Writing a Lone-Worker Policy That Actually Gets Used

The policy is eight pages. It lives in a folder labelled “H&S”. It says workers will “maintain regular contact” and “use a suitable device”. Your cartage driver is on a rural run with two dead zones and a load that needs a tail-lift. His system is a text at 5pm that says “all good”. If he does not send it, the office notices tomorrow. That is not a lone-worker policy. That is a document that would look embarrassing after a quiet roadside.

Plenty of small firms in New Zealand and Australia are in this gap. Working alone is normal: first one in, last one out, a solo service call, a night drop, a farm delivery, a roof measure. The policy is written for an inspector. The actual method is hope and a mobile phone.

HSWA 2015 and the model WHS Act do not ban working alone. They expect you to manage the risks of remote or isolated work, including a way to communicate that actually works where the person is. This reflects that general intent. It is not legal advice. The rest is how to write a short policy people will use on a Tuesday.

Write the policy for the work you already do

Start by naming who is alone, when, and doing what. If you skip that, you will write a generic essay. Walk the roster. First opener. Last locker. Drivers between drops. The estimator who measures occupied houses. The cleaner after hours. The on-call tech.

A useful definition is practical, not poetic. A lone worker is anyone who cannot get immediate help from a colleague when something goes wrong. That can be ten minutes in a locked yard at 5:30am. It can be a whole shift. If they are out of earshot and out of sight, treat the stretch as lone work.

Then split the work into three piles:

• May work alone with the agreed controls. Routine driving on known routes. Opening a workshop with lights, a phone that works, and a check-in.
• May work alone only with extra controls. A rural delivery, a night drop, a visit to a site with a history of aggression, a roof measure.
• Must not work alone. Work at height that needs a rescue plan, live electrical work, confined spaces, a task where one person cannot use the control.

If everything is in the first pile, you have not been honest. If everything is in the third, the policy will be ignored. Put real jobs in each list. Use the names your team already uses.

The five parts a usable policy needs

Keep the document to a few pages. If it needs a contents page, it will not ride in the glovebox.

1. Who it covers

Employees, labour hire, regular contractors you send out under your system. New people and visitors who open up with a key. Say who is out of scope, and who owns those people if they are not yours.

2. What is allowed alone

The three piles above, with examples from your work. Review them when you take a new type of job. A firm that starts doing farm tanks has changed the policy, even if nobody edited the file.

3. How people stay in contact

Name the method, not “appropriate communication”. Phone. Radio. Satellite messenger. A check-in app. A call to the office. Then name where it fails. If the gorge has no signal, the method is not a mobile phone. Write the fallback: last known time, a scheduled call before the dead zone, a device that works there, or do not send them alone.

4. Check-ins that have a clock

A check-in without a time is a wish. Set the interval by the risk of that job, not by what is convenient for the office. A quiet workshop open at 6am is not the same as a driver on a remote beach road. Write who receives the check-in. Write what happens at the first miss. Write the second and third steps with minutes attached.

5. Escalation that someone will actually run

This is where most policies die. “Contact next of kin” is not a first step. A usable ladder looks like this: call the worker. Call again. Call the last known site or the previous drop. Call a colleague who can roll. Then emergency services with a last-known location and vehicle. Put names or roles on each step, and a number that is answered. A missed check-in at 7pm cannot wait for the owner to finish dinner unless the owner has agreed to be the person who answers.

What good looks like, and what fails

Good is short, specific, and tested. The driver can say the rule in one breath. The office can run the miss drill without looking it up. You have tried a missed check-in on a quiet afternoon and timed the response. Dead zones are on a map or a route card. High-risk tasks are on the “not alone” list and people have been told no.

Fail is long and hopeful. “Workers shall take reasonable care.” No interval. No named person after hours. A phone app that needs login and signal in a valley. A next-of-kin list that is two years old. A policy that says “buddy system” while the roster still sends one person.

• Good: “Rural tanks: check in every 60 minutes to the duty phone. One miss: call. Two misses: roll a second person and call 111/000 with last GPS.”
• Fail: “Maintain regular contact using a suitable device.”
• Good: the after-hours phone is on the wall and it rings when you test it.
• Fail: the after-hours phone is a mobile that lives in a drawer.
• Good: a new driver gets the rule on day one and a dry run before the first solo trip.
• Fail: they get a PDF in the induction pack and a wave at the gate.

A cartage driver on a rural Friday

Hypothetical. One driver, a tautliner, three farm drops west of town. The middle drop sits in a known dead zone. The old rule was “text when you’re done”. The new rule, written with the driver, is this. He sends a start check from the yard with the three drop names. He calls or pings before he enters the dead zone, and again when he is out. If the office has not heard by the agreed time, they call. If no answer, they call the farm. If the farm has not seen him, they roll another driver and treat it as urgent. He does not start the tail-lift job on the side of a quiet road without that start check. The policy is two pages. The route card has the times. They test the miss on a Tuesday when nobody is hurt. The driver uses it because it is clear and it fits the day.

That is a policy that gets used. Not because it is clever. Because it was written from the run, not from a template.

How to write it with the people who work alone

Do not draft this in the office and email it. Sit with one driver, one opener, one after-hours person. Ask four questions.

1. When are you actually alone, including the short stretches we pretend do not count?
2. Where does the phone die?
3. What job would you not want to do without a second person?
4. If you missed a check-in at 7pm, who do you trust to notice, and how fast?

Write their answers into the piles and the ladder. Then read the draft back in plain words. If they laugh, it is still a brochure. If they argue about the interval, you are in the right conversation. Pick an interval you will staff. An hourly check-in that nobody answers after 5pm is worse than a two-hour check-in that rings a person who is on duty.

Train the receiver as hard as the worker. The lonely part of lone work is often the office end: a missed ping that sits on a desk. If you already run check-ins through Horizon Ops, the policy only works if a missed check-in still rings someone who can act. An alert with no owner is still the old 5pm text.

A one-hour write method

You can get a first version done in a single sitting if you stay specific.

1. List every role that is alone for any part of the week.
2. Sort each task into may / extra controls / must not.
3. For each “may” and “extra”, write the contact method and where it fails.
4. Set check-in intervals and the person who owns the inbox or the phone.
5. Write a three-step miss ladder with minutes, not adjectives.
6. Add emergency information: vehicle, plate, last-known, medical facts the worker has agreed to share.
7. Name who can approve a one-off exception, and that they do it in writing that day.
8. Test a miss this week. Time it. Fix the bit that stalled.
9. Put the two-page version in the glovebox and on the office wall. Archive the eight-page draft.

Review after any incident, any new route, and at a date you will keep. If the roster changes, the policy changed. Edit it then, not at the next visit.

Make it short enough to use

If the team cannot say the rule without the folder, the folder has failed. Name who is alone. Name what they must not do solo. Name the contact method that works in the dead zone. Put a clock on the check-in. Put a person on the miss. Test it once while everyone is still fine. That is a lone-worker policy. The rest is filing.