What is a connection health score in church software?

Every church management dashboard can report how many people showed up. Almost none of them can answer the question pastors actually lose sleep over: which specific people are quietly coming loose? A connection health score exists to answer that second question, and whether it answers with a status word or a percentage changes what a pastor does next.

Key takeaways

• A connection health score is a per-person read of how connected someone is to the life of their church, built from signals like recent care contact, attendance rhythm, group membership, and real friendships.
• It surfaces as a plain status label (thriving, engaged, drifting, disconnected, or unknown when there is not enough signal), never as a number or a percentage shown to the user.
• The rules behind it are explicit and church-owned: every signal goes green, amber, or red by thresholds the church can see and change, so there is no black box.
• A status word leads to a next step; a percentage leads to a ranking. That design choice is deliberate.
• The score never acts on its own. It puts the right person in front of a pastor, and a human decides what happens next.

Quick answer: What is a connection health score?

A connection health score is a per-person read of how connected someone is to their church, drawn from signals like recent care contact, attendance rhythm, group membership, and real friendships. It surfaces as one of five plain status labels: thriving, engaged, drifting, disconnected, or unknown when there is not yet enough signal to say. Never a number, so the reading always points to a pastoral next step instead of a ranking.

The problem the score answers

Churches are good at counting the things that are easy to count. Attendance totals and giving trends roll up neatly and land on a dashboard every week. What none of them reveal is which specific member is quietly drifting away before it is too late.

That gap has a shape. A person can attend for years while remaining unknown, which is why a church connection is defined as a real, two-way relationship rather than a mark in an attendance column. And the people most at risk rarely announce it; they simply appear less, then not at all. Spotting them early is its own discipline, covered in depth in how to identify isolated church members before they leave.

A connection health score is the software-feature answer to that gap: one term, introduced by FlockConnect to church management software, for a per-person reading of connection that a pastor can actually act on. Traditional church software reports top-down data about the institution. Connection health reads the peer relationships around each person, which is where belonging tends to form. It is one of the core ideas inside Church Relationship Management.

What actually gets measured: the five rings

In FlockConnect, connection health is drawn as a set of rings around each person, one ring per signal family. Each ring is a plain rule the church owns, with sensible defaults out of the box.

• Care contact. Has someone from the church reached out recently? The default expects contact within the last 30 days and at least twice a quarter.
• Attendance. Has this person been seen recently and regularly? The default expects attendance within the last two weeks and at least six times a quarter.
• Groups and ministries. Is the person part of at least one group or serving team? Belonging to a smaller circle is one of the strongest signals that someone is held by more than habit.
• Network. Does the person have real friendships inside the congregation? The default aims at roughly seven, because the 7-friend threshold research found that people who form several real friendships in their first months are far more likely to put down roots and stay.
• Giving cadence. Has a previously steady rhythm of giving gone quiet? This ring reads cadence only, never amounts, and it is permission-gated so only staff the church explicitly permits can see it.

Every one of those defaults is a starting point, not a verdict. The thresholds are stored per church and can be tuned to fit a rural congregation of eighty as readily as a multi-site church of two thousand.

How the status is derived, with no black box

The mechanics are deliberately simple enough to explain in a hallway conversation.

Each rule checks one signal against explicit thresholds and comes back green, amber, or red. A ring's outcome is the worst outcome among its rules, on purpose: an explainable health surface should not quietly average away a signal it knows is red. If a ring has no usable data at all, it stays unknown rather than guessing.

The per-person status then follows from the rings:

• All rings green: thriving.
• No reds and no ambers, with at least one ring known: engaged. The rings with data are green; some simply have nothing to report yet.
• Any ring amber, with no red: drifting.
• Any ring red: disconnected.
• Everything unknown: unknown, stated plainly instead of dressed up as a guess.

Alongside the status, the evidence view shows exactly what counted, what did not, and what is missing. A pastor never has to take the word for it; the reasoning is right there.

New people get their own treatment. Someone who joined three weeks ago has not had time to build seven friendships or attend six times a quarter, so newcomers are read on a newcomer curve rather than being penalized by rules written for established members.

Why it is a status word, never a percentage

This is the design decision behind the whole feature, and it is worth explaining plainly.

