Activity Dashboard Updated Every 15 Seconds

A live activity dashboard is one of the most useful features in workforce software and one of the most frequently misused. The difference comes down to a single question: does seeing this information *now* change a decision you can act on *now*?

For a handful of roles the answer is clearly yes:

• Support and contact centre supervisors rebalancing queues during a volume spike, where a twenty-minute delay means abandoned calls and a missed SLA
• Field dispatchers routing the nearest available technician to an urgent job
• BPO floor managers holding to a contractual staffing commitment that is measured in real time and penalised when missed
• Shift leads in operations covering an unplanned absence before the gap affects throughput
• On-call and incident coordinators confirming who is actually available right now

What these have in common is a decision with a short expiry. The information is operational, the action is immediate, and a next-day report would be useless.

For most other management roles — engineering leads, marketing managers, finance — the honest answer is no. Nothing they will do differs based on whether a team member is typing at 11:42. For those roles a weekly aggregate is more useful and considerably less harmful.

This is worth naming directly, because it is the most common way organisations lose value from real-time tracking, and it is entirely self-inflicted.

A manager with no operational need discovers the live view and starts watching it. Within a fortnight two things happen. The manager's attention shifts from what the team produced to whether the team looked busy. And the team, who are not stupid, work out they are being watched in real time and begin optimising for the indicator — keeping a document in focus, avoiding the reading and thinking that registers as idle, timing breaks around the dashboard rather than around the work.

The tool ends up measuring the performance of work rather than work. The data degrades precisely as reliance on it increases, which is the worst possible combination.

The practical safeguard is access control rather than policy: restrict live views to roles with a documented operational reason for them. Everyone else gets aggregates. This is a configuration decision made once, and it is far more effective than asking managers not to watch a dashboard they can see. The broader principle is covered in our guide to productivity trackers.

Every real-time dashboard has to decide when someone stops counting as active, and that threshold quietly determines whether the whole display is fair or misleading.

Set it too short — three minutes is common and usually wrong for knowledge work — and you mark people idle for reading a specification, thinking through a problem, or being on a call on another device. Set it too long and an unattended machine reads as a full working day, which defeats the purpose entirely.

There is no universally correct value, which is why a single global threshold is a poor design. Sensible starting points differ sharply by role:

• Contact centre and transactional work — short thresholds are appropriate because the work genuinely is continuous input
• Engineering, design and analysis — considerably longer, because sustained thinking is the job and penalising it produces exactly the wrong behaviour
• Client-facing and advisory roles — longest, since much of the work happens in meetings and on calls that generate no device input at all

Track Nexus supports per-team idle thresholds and treats calendar occupancy as active rather than idle, which removes the most common source of unfair readings. Combining input signals with application and website context produces a more defensible boundary than input alone.

The most useful inversion in this category is to stop expecting anyone to watch a screen. A busy supervisor will not, and a supervisor who does has stopped doing their actual job.

Exception-based alerting delivers the operational value of real-time data without the surveillance dynamic, because the system watches and a human is interrupted only when something needs a decision:

• Coverage below threshold — a queue or shift dropping under its committed staffing level
• Scheduled start missed — nobody clocked in for a shift that should be running, surfaced within minutes rather than at end of day
• Sustained idle during committed coverage — genuinely useful when a contractual staffing obligation exists, and not otherwise
• Geofence exceptions — a field team member clocked in from outside the expected site boundary
• Overtime approaching — the point at which a scheduling decision is still possible, rather than a payroll surprise later

Each of these maps to an action. That is the test any alert should pass before it is enabled — an alert nobody acts on is noise that trains people to ignore the ones that matter. For a shared operational display rather than individual dashboards, office TV dashboards put the same data on a wall where the whole team sees the same picture, and productivity analytics covers the retrospective view where most of the analytical value actually sits.