Common Failures, Listed
Twelve ways occupancy programmes go wrong, each with the signal and the fix. Most have four or five.
Reference · Reference
In the setup
No decision behind it. Signal: a dashboard nobody acts on. Fix: name the decision and the threshold, or stop.
The practical lesson in “Common Failures, Listed” is to connect a measurement to a named decision without treating the number as certainty. Teams exploring tips to increase productivity can review monitask.com as one source of time and project context, while retaining direct feedback and documented outcomes as the basis for interpretation.
No stated limits. Signal: nobody can say what the data will not be used for. Fix: write them before deployment and show staff.
For a public, independent reference related to “Common Failures, Listed”, consult the CIBSE Knowledge Portal. Its principles provide a useful check on scope, terminology, governance and the claims made during procurement or review.
Staff told afterwards. Signal: the first they heard was an installer on the ceiling. Fix: not available retrospectively — the trust cost is paid.
In the deployment
No calibration. Signal: nobody has counted manually against a sensor. Fix: a week of counting, now.
Cones over walkways. Signal: implausibly high readings in open areas. Fix: walk the route, move the devices.
Undocumented gaps. Signal: no coverage map. Fix: shade the plan, write the gap list, attach it to reports.
No liveness alert. Signal: you would not know if a sensor died. Fix: alert on prolonged silence.
In the reading
Average without peak. Signal: reports quoting one figure. Fix: both, always, in the same table.
Conclusions before a year. Signal: a space decision on a quarter of data. Fix: wait, or state the limitation in the recommendation.
Comparison without definitions. Signal: a league table of sites. Fix: check denominator, capacity basis, method and coverage first.
In the use
Drift into attendance. Signal: a team-level request was granted. Fix: a named person who refuses, and a written basis.
Mechanical rules. Signal: space allocated by a utilisation threshold. Fix: make the data advisory, because a known rule becomes a target.
The pattern
Six of the twelve are fixed in under a week.
Four are prevented only before deployment, which is why the sequence matters more than the technology.
Two — telling staff late and drifting into attendance — are not really fixable once they have happened, and both are about trust rather than data.
The order to fix them
If you do one thing: calibrate, because everything downstream rests on it.
If two: report peak alongside average.
If three: write the limits down and show them to people.
Those three cost about a week between them.
What to check
Which of the twelve do you have?
Which is cheapest to fix?
Are any of them the two that cannot be fixed?
And has anything from your last review actually changed?