Skip to content
Church Posting
Search
Menu

Latest

See all posts

The library

See all resources

Apps and tools

See all apps

One email when a new post goes up, nothing held back. Subscribe

Average attendance hides whether your church is growing or shrinking

Why the mean lies, how to count the same way every week, and a rolling eight week method that reports real attenders and the frequency ratio instead of a flattering number.

Church Posting 7 min read Church growth

Updated

A man standing at the back of a full church aisle holding a small notebook behind his back
Illustration made for Church Posting

Two churches report 200 in worship. The first has 200 people who come nearly every week. The second has 380 people who come every other week. Those are different congregations with different budgets, different volunteer capacity, and different pastoral needs, and the number 200 hides the difference completely.

Counting well tells you whether you know your church or are guessing about it.

Count the same way every week

Pick a method, write it down, and never change it mid-year. Changing method mid-year destroys your ability to compare anything.

Head count at a fixed moment. An usher counts the room during the second hymn or the offering, every week, same moment. Count adults and youth in the room. Count children in classes separately and label them separately. One counter, one backup, a card handed to the office. Accuracy on a manual count runs a few percent off in a full room, and that is acceptable as long as the error is consistent.

Check-in systems. Reliable for children, incomplete for adults unless your congregation actually checks in, which most do not. Use children’s check-in numbers as a hard count for kids and nothing else.

Chair count. Count empty chairs and subtract from chairs set. Faster and more accurate than counting heads in a room over 150. Requires knowing the true number of chairs, so count the room once and write it on the wall backstage.

Giving units. Not attendance. It measures households who gave, which is a separate and useful trend line. Never substitute one for the other.

Record online viewers in a separate column with a separate label, and never add them to in-person attendance. A fifteen second autoplay in a scrolling feed is not a worshipper. If you report streaming at all, report minutes watched or the count of devices past ten minutes, and say which one you used.

Averages lie in three specific ways

The mean gets wrecked by Easter and Christmas. One Easter of 410 against a normal 180 pulls a fifty-two week mean up by four or five people, and the pull is bigger for small churches. Report the median week alongside the mean. When the mean sits well above the median, big holiday services are inflating your picture.

Year over year is the only honest comparison. Attendance is deeply seasonal. August against July tells you about vacations. Comparing a four week block to the same four weeks last year tells you about the church.

A single week means nothing. Weather, a flu wave, a school event, a home football game. Use a four week trailing average for anything you report to a board, and an eight week average for anything you are about to make a decision on.

The number most churches are missing

Frequency erosion is the quiet story in most congregations. People who once attended four Sundays a month now attend two. Head count stays flat while the number of households connected to your church rises, then falls, and the head count only reveals it two years later.

The national numbers say to expect it. Pew Research Center’s 2023 to 2024 Religious Landscape Study, fielded to 36,908 respondents between 17 July 2023 and 4 March 2024, found 33 percent of US adults attend religious services at least once a month, and among adults currently raising children, 26 percent attend at least weekly with another 9 percent once or twice a month (Pew Research Center). Read that against your own roll. In a country where monthly is the more common commitment than weekly, a congregation that has not measured frequency is not measuring the thing that changed.

Measure it with a rolling window:

  1. Take an eight week window.
  2. Count the unique people who attended at least once in those eight weeks. Call this your active attenders.
  3. Divide total weekly attendance across those eight weeks by 8 to get average weekly attendance.
  4. Frequency ratio = average weekly attendance divided by active attenders.

A ratio near 0.7 means your people come most weeks. A ratio near 0.45 means roughly every other week is normal in your congregation. Watch that ratio across quarters. A ratio that falls while attendance holds steady means you are recruiting new people fast enough to mask people drifting away, which is the exact condition that looks healthy for two years and then collapses.

The four steps of the eight week window beside two worked cards. Church A reports 200, has 200 average weekly attendance against 215 active attenders, and a ratio of 0.93. Church B also reports 200, has 190 average weekly attendance against 380 active attenders, and a ratio of 0.50. Both are plotted on a scale from 0.40 to 1.00 marked with the two landmarks: 0.45 where every other week is normal, and 0.70 where people come most weeks.
Church B carries nearly twice the households and nearly twice the pastoral load, and reports the same number as Church A. The ratio is the only line on the monthly report that tells the two apart.

