TL;DR
- ARRCS is an aged care provider with approximately 900 staff across the Northern Territory and South Australia, in sites ranging from ten people to more than a hundred.
- Its engagement survey used to be run by the larger network it belongs to, which set the questions and decided the reporting. ARRCS became more autonomous around two years ago and had to run the survey itself.
- It chose CultureMonkey on cost and on the ability to write its own questions, which mattered as a not-for-profit with a limited budget.
- The old survey took staff up to 25 minutes and people were abandoning it halfway. The new one takes five to ten minutes on a phone.
- A custom question about incident reporting, which the old survey never asked, came back as their highest scoring item.
- Results went out as separate reports by service area, and managers with lower scores were met individually rather than sent a document.
Australian Regional and Remote Community Services, known as ARRCS, is an aged care provider working across the Northern Territory and South Australia. It runs residential aged care homes, home care and community services from bases including Darwin, Alice Springs, Katherine, Tennant Creek and Nhulunbuy, and employs approximately 900 people. Its sites are not the same size as one another: some carry more than a hundred staff, some carry ten.
For years, its engagement survey was not its own. ARRCS sits inside a larger care network, and that network ran the listening programme on its behalf: it set the questions, it determined the reporting, and it sent results back in the shape it chose. Around two years ago ARRCS became more autonomous, and the survey went the same way as everything else it now had to run for itself.
This is what the cycle looks like now that the questions, the timing and the reports belong to them.
Challenges: a survey they did not own, and a workforce that is hard to reach
Someone else's questions, someone else's reports
The old arrangement worked, in the sense that a survey went out and numbers came back. What it did not do was answer anything ARRCS specifically needed to know. The question set was designed for a network, not for aged care in remote Australia, and the reporting arrived as a single package rather than something the organisation could cut by service area or site.
“They just asked for our data, and then they would determine the questions and they would determine the reporting. We didn't have any ability to streamline the reports how we wanted, which departments, which area. We just got the whole lot sent to us however they wanted to send it.”
A workforce spread across remote Australia
Aged care is shift work, and much of ARRCS is not sitting at a desk with a company inbox open. Staff in remote communities often do not check work email at all. Any survey that assumes an email client, a laptop and a quiet twenty minutes is a survey those people will not complete. The spread also makes a single organisation-wide number close to useless on its own, because a site of a hundred and a site of ten are not facing the same conditions.
“Each site would be different in terms of how they would approach that feedback. Some sites have over 100 staff, some sites have 10. So we needed to approach that differently.”
A survey people were abandoning halfway
The previous instrument was long. Long enough that staff started it, ran out of shift, and left it unfinished. A partially completed survey is not a smaller data set, it is an unusable one: the responses skew toward whoever had time, and the drop-off itself distorts everything after the point where people gave up.
“The previous survey, it took too long for staff to complete. They were taking up to 20 to 25 minutes to complete, and sometimes they'd give up halfway through and they wouldn't finish the survey, so the results became invalid.”
Solutions: their own questions, and a survey that fits in a break
Writing the questions that actually mattered to aged care
The first thing ARRCS did with control of the question set was add something the network survey had never asked. Incident and accident reporting is a live obligation in aged care, and the organisation had no read at all on whether frontline staff knew what to do when something happened. So they asked directly. It came back as their highest scoring item, which is a genuinely useful thing to learn about a risk you had been unable to measure.
“We decided to ask a question quite specific, which was, I understand what to do if there is an incident or accident at work... And that was our highest rated score, a favourable score. So that told us something that we didn't really have much of a gauge on previously.”
Removing the step that stopped remote staff
The default survey flow sends a verification code to the respondent's email address before they can begin. For staff who do not open work email, that is not a security step, it is a wall. ARRCS asked for it to be turned off so that people could enter their email address and go straight into the survey on their phone. Peter is clear-eyed about what that trade costs: it means the organisation is not verifying who sits behind each response. For a workforce that would otherwise not have answered at all, ARRCS judged it worth making.
“It made it a lot easier for our people who don't access their emails out in the remote areas to just grab their phone, enter their email address, complete the survey in five, ten minutes and be done.”
A survey short enough to finish
The new cycle takes five to ten minutes, against up to 25 minutes before. That is the difference between something a carer can finish between handover and a break, and something that needs an evening. Peter puts the response rate down to length more than anything else, and the survey ran on phones and browsers alike without anyone needing to be walked through it.
“In terms of the participants and the end user, we had a huge participation rate. It was very quick, it was very easy. People were doing it on their phone or they were doing it online, either or.”
Reports handed to the managers who could act on them
Results were cut by service area rather than sent out as one organisation-wide document. Where engagement came back lower, the People and Culture team met those managers individually instead of leaving them to interpret a deck alone, and the conversation was about what that site could do rather than how it compared. The point of splitting the reporting was to move ownership out to the sites, so that the next cycle has something to measure against.
“Here's the data to the managers. You set up your own teams locally to work out what's best for you, so that when we do the survey again, you can see an improvement.”
Run a survey your frontline staff will actually finish
See how CultureMonkey supports listening programmes for shift-based and remote workforces, with your own questions, mobile-first surveys and reporting managers can use at site level.
Results: what the cycle produced
The cycle reached a workforce that the previous survey had struggled with. Staff completed it in five to ten minutes, most of them on a phone, including people in remote communities who do not open work email. Peter's own summary is that participation was high and that the length of the survey is the main reason.
It also answered something ARRCS had never been able to ask. The custom question on incident and accident reporting came back as their highest scoring item, which is a useful thing to establish about a live obligation that no network-designed question set had covered. And the results went out cut by service area rather than as one document, with the People and Culture team sitting down individually with the managers whose scores came back lower.
The cycle also produced something that is easy to miss. ARRCS now knows what it does not know: the questions it wants to keep, the ones it wants to add, and the shape of report each service area actually reads. That is the ordinary payoff of owning a survey rather than receiving one.
What transfers to another frontline workforce
Own the questions, or you cannot ask about your own risks
The single most valuable item in the ARRCS survey was a question no generic instrument would have contained. If the question set belongs to somebody else, the things specific to your industry and your obligations are precisely the things it will not cover.
Count the minutes, because your people are counting them
A survey that runs 25 minutes on a desk job runs longer on a shift floor, and it competes with a resident who needs attention. Cutting completion time to under ten minutes did more for ARRCS's response rate than any reminder campaign would have.
Meet people on the device they actually have
Email-first assumptions quietly exclude the staff you most need to hear from. ARRCS looked at the steps between a remote carer and a completed survey and removed the one that was stopping them, and made that call knowingly, with the trade-off named rather than ignored.
Give managers the data, not a summary of it
Splitting results by service area and sitting down with the managers whose scores were lower turns a report into a local plan. The organisation-wide number tells you almost nothing about a site of ten people, and the manager of that site is the only person who can change it.
“Some people dig into the detail and get right down to the nitty gritty and do a survey that goes for an hour for people to complete, and then the information just sits in a drawer. This really met our needs, and it gives a good gauge of how our people are feeling out in the workplace.”