How to Track Work-from-Home Productivity Without Being Creepy

The line between productivity tracking and surveillance is the most important thing to get right with WFH employees. Too little measurement, and managers feel blind; too much, and employees feel watched. Here is the framework for measuring WFH productivity that respects the trust contract.

The 3 things worth measuring

1. Active work hours (not "logged-in hours")

Logged-in time includes the 90 minutes the employee spent making lunch and watching news. Active work hours = time actually using work tools (typing, clicking, in meetings, scrolling work apps). This is the true work-time signal. See: productivity attendance explained.

2. Output delivered

For developers — commits, tickets resolved, PRs reviewed. For designers — assets delivered, iterations completed. For support — tickets closed, response times. For sales — calls made, deals progressed. The real productivity signal is always output, not hours.

3. Response time within agreed windows

Is the employee reachable during their agreed online hours? Slow responses to legitimate work messages = pattern to address. 4am responses to non-urgent messages = also a pattern to address (burnout signal).

The 3 things NOT worth measuring

1. Continuous screen recording

Counterproductive. Creates surveillance theatre, hurts trust, generates terabytes of useless data, and rarely catches anything not already visible in output and active-hours data.

2. Webcam monitoring

Crosses the line for almost everyone. Acceptable only in specific high-security contexts (finance, defence) and even there with explicit consent.

3. Keystroke logging

Tracks the content of what is typed, not just the fact that input is happening. Privacy violation. The fact of input (which is what you want) is captured by active-hours measurement without the content.

The trust contract

For productivity tracking to work, both sides must understand:

• What is measured (active hours, output, response times)
• What is NOT measured (no keystroke logging, no screen recording, no webcam)
• Who sees the data (employee + their manager, not peers)
• How long it is retained (typically 6-12 months)
• What triggers human review (consistent below-threshold weeks, not single bad days)

Put this in writing in the WFH policy. Have employees sign it. This is the legal + trust foundation.

The 4 productivity patterns to actually act on

1. Sustained drop in active hours

One bad week — ignore. Four bad weeks — conversation about workload or motivation. This is the right reaction, not "your active hours dropped 12% on Tuesday, explain."

2. Output disconnected from hours

High active hours but low output — coaching opportunity (work patterns might be inefficient). Low active hours but high output — celebrate (this is exactly what you want from senior staff).

3. Always-on signals

Active at midnight, weekends, holidays — burnout risk. Worth an HR conversation, not a celebration.

4. Response gaps during agreed hours

If "online 10am-6pm" but responses take 4 hours — pattern. Either the policy is wrong (they actually work different hours and it's fine) or the employee is non-responsive (which is the actual problem).

What the WFH policy should say about tracking

• The HR system records active work hours and application categories
• It does not record keystrokes, does not screen-record, does not monitor webcam
• Data is visible to the employee and their direct manager only
• Data is retained for [X] months
• Data is used for productivity feedback, not for surveillance
• Employees can request to see their own data at any time

The Zaffre HRM fit

Zaffre HRM's desktop monitoring agent tracks active hours and application categories without keystroke or screen-recording overreach. Productivity-based attendance derives presence from real work activity, with full transparency to the employee. See also: remote management playbook.

Book a demo to see the monitoring agent + productivity attendance live.