Nobody enjoys the conversation where you tell a volunteer, gently, that something is not working. It is easier to keep hoping the group turns a corner on its own, especially when the leader is someone you like and trust and who said yes to a hard job nobody else wanted. But a group that has been quietly shrinking for three months is not going to fix itself because you avoided mentioning it.
The good news is that this conversation goes better than most people expect, if you walk in with facts instead of a feeling. A leader who hears “I've noticed something and want to understand it with you” responds very differently than one who hears a vague sense that things “don't feel right.” Here is what to look at before you sit down, and how to have the conversation once you do.
Separate a slump from a slide
Every group has bad months. A leader travels, a holiday falls on the wrong Wednesday, half the group is out with the flu at the same time. None of that is a signal worth acting on. What you are looking for is a slide — a pattern that holds up over eight or ten weeks, not a single rough patch that a normal calendar would explain on its own.
The distinction matters because acting too early teaches leaders that any dip gets them a talking-to, which makes them defensive before you have even said anything. Give it real time. Look at attendance charted against its own history rather than against last week alone, so a two-month view, not a two-week snapshot, is what tells you whether this is worth raising.
Look at who left, not just how many
A headcount going from fourteen to nine tells you something changed. It does not tell you what. Before the conversation, pull up the people who stopped coming and look at each one individually — their engagement timeline, whether they are still showing up anywhere else at church, whether they left around the same week as someone else in the group.
If five people left within a month of each other and none of them are otherwise disengaged from the church, that points toward something happening inside the group itself — a conflict, a format that stopped working, a leader who has checked out. If the departures are scattered over months and each person also stopped showing up elsewhere, the group may not be the real story at all; the person is drifting from the whole church, and the group is just where you noticed it first.
Check whether the leader is showing the same signs
Attendance is the easiest thing to see, but it is worth glancing at the leader's own pattern too. Are they still signed up anywhere else, still serving, still around on Sundays? A leader whose own engagement has quietly thinned out is often the actual explanation for a group that is thinning out under them — not because they stopped caring, but because whatever pulled their attention away from the group probably pulled it away from other things too.
This is not about catching someone in a failure. It is about walking into the conversation already knowing whether you are talking to someone who needs a new format, a smaller group, a break from leading altogether, or just to hear that someone noticed and cared enough to ask.
Bring the specific numbers, not an impression
“Your group seems to be struggling” is a hard sentence to respond to, because it hands the leader a feeling and asks them to defend against it. “Attendance was averaging eleven in the spring and it's been six the last two months” is a different kind of sentence. It is not an accusation. It is a fact both of you can look at together, and it opens a conversation instead of closing one down.
Have the number ready, but do not lead the whole conversation with it. Say what you noticed, then ask what the leader has noticed too. Most leaders already know their group has thinned out — they have been sitting with that feeling alone for weeks, wondering if anyone else has seen it. Being the first person to say it out loud, kindly, is often a relief rather than a blow.
Ask questions before you offer solutions
The instinct, once you have the numbers in hand, is to walk in with a plan: change the meeting night, split the group, recruit a co-leader. Resist that instinct until you have actually heard from the leader. They may already know exactly why attendance dropped — a member moved away and took three friends with her, the group outgrew its living room, the curriculum stopped landing — and your plan built without that information will miss the real problem entirely.
Open with something like: “What's your read on why the numbers have shifted?” Then actually wait for the answer before offering yours. A leader who feels heard before they feel managed is far more likely to stay in the role and try something new, rather than quietly deciding this is the semester they step down.
Watch for burnout disguised as a leadership problem
Sometimes what looks like a struggling group is really a tired leader. Someone who has led the same group for three years without a break, who has never had a co-leader to share the load, starts to run on fumes in ways that show up as a flatter meeting, less follow-up between members, fewer people invited. The group did not get harder to lead. The leader got worn down leading it.
If that is what the numbers and the conversation point to, the fix is not a talk about performance — it is rest, a co-leader, or permission to hand the group to someone else without that being treated as a failure. The same warning signs covered in avoiding burnout on a serving team apply just as much to someone leading a small group as to someone on a Sunday rotation.
Consider whether the format itself has aged out
Sometimes the leader is fine and the group is fine, and the problem is simply that the format stopped fitting the people in it. A group built around young families with toddlers does not automatically keep working once half the kids are in middle school and bedtime is no longer the reason everyone leaves by eight. A study written for people new to church does not keep landing once most of the room has been there for six years and wants something with more depth.
This is worth raising as a real possibility before you assume the leader needs coaching at all. Some of what makes small groups that people keep coming back to work is not the leader's skill so much as a format that still matches who is actually in the room. A leader who has been quietly fighting a format that no longer fits will look, from the outside, exactly like a leader who has lost their touch — and the two problems have very different fixes.
Have the conversation in person, and make it two-way
This is not a conversation to have by text or in a hallway between services. Sit down, ideally somewhere other than church, and make clear from the first sentence that you are there to understand, not to deliver a verdict. Bring the numbers, but bring them as a starting point for a conversation, not the conclusion of one you already had without them.
End the conversation with a concrete next step, agreed on together — a different meeting format to try for six weeks, a co-leader to recruit, a smaller and more manageable group size, or simply a check-in in a month to see whether anything shifted. A leader who leaves with a plan they helped build is in a completely different position than one who leaves with a private sense that they failed.
Follow up before the pattern repeats
The conversation is not the end of the process. Put a date on the calendar, four to six weeks out, to look at the same numbers again and talk about what changed. Groups that get a plan and then never hear about it again quietly slide right back to where they were, because the leader reasonably assumes that if nobody followed up, nobody was really watching in the first place.
Following up is also what makes the next hard conversation, if you ever need to have one, easier — with this leader or another one. It establishes that when attendance numbers get brought up, it is part of an ongoing, caring rhythm, not a one-time ambush that only happens when things have already gone badly wrong. A church using SundayBridge can pull that same attendance history again at the check-in date without having to reconstruct it from memory.