82% of the People You'll Lose, You Lose in Year One

When somebody leaves a church after fifteen years, everybody notices. We make calls. We take it personally, sometimes for months. But that isn't where most of the leaving happens. Gary McIntosh and Charles Arn, working on why people drop out of churches, found that 82% of the people who leave do it inside their first year. The back door isn't over by the long-timers. It's standing wide open right next to the newest people in the room.

Put it next to the other number

Lifeway surveyed a thousand pastors in 2023 and found that the average church holds onto 79% of its new commitments. Read those two findings together and you get the shape of the thing. Roughly one out of five new people you reach is going to slip away, and when it happens it'll almost always be during the stretch when everybody still assumes they're just settling in.

That Lifeway study is worth sitting with for another reason. Out of everything they tested, only four things actually predicted whether a church grew: evangelism, assimilation, small-group discipleship, and church size. Assimilation, which is just a churchy word for what happens to a person after they walk in, is sitting right there next to evangelism on a very short list.

Which means reaching a family in March and losing them in October isn't really evangelism. It's a slower way of telling them they weren't seen.

Why the first year

Because that's the stretch when belonging either takes hold or it doesn't. Somebody in their first year has shallow roots by definition. Few friendships, no role, no habits yet. So every miss costs more. The Sunday nobody said their name. The interest card nobody followed up on. The week they were out sick and nobody called. A member of fifteen years can absorb a bad month on the strength of a hundred good ones. A family in month two has nothing saved up.

The decision about whether to come back at all happens even faster than that, usually inside the first three visits, which I wrote about in You Have 45 Days, Not Six Months.

The encouraging side of this is that the same window is when small things count the most. Attention that would feel routine to somebody who's been around a decade feels like a big deal to somebody in week three. You will never again get as much return on a phone call as you do in the first ninety days.

What it takes to work that window

Most churches have exactly one tool for this, which is being friendly. Ours was friendly for years and we still lost people. Friendliness is in the air, and year one gets lost in specifics. Every checkpoint that actually holds has two things on it: a person's name and a date.

In week one that means somebody real makes contact the same day, a short sequence follows, and the second visit gets written down rather than guessed at. If you can't say which of your first-time guests came back, you can't tell where year one even started for them, let alone how it's going.

In the first month it means an invitation to something smaller than a service. A meal, a group, a serving team. What the research keeps pointing at is early relationships, and the move most churches skip is simply asking. Only 23% of guests ever get invited to anything outside a service, which I dug into in The Most Underused Move in Church Hospitality.

And for the rest of the year it means somebody notices absence. A first-year family that misses three Sundays in a row should get a phone call, not a shrug. Absence early on is information, and information needs somebody assigned to hear it.

Where we started

We started at the very front of the window, partly because that's where the leverage is and partly because it's the only part a church our size could actually staff. Two personal touches in the first week, a second visit we actually mark down, and a handoff to a group leader. The whole thing is written out in We Rebuilt Guest Follow-Up Around Two Texts and a Gift Card, and the software half of it became GuestPilot.

The other eleven months still belong to the church, and they should. No software is going to disciple anybody. But the window opens in week one, and 82% is roughly what it costs to assume somebody else is watching it.