Ten years ago we started keeping simple records on every group we knew about: when it started, who led it, and whether it ever produced a second group.
We expected the data to show that group size mattered, or that certain neighborhoods were simply more receptive than others. Neither held up.
The single strongest predictor turned out to be whether the original leader had a coach who met with them at least monthly during the first six months. Groups with that kind of coaching multiplied at more than three times the rate of groups without it, regardless of size, location, or leader experience.
It's a less exciting finding than we hoped for. There's no clever technique buried in it, just a very unglamorous case for consistent, ordinary follow-up.
