a long phone call, a walk to think through a problem — pulls the whole day down more than most people expect, and no amount of frantic typing later in the day fully cancels it out.

Keyboard and mouse: what actually registers

Only two kinds of signal move the number: the keyboard and the mouse. Keystrokes count, mouse movement counts, clicks count, and scrolling counts. Hubstaff records that these events happened, not their content — it is measuring that input occurred, not logging the words you typed. Everything else your body and brain do during real work is invisible to it.

This is why the carrier of the input barely matters and the shape of it matters a lot. A mouse that drifts in one dead-straight line and a person nudging the trackpad while reading both produce "activity", but a tracker that looks at patterns treats them very differently. The mechanics of that are covered in the detection write-up in the related links.

• Keystrokes — any key, regardless of what you type.
• Mouse movement, clicks and scrolling.
• Not the content of your input, and not what is on screen.

What the percentage does not measure

Input is the only signal, so a shipped feature and an aimless scroll look identical to the tracker. The number carries no information about quality, output or value — it cannot, because none of those things touch the keyboard or the mouse in a way it can read.

That gap is the single most important thing to understand about the metric. Managers who treat the percentage as a productivity score are reading a presence signal as if it were a results signal, and the two come apart constantly for anyone whose work involves thinking rather than typing.

• Not the usefulness of what you did.
• Not code, words or tickets delivered.
• Not thinking, reading, listening or planning.

Why honest work can score low

Reading a spec, watching a recorded call, reasoning through a bug or sitting in a meeting generates almost no input. A research-heavy day lands far below a data-entry day even though both are entirely real work — often the harder, more valuable day is the one that scores lower.

The effect is structural, not a sign of slacking. Designers reviewing mockups, engineers reading logs, writers staring at a paragraph, and anyone on back-to-back calls all produce naturally low numbers. If a role is judged on the percentage alone, the people doing the most thinking are penalised the most.

What a healthy number looks like by role

There is no universal "good" percentage, because the honest baseline depends entirely on how much of the job runs through the keyboard and mouse. A useful way to read your own number is to compare it against your own quiet weeks, not against a colleague in a different kind of role.

As rough context: data entry, support and QA tend to sit high because the work is continuous input; development, design and analysis sit in the middle because thinking and reading break the input up; and research, strategy or meeting-heavy days sit lower still. All of these can be a full, honest day of work.

• Continuous-input roles (support, data entry, QA) — naturally high.
• Build-and-think roles (dev, design, analysis) — naturally mid.
• Read-and-meet days — naturally low, and that is fine.

Why a flat 100% is a red flag

Nobody produces input in every sampled moment of every interval — real work has pauses built into it. A report pinned at or near 100% for hours, day after day, does not read as a heroic worker; it reads as automation, because no human hand behaves that evenly.

A slightly uneven line in the 60–85% range, dipping for calls and reading and climbing during hands-on work, is what a person actually looks like. Consistency at the top of the scale is more suspicious than a mid-range number, which is exactly backwards from how most people assume the metric is judged.

How to read your own report

Open your own timeline before anyone else does. Look at the shape across the day rather than the single daily number: where the dips are, whether they line up with meetings and reading, and whether any interval is oddly flat. That context is what turns a bare percentage back into a story about a real day.

If a genuinely low day needs explaining, the interval view is your evidence — the dip that matches a two-hour design review is self-explanatory once someone can see it. Understanding the metric is what lets you talk about your work on your terms instead of apologising for a number you did not understand.

Where Reaction fits

Reaction sits in one narrow place: the stretches when you step away from the keyboard but the session should stay alive — the coffee, the stretch, the school run. It keeps a natural, human-like level of presence within a range you set, using varied cursor motion, a living keyboard rhythm, occasional scrolling and window switches, and it starts only when the computer is genuinely idle and steps aside the moment you touch the mouse.

It does not, and cannot, make low-input work look like high-input work while you are actually at the desk doing it — that is a limitation of the number, not a feature to sell. What it does is stop a normal break from reading as an absence. Use it within your agreements; see the acceptable-use note below.