With a normal focus group, everybody is invited a week or so in advance, a time and place are set, and the participants turn up for a fixed period. That's all very well, but sometimes research needs to be more spontaneous.
In 1988 I visited Cooktown, some 200 km north of Cairns. This is one of Australia's most isolated towns. The roads (at the time) were terrible - it took about 8 hours by 4-wheel-drive to get to Cairns. There was a ferry service a few times a week, and one flight a day in a small, poorly maintained aircraft. I was there because Cooktown was to get a radio transmitter. Several networks were available, and the research task was to find out which would be most popular.
I'd never been to Cooktown before. It turned out to be one of the most laid-back towns ever. I'd organized a group discussion by phone, for a Saturday morning, and hired the only public meeting place I could find in Cooktown - the Country Women's Institute. This turned out to be a small corrugated-iron hall in the main street. The helpful and friendly Country Women were there too, serving tea and coffee, and baking scones. It was hot and sticky, and the door to the street was left open to cool the building. But the smell of scones wafted out, and passers-by began dropping in to see what was happening.
Seeing a group discussion taking place, they joined in. Some had a few words to say, then left. Others came and stayed. Maybe 30 people came and went. Of course, they mostly knew each other. It was nothing like any focus group I'd ever run before, but it was a lot of fun. It also gave me a very clear picture of which program they preferred, and why. Every now and again there were thunderous showers; all discussion had to stop, because of the noise of rain drumming on the iron roof. To escape the showers, more people ducked in. We went through a lot of scones that morning.
After that unplanned but successful experience, we organized some more rolling focus groups. One, for example, was held at a farmers' show at Paskeville, in South Australia. Paskeville normally has a population of approximately zero (give or take a few cattle), but for a few days every second year, it becomes home to 20,000 people.
We were doing a project for rural radio - basically finding out what listeners thought of the programs. So I borrowed a room in the back of one of the sheds,and stationed an interviewer at the front to lure people in. She asked passers-by a few screening questions, and if they turned out to be listeners to the station we were doing the project for, she sent them to the back room, where I conducted the rolling group discussion. It turned out very much like Cooktown: some people stayed only a few minutes, expressing themselves succinctly, while others hung around for hours. Sometimes there were a dozen people in the room, at other times only one or two. We were there all day, and the tape recorder was running the whole time. Everybody also got a simple questionnaire to fill in, summarizing their background and radio listening habits.
One of the main benefits of the rolling format is that ideas brought up by early participants can be checked with later participants, e.g. "Some people have said so-and-so - what's your reaction to that?"
This format, because of its interactivity, produces much more insight than the more traditional alternatives of handing people questionnaires to fill in, or interviewing passers-by one at a time.