When we started the COVID-19 coaching groups, Daniel Yang, Todd Milby, and I were hoping we’d have somewhere between fifty churches join in. As we continued to talk with leaders of other networks, we began to realize the number might be a little bigger, but our expectations didn’t massively shift.

Early on, we made a strategic choice: We decided not to brand the work under any particular organization. If we believed we were better together, it needed to function and be communicated as a collaboration. It was a trade-off: No single organization got the credit, but it meant more people (Mass) might be part of it. So from day one, these coaching groups were a combined effort of Catapult, the SEND Institute, the NewThing Network, Christ Together—a collection of leaders who were already working together in one way or another, with each of us bringing some of our coaches to the table.

We made another strategic choice: We would not charge for participation in a coaching group. Every coach gave their time away. That made it accessible to anyone in the world with internet access. We offered groups on every day of the week (except Saturday and Sunday), including some with early morning slots, to account for global time zones. We asked coaches to help lead at least two groups. This was our radical minimum. Not surprisingly, the first and fast followers of our coaches quickly turned into a hotbed with a center of gravity. And as more groups developed, more hotbeds started to emerge and multiply, and they started to share best practices. Very quickly, a tribe was developing, and it had all the relational thickness that Alan Hirsch calls Communitas—the friendship, community, and relational bonds formed in the fires of being on mission together.

The groups themselves happened on Zoom, so if we had five hundred people register for one time slot, the only thing keeping us from scaling was the number of coaches, because each virtual breakout room required one coach to every six to eight participants.

We started Week 1 with a few hundred churches participating in groups, and we decided to lower another barrier that might hurt scalability: We didn’t close registration after Week 1. As it turns out, the experience of the first week was sticky and sneezable. We grew to 1,094 churches at the end of Week 2.

But it didn’t stop after Week 2. Word got out what was happening; not only was it helping people stabilize and re-normalize in the midst of the crisis, but it was helping leaders mobilize their people into mission. We had more and more people clamoring to get into groups.

Like Rent the Runway, we had a problem on our hands: a lot of new churches wanted to get into coaching groups, but our infrastructure was starting to creak. We were running out of coaches, the IT support needed to sustain the team maxed out, and the logistics of running that many groups, with that many people, were redlining the effort. At this point, the way we were scaling the innovation was using the Resourced model. This was only possible because everyone was donating their time, and we were using technology already in our budget.

A number of networks, denominations, and mission agencies asked if we could start groups for churches in their tribe. There was just one problem: the infrastructure built for the Resourced model was tapped out. But if we pivoted to the Groundswell model? It was suddenly scalable to a new level. However, doing this would mean sacrificing control.

In the end, we made a choice. We gave those leaders everything we had and held nothing back: Detailed notes of every session, scripts, worksheets, slides, training videos we’d recorded for coaches, video replays of each week, email templates. Everything we had, we gave it to them, free of charge. We trained the leaders of those tribes of churches, walked them through the essentials of the radical minimums and what they’d need to do. We then released them to be the yeast in the dough of their specific tribe.

This pivot worked.

At the end of Week 2 there were 1,094 churches in a coaching group. At the end of Week 5, there were more than ten thousand churches in a coaching group, spread across thirty-nine countries, speaking eleven languages.

And after that? We simply stopped counting.