I keep coming back to this research. It's been replicated for fifty years and the result doesn't budge. The difference between a member who stays and one who drifts out isn't the sermon, isn't the worship, isn't the programs. It's how many close friends they made in their first six months.
Key takeaways:
- Research popularized by Win & Charles Arn, building on Flavil Yeakley's church retention studies, found that members with seven or more close friendships in their first six months almost never leave.
- Members with two or fewer close friendships in the same window almost always leave within two years.
- The finding has been replicated across the 1970s, 1990s, and 2010s with the same qualitative result. It is generationally stable.
- Retention is relational. Sermon quality and worship style matter for formation, but they do not predict who stays.
- Pastors need a way to see relational formation in their congregation, because it does not show up on the attendance report or the giving summary.
Quick answer: what is the 7-Friend Threshold?
The 7-Friend Threshold is shorthand for a finding that keeps showing up in church-retention research. Members who form seven or more close friendships inside their congregation within the first six months stay at extremely high rates, often close to 100%. Members who form fewer than two close friendships in that window leave at extremely high rates, usually within two years. The principle is associated with Win & Charles Arn, whose church-growth work built on earlier retention studies by Flavil Yeakley. Later replications have produced the same result across more than four decades.
I run FlockConnect, and I've talked to a lot of pastors about this research. Most of them have heard of it. Almost none of them have a way to act on it, because their church management software was built to track attendance and giving, not friendships.
The research, decade by decade
Flavil Yeakley's original work (1970s)
Flavil Yeakley is a communication researcher whose doctoral work in the late 1960s looked at persuasion and church growth. He started tracking new members over time and asking them a specific question: who do you know well in this congregation, and when did you get to know them? What he found, and wrote about through the 1970s and 1980s, was the same thing again and again. The strongest predictor of whether a new member was still active two years later was not doctrinal agreement, not attendance frequency, not giving patterns. It was the number of real friendships they had formed during onboarding.
His book Why Churches Grow summarized this for a pastoral audience. The research is rigorous and it still gets cited in contemporary retention literature.
Win & Charles Arn's popularization (1980s–90s)
Win Arn was one of the central figures in the American church growth movement. His son Charles joined him, and together they ran the Institute for American Church Growth through the 1980s and 1990s. They took Yeakley's findings out of academic journals and put them in front of pastors in trainings, books, and consultations. That's how the "seven friends" number became the shorthand. Read any serious pastoral-training material on retention from the last thirty years and you will find the Arns cited somewhere.
Win Arn's book The Master's Plan for Making Disciples, co-authored with Donald McGavran, is the cleanest entry point for the original frameworks.
Replications (2000s and 2010s)
Denominational research offices and independent ministry researchers have tested the finding in successive decades. Barna has published repeatedly on church dropout. Thom Rainer's work at LifeWay and Church Answers hit the same pattern. Studies inside Southern Baptist, Assemblies of God, and mainline Protestant tracking have produced the same result qualitatively: relational integration is a stronger predictor of retention than program quality.
The specific number varies a little — some report five, some eight, some ten. That's not important. The threshold effect is what matters. Above a certain count of close friendships, retention approaches certainty. Below a very low count, attrition does.
For practitioner-accessible writing on the modern version, see LifeWay Research and the Church Answers archives.
Why seven, and why two
Neither number is magic. They describe a non-linear relationship between friendships and staying power.
The lower threshold: fewer than two friends
If a member has zero or one close friendship in the congregation, there is nothing holding them in relationally. They can miss a Sunday and no one notices. They can go through a hard season and get no calls. They can drift out without any single conversation, because no one is close enough to have the conversation. At that level of isolation, attrition is almost structural. They aren't leaving because of a friend count. They're leaving because nothing in the community is holding them.
The upper threshold: seven or more
With a cluster of real friendships, a member is inside a web of mutual expectation. A life crisis triggers multiple calls. A missed Sunday gets noticed the next week. The small group, the serving team, the coffee after service — all the same people holding each other together. Leaving the church means leaving the friendships. That is a much larger cost than leaving a program.
The middle: three to six
Here the data gets noisy. Some members stay, some leave. The tipping factors are the ones pastors can actually influence: introductions, group placements, serving opportunities, pastoral attention. This is where a ChRM tool earns its keep, because these are the members who are at risk but not lost, and they are the easiest to miss without visibility.
Why the finding holds across generations
What I find most compelling about the 7-Friend Threshold is how boring it has been across generations. Baby Boomers, Gen X, Millennials, early Gen Z — the threshold effect holds for all of them. The mechanics of friendship formation are completely different. In 1978 it was Sunday school class and church potlucks. In 2018 it was small groups and GroupMe threads. The surface changes. The need does not.