Under the hood, an internal score does exist; software needs a number to sort a list. What FlockConnect refuses to do is show that number to anyone. No member is ever presented as "73% connected."

There are three reasons.

First, a percentage manufactures false precision. The difference between 71% and 76% is noise, but a number that specific begs to be read as measurement. Relationships do not resolve to two significant figures, and pretending they do teaches staff to trust the number over their own pastoral instincts.

Second, percentages invite ranking. The moment people carry numbers, someone will sort the congregation from highest to lowest, and the congregation becomes a leaderboard. C.S. Lewis argued in The Weight of Glory that there are no ordinary people, that every person carries a weight no metric can hold. A tool for shepherding software has to be built as if that is true, because it is. Michael Tribett, the founder of FlockConnect, has made the same point in plainer terms: when a church loses someone, it loses a person, not a number, and the tooling should never train anyone to think otherwise.

Third, a percentage hides the thing a pastor actually needs. "68%" does not say whether the problem is attendance, friendships, or a care contact that never happened. A status word backed by ring-level evidence does: this person is drifting, and here is the signal that went amber. The status invites a next action. The number invites a spreadsheet.

As Tribett puts it, "The goal is to strengthen the human element of the church." A percentage quietly works against that goal. A status word serves it.

Who sees it, and what happens next

Connection health surfaces where the work already happens: on the people directory, on each person's profile, and in church-wide insights. A pastor scanning the directory sees status at a glance; opening a profile shows the rings and the evidence behind them.

The score itself never creates a task, sends a message, or schedules anything. When a pastor sees a drifting status, the human move is to turn it into a follow-up with a named owner and a due date, which is exactly the workflow a church follow-up system exists to carry. The score finds the person; a person does the caring.

Churches also keep full ownership of the rules. A built-in rule editor lets a church adjust windows, thresholds, and newcomer handling so the definition of "connected" matches its own congregation rather than a vendor's average. The reading is recalculated nightly and refreshed when new interactions are logged, so it stays current without anyone tending it.

About the author

Michael Tribett is the founder of FlockConnect, the first purpose-built Church Relationship Management (ChRM) platform. He holds a Master of Divinity in Christian Ministry from Southeastern Baptist Theological Seminary and is based in Raleigh, North Carolina. FlockConnect participated in the Missional Labs Faith & AI Accelerator during Winter and Spring of 2026 and is an official Planning Center integration partner.

Frequently asked questions

Is a connection health score the same as an engagement score?

No. Generic engagement scores usually blend activity metrics into a single number that measures interaction with the institution. A connection health score reads relational signals (care contact, attendance rhythm, group membership, friendships) and reports a status word per person. The aim is pastoral action for one individual, not an aggregate engagement metric.

Can a church change what counts toward a connection health score?

Yes. Every threshold is stored per church and editable through a built-in rule editor: the care-contact window, the attendance expectation, the friendship target, and how newcomers are treated. The shipped defaults are a starting point calibrated to common patterns, but the church owns the definition of connected for its own congregation.

Does a connection health score look at how much people give?

No. The giving ring reads cadence only, meaning whether a previously steady rhythm has gone quiet, and it never stores or displays dollar amounts. It is also permission-gated, so the signal is visible only to staff the church explicitly permits. Giving is treated as one signal of belonging, never as a measure of worth.

Is AI deciding who is spiritually healthy?

No. The status comes from explicit, human-readable rules the church can inspect and change, not from a model making judgments. Nothing here measures spiritual health; it measures observable connection. FlockConnect's position is that AI should work like a sheepdog, not a shepherd: it can point, but the shepherding stays human.

Is there a numeric score under the hood?

Yes, and it is fair to say so plainly. The software computes an internal score to order lists and derive the status band. What FlockConnect never does is display that number to a user. The interface shows a status word and the evidence behind it, because a word points to a next step while a percentage points to a ranking.

What does an "unknown" status mean?

It means the system does not have enough signal to say, and it refuses to guess. If every ring lacks usable data, the person shows as unknown, and the evidence view lists what is missing. That honesty is deliberate: the missing data itself tells the church which signals it is not yet capturing.

How often is the score updated?

Connection health is recalculated in a nightly batch and also refreshed when new interactions are logged, so a care contact recorded this morning is reflected without waiting a week. Nobody has to run a report or trigger anything; the reading stays current in the background.