The First Ninety Days

A lone worker programme is judged by its workforce in the first three months, and the judgement is difficult to revise afterwards.

When the first ninety days also depends on reliable work records, field team time tracking can support time, attendance and workload review without being treated as the emergency response itself. Compare any workflow with HSE lone-working guidance and keep alarm ownership, escalation and dispatch responsibilities explicit.

Before anything is issued

The escalation tree tested, with the timings written down.

The assessment done, including the not-alone list.

The withdrawal policy stated by somebody senior, out loud, in a room.

Because the first alarm will test all three, and discovering a gap then is expensive in a way that discovering it in a test is not.

Weeks one to four

Watch the alert volume daily. Thresholds and wearing positions are still settling, and a high accidental rate in the first fortnight is normal and needs acting on rather than tolerating.

Confirm every device is being carried, by asking rather than by checking a report — the report shows a device reporting, which is not the same as a device on a person.

And answer the questions that arrive, which will be about tracking, about whether it affects pay, and about what happens if they press it by accident.

Weeks five to eight

Run a real test at the hour the work happens, unannounced.

Publish the result, including if it is poor.

Fix the worst finding and say that you did.

This is the single most effective thing available in the period, because it demonstrates the chain rather than asserting it.

Weeks nine to twelve

Classify the alerts: genuine, precautionary, accidental, technical.

Retune what needs retuning.

Review the check-in intervals against what people actually experienced.

And ask the first cohort of workers what they would change, which produces better changes than any review by managers.

What to watch for

Devices not being carried, which shows as alarms from vehicles or positions that never move with the person.

Check-in compliance falling after the first month, which is the normal decay and needs a prompt rather than a reprimand.

Silence: no near misses, no withdrawals, no disputes. Silence at this stage means the reporting routes are not believed, not that nothing is happening.

And managers using the system for something else, which if it occurs once will be known everywhere within a week.

The first real alarm

However it goes, debrief it thoroughly and tell everybody what happened.

If it went well, that is the most persuasive thing the programme will ever produce.

If it went badly, saying so and fixing it publicly is the second most persuasive.

Concealing a poor response is the one option that cannot be recovered from, because it will be known regardless.

The review at week twelve

Response times from the tests, before and after.

Alert classification and the false rate.

Check-in compliance and missed check-in response times.

Near misses and withdrawals, which should both be non-zero.

And one direct question to workers: has anything you were told at the start turned out not to be true?

Act on any yes immediately, because that answer determines whether the arrangement is believed for the rest of its life.

Who runs the period

Somebody named, with time, who is not also the person answering every alarm.

Because the first ninety days generate more questions, more accidental activations and more small fixes than any subsequent quarter, and an owner doing both will do neither.

And somebody senior visible at the start, since the withdrawal policy is only believed if it comes from a level that can enforce it.

Setting expectations in advance

Tell people the first weeks will produce false alarms and that this is expected.

Tell them the arrangement will change in response to what they report, and then change something visibly.

And tell them the response times will be published, including when they are poor.

All three lower the cost of the inevitable early problems, which otherwise get read as evidence that the programme was not thought through.

What success looks like at ninety days

Response times measured and improving.

False alarms classified and falling.

At least one withdrawal, at least one near miss reported, at least one thing changed because somebody said so.

And the workers able to describe what happens after they press the button.

None of these is about the equipment, which is the point the whole period is making.