That is because the underlying thing is not generational. It is anthropological. Humans are made for belonging, and a church that cannot produce belonging for its new members will lose them, regardless of the decade.
C. S. Lewis makes a version of this point in The Four Loves. He argues that friendship, what the ancients called philia, is the love modern people most often neglect because it looks least necessary. "Friendship has no survival value; rather, it is one of those things which give value to survival." Lewis is saying friendship is what turns existence into a life worth continuing. For a congregation, friendship is what turns mere membership into a commitment worth keeping. The data has been agreeing with Lewis for fifty years.
The theology underneath the data
The research is compelling on its own, but I don't think it stands alone. It matches what Scripture has said about belonging from the start.
The "one anothers"
The New Testament contains roughly fifty "one another" commands. Love one another. Bear one another's burdens. Confess to one another. Encourage one another. Forgive one another. None of them can be done alone. Each one assumes a web of relationships dense enough that the verb makes sense. A member with two acquaintances cannot "bear one another's burdens" at any scale, because no one has let them near enough to a burden to bear it.
Tim Keller, in Center Church, argues that gospel-shaped community isn't a program the church runs. It's the byproduct of a certain kind of belonging. The friendships are not the strategy. They are the substance. The strategy is whatever makes the friendships possible.
John 10 and 1 Peter 5
Both passages describe pastoral ministry in terms of knowing. The shepherd knows the sheep by name. The elder cares for the flock that is among them. This is not knowledge about a database. It is the knowledge of relationship. A congregation that hasn't formed real friendships between its members is one where the shepherd can't, in practice, know the flock, because the flock's identity is partly its relational life, not just its individual profiles.
John Piper, in Brothers, We Are Not Professionals, refuses the managerial picture of ministry. "Pastoral work is not a profession; it is a calling to care for souls." Souls, for Piper, are not atomized units. They are people embedded in relationships, and shepherding them means watching over those relationships, not just the souls in isolation.
Friendship as a means of grace
Francis Chan's Multiply and Gavin Ortlund's work on retrieval theology land on the same place from different directions: the Christian life is not primarily individual. It is communal in a way that means friendships are not an accessory to discipleship. They are part of its machinery. You become like Christ in large part by being near people who are becoming like Christ. Cut the friendships, cut the mechanism.
That is why the 7-Friend Threshold is more than a retention metric. It is a measurement of how much gospel-shaped life a new member has actually been drawn into. Members without friendships don't just leave. They leave before discipleship has the chance to take hold.
What pastors actually do with it
Knowing the research is cheap. Acting on it is hard. Here is what pastors at different sizes of church actually do.
Under 100 members
At this size the pastor can usually hold the relational map in their head. The right move is to be deliberate about introductions, especially for new members in the first 90 days. A weekly prayer-and-planning rhythm is enough. Who visited, who do they need to meet, who will I ask to make the introduction. Write it down or you will forget.
100 to 300 members
The map starts to exceed any one pastor's memory. The common failure mode at this size sounds like "I thought someone else was going to connect them." Churches here benefit from assigning a specific team — connection coordinators, care partners, or the small-group leaders themselves — explicit responsibility for new-member integration. Written records beat memory once you're past about 120 members.
300+ members
Manual tracking breaks down. Members fall through the cracks, not because anyone is negligent but because the information isn't visible anywhere. Pastors at this size need a tool that measures relational formation directly, not just attendance and giving. See the 2026 pastoral care tools post for the current options.
How to measure the threshold in your church
You cannot act on a number you don't have. Four practical ways to get one.
- Ask directly. A simple new-member check-in at the 3-month and 6-month mark with one question: "how many people at this church do you know well enough to call if you had a hard week?" Most members will answer honestly.
- Survey through small group leaders. Ask leaders to observe and report on the integration of the members in their care. This catches what members themselves may not notice or report.
- Proxy through group participation. Consistent small-group attendance for three months sets a floor of 3–5 weekly relational interactions. Not a substitute for direct measurement, but better than raw attendance alone.
- Use a ChRM tool. Platforms like FlockConnect compute a live connection score across the whole congregation from observed relationships, group participation, pastoral interactions, and optional member-reported data. The dashboard surfaces members below the threshold automatically.
Whichever method you pick, use it consistently. Ad-hoc gut tracking over-reports connection, because the connected members are the ones pastors see. The isolated ones are by definition harder to notice.
