Say "screen tracking" in a standup and watch the room tense up. For most engineers the phrase means one thing: a manager who doesn't trust them, watching a mouse-jiggle counter tick down. That reputation is earned — a lot of monitoring software is built to catch people, not to help them.
But there is a legitimate reason teams want time data. Estimates drift. Invoices to clients need hours behind them. Someone quietly carries three projects while another person is blocked for two days and nobody notices. The problem is almost never tracking — it's tracking done secretly, applied unevenly, and used as a weapon.
This post is about the other version: how to introduce time and screen tracking so that the people being tracked actually want it on. The mechanics matter less than the contract you set around them, so we'll cover both.
Why monitoring usually backfires
When tracking creates resentment, it's almost always one of these five failures — not the feature itself:
- It's covert. Software installed without a clear conversation reads as spying, full stop. Trust never recovers from finding out after the fact.
- Visibility is asymmetric. The manager sees everything; the employee sees nothing about themselves. That imbalance is the surveillance.
- It runs all the time. A tool that watches you during lunch, during a bathroom break, during a 1:1 is measuring presence, not work.
- The data punishes. The first time a screenshot shows up in a performance review as a "gotcha," every timer on the team becomes an enemy.
- It measures the wrong thing. Keystrokes-per-minute and active-window time reward looking busy. Good engineering often looks like staring at a wall.
Notice that four of the five are policy choices, not technical limits. You can buy the most invasive tool on the market and use it humanely, or buy a gentle one and wreck a team with it. The tool sets the ceiling; you set the norm.
The trust contract
Before you turn anything on, write down — and share — the answers to these questions. This is the single highest-leverage thing you can do.
- Who can see the data? Name the roles. "Admins only" is very different from "your whole reporting chain."
- What is captured, and when? Be specific: presence, timer state, periodic screenshots only while a task timer runs — not a 24/7 feed.
- What is it used for? Estimation calibration and client billing are legitimate. Ranking people against each other is how you lose them.
- What is it never used for? Say the quiet part out loud. "We will not use screenshots in performance reviews" is a promise worth making in writing.
- Can people see their own data? They must. Anything a manager can see about an engineer, that engineer should see about themselves first.
A team that has read and agreed to those five answers reacts to a task timer the way they react to a git commit — as a normal part of the work, not a camera in the corner.
Opt-in beats mandatory
The instinct is to make tracking mandatory and universal on day one. Resist it. A better rollout looks like this:
| Phase | What you do | What it buys you |
|---|---|---|
| Week 1 | You (the lead) turn it on for yourself only | You feel exactly what you're asking of others |
| Week 2–3 | Volunteers opt in; data stays private to each person | Proof it isn't a gotcha machine |
| Week 4+ | Team agrees on a shared norm for billable/estimated work | Buy-in instead of a mandate |
By the time it's a team norm, everyone has already seen their own numbers and knows nothing jumped out to bite them. Mandating it later, if you still need to, is a much smaller ask.
What "humane" looks like in the tooling
This is exactly the line ETracking is built on, so it's a concrete example of the settings that matter:
- The agent is opt-in and visible. It's a desktop app for Windows, Mac and Linux that the person installs and can see running — not a hidden background service.
- Screenshots are tied to the timer. Periodic screen captures happen only while a task timer is actively running. Stop the timer and nothing is captured — lunch, breaks and personal time are simply off the record by design.
- Visibility is admin-only, and the person sees their own. The raw stream isn't broadcast to the whole org; it's scoped, and the employee has their own view.
- Hours attach to the actual work. Time lands on the issue or board card it belongs to, not on a separate spreadsheet. The data point is "this task took six hours," which is useful, instead of "this person was 73% active," which isn't.
The payoff is that the same data serving your invoices also calibrates your estimates. After a few sprints you can line up estimated versus actual time in Reports and watch your team's "small / normal / scary" instinct get sharper — which is the version of tracking engineers will actually defend, because it makes their estimates less of a guess.
Rolling it out: a short checklist
- Write the trust contract first, share it, and invite pushback before anything is installed.
- Turn it on for yourself for a full week before asking anyone else.
- Keep individual data private during the trial — no dashboards ranking people.
- Point the data at estimation and billing, and say plainly what it's not for.
- Review the norm after a month and drop anything that isn't earning its keep.
The real test
Healthy tracking passes one test: if you announced the exact policy to the whole team in an all-hands, would anyone be surprised or upset? If the answer is no — because visibility is symmetric, capture is scoped to real work, and the data is never a weapon — then the timer stops being surveillance and starts being what it should have been all along: a shared, honest record of where the work went.
If you want the humane version of this baked into the tooling instead of bolted on, ETracking is your.team's flagship, and it's priced as part of the platform rather than a per-seat spy tax. Start free at app.your.team/signup and turn it on for yourself first.