Common misreadings of the research
It is not a scorecard
The threshold is not a ranking of which members are "better" Christians. It is a retention-risk indicator. A new member at two friends isn't spiritually deficient. They are relationally exposed. The pastoral response is to help them form friendships, not to judge them.
It is not a case for forced community
Using the research to guilt members into small groups or to stage awkward mixer events backfires. Friendships form in the natural overlap of shared interest and repeated contact. The pastoral task is to create the conditions for that overlap, not to manufacture it. Well-designed small groups, serving teams, and discipleship paths work because they create proximity with purpose.
It is not a replacement for good teaching, worship, or discipleship
The research does not claim sermons don't matter. It claims sermon quality alone doesn't predict retention, and relational formation does. Healthy churches teach well and form friendships. Doing only one and skipping the other is the failure mode.
How FlockConnect uses the threshold
I built FlockConnect specifically because I was tired of seeing pastors cite this research in trainings and then go back to their church management software that couldn't act on it. The whole point of FlockConnect is to make the threshold operational.
Every member in FlockConnect has a live connection score, computed from observed relationships, group participation, pastoral interactions, and optional member-reported data. The dashboard surfaces three things:
- Members below the threshold. The short list of people most at risk of drifting. This is where a pastor's week-to-week attention should go.
- Trend over time. Whether a member's integration is improving, flat, or declining. A member whose score drops over two months, even if still above threshold, is a signal worth acting on.
- Suggested introductions. Members who should meet each other based on life stage, proximity, and shared interests. This is how a church scales introductions past the point where one pastor can hold every match in their head.
FlockConnect integrates with Planning Center through one-click OAuth. For churches on other ChMS platforms, FlockConnect imports members via CSV from any system. Pricing starts at $10/month for solo pastors with a 14-day free trial and no credit card required.
Related reading:
- The Best Pastoral Care Tools for Pastors in 2026
- The Ultimate Church Management Software Comparison 2026
- What is a ChRM?
About the author
Michael Tribett is the founder of FlockConnect, the first purpose-built Church Relationship Management (ChRM) platform. Based in Raleigh, North Carolina, he built FlockConnect because the research on church retention has been clear for fifty years and the software had not caught up. FlockConnect is a member of the Missional Labs Faith & AI Accelerator and an official Planning Center integration partner.
Frequently asked questions
What is the 7-Friend Threshold?
The 7-Friend Threshold is a church retention principle from research by Flavil Yeakley, popularized by Win & Charles Arn. It says that members who form seven or more close friendships in their first six months stay at very high rates, and members with fewer than two close friendships in the same window leave at very high rates, usually within two years.
Who is Flavil Yeakley?
Flavil Yeakley is a communication researcher whose work in the 1970s focused on church growth and retention. His longitudinal studies of new members produced the core finding that friendship formation during onboarding was the strongest predictor of long-term retention. His book Why Churches Grow is the pastoral-audience summary.
Who are Win & Charles Arn?
Win Arn was a leader in the American church growth movement. Charles is his son and coauthor. Through the Institute for American Church Growth and their writing in the 1980s and 1990s, they translated Yeakley's research for pastors and made the "seven friends" formulation the recognized shorthand.
Why specifically seven friends?
The specific number varies across studies. Some report five, some eight, some ten. What matters is the threshold effect. Above a certain count of close friendships, retention approaches certainty. Below a very low count, attrition does. Seven is the number that stuck because it appeared in the Arn-popularized literature.
Why does the finding hold across generations?
The underlying need for belonging is anthropological, not generational. How friendships form has changed drastically between Baby Boomers and Gen Z, but the threshold effect on retention has been stable in replicated studies across the 1970s, 1990s, and 2010s.
How do I measure the 7-Friend Threshold in my church?
Four common methods. Ask new members directly at 3- and 6-month checkpoints. Survey through small group leaders. Proxy through consistent small-group participation. Or use a ChRM platform like FlockConnect that computes live connection scores across the whole congregation. Whatever you pick, use it consistently. Ad-hoc gut tracking over-reports connection because isolated members are harder to notice.
How does FlockConnect use the 7-Friend Threshold?
FlockConnect computes a live connection score for every member from observed relationships, group participation, pastoral interactions, and optional member-reported data. The dashboard surfaces members below the threshold, shows trends over time, and suggests introductions based on life stage and shared interests.
Does this mean sermon quality doesn't matter?
No. The research does not claim preaching or worship are irrelevant. It claims those factors alone do not predict retention, and relational formation does. Healthy churches teach well and form friendships. Doing only one and skipping the other is the failure mode.