Getting unique attenders requires knowing who was there. Children’s check-in, small group rolls, and serving rosters get you most of the way in a church under 250. Planning Center People is free with an unlimited database, and Check-Ins has a free tier capped at 10 daily check-ins, which is enough for a small children’s ministry and not enough for a large one (Planning Center pricing, checked 29 July 2026). If you have no system at all, free church management software states where each free tier actually stops, and choosing church management software covers what to ask before you pay for one.

Do not build a surveillance system to get a cleaner number. Approximate is fine and honest, and a roll that tracks individual attendance week by week is a record of people that has to be governed like one. Decide who can see it before you build it, using the tiers in protecting the data a church holds about its members.

A one page monthly report

Elders and staff should see the same page every month. Six lines:

  • Median weekly worship attendance, this month
  • Four week trailing average, and the same four weeks last year
  • Children counted in classes, four week average
  • Active attenders in the last eight weeks, and the frequency ratio
  • First time guests this month, and return rate from two months ago, per the guest follow-up system
  • Giving households, and the count who gave in each of the last three months

Add one sentence of interpretation and no more. “August is down 14 percent from July, which matches last August’s 12 percent drop.” That sentence is worth more than a chart.

Rules that keep the number honest

  1. One method, written down, applied every week. Include what counts as a person: nursery workers yes, the sound volunteer yes, the pastor yes, the toddler in the sanctuary yes.
  2. Never report a range when you have a number. “About 200” usually means someone did not count.
  3. Never quietly change what is included. Adding the second service, the Saturday night gathering, or online viewers to a trend line without labeling the change is how a church convinces itself it grew 30 percent.
  4. Publish the bad months to the board. A leadership team that only sees good months will not believe good months either.
  5. Do not report attendance from the pulpit as a scoreboard. The congregation is not an audience being told its ratings.

What to do with a real decline

First, check whether it is departure or frequency. Departure means people left. Frequency means the same people come less. The responses are entirely different.

For frequency erosion, look at what changed in the last eighteen months: service time, Sunday schedule, a beloved teacher who stopped, a small group that folded. People stop attending weekly when the week stops requiring them.

For departure, look at the exit list by name and ask an elder to call ten of them. Not a survey. A phone call from a person who knew them. You will learn more in ten calls than from any consultant’s report, and you will occasionally get someone back.

For seasonal decline, do nothing. Some of the calendar is weather.

A decision tree with three questions and four exits. Is the four week trailing average down against the same four weeks last year: no exits to seasonal, do nothing. Did active attenders fall too: yes exits to departure, read the exit list by name and have an elder phone ten of them. Did the frequency ratio fall while active attenders held: yes exits to frequency erosion, look at the last eighteen months for a changed service time, a teacher who stopped, or a group that folded; no exits to a measurement change, where a service or a room was added or online viewers were folded in. A band at the bottom names the pair to watch, a flat head count beside a falling ratio.
Four different problems produce the same falling number on a board report. The exit you land on decides whether you make ten phone calls, read eighteen months of schedule changes, or do nothing at all.

The tradeoff worth naming

Better counting costs something. It requires a volunteer to count every week, a system to store names, and the discipline to publish numbers that sometimes embarrass leadership. In exchange you learn whether your church is growing, aging, drifting, or holding, which determines whether the building project makes sense, whether you can afford the second staff hire, and how many volunteers you can realistically ask for.

Guessing is cheaper. Guessing is also why churches hire staff they cannot sustain and build rooms they cannot fill. Before any decision that costs money for more than three years, run the eight week window and look at the frequency ratio. Then decide. Also audit what the calendar is actually asking of these people, using what actually fills a church calendar, and if the decision in front of you is a building, take the numbers into the announcement sequence in communicating a building project without losing the room.

Sources

Gallup’s religion topic page was opened while writing this and served no attendance table, so no Gallup figure is quoted here.

An older man and a young woman filling in a hand ruled paper grid on a table in a church

Social media

A church social media strategy that survives the volunteer

What the account is for, a three post week, who owns the login, the consent rule for a child's photo, what to do when a member dies, and scheduler prices checked 29 July 2026.

Jul 22, 2026 · 22 min read

An older woman in glasses pointing at a page of handwritten notes on a wooden desk while a younger woman beside her holds a phone

Social media

Training a church social media volunteer: the first four weeks

A four week onboarding for someone who has never run an account: what they learn each week, the access they get and when, the eight shot capture card, the approval rule, and the handover document.

Jul 17, 2026 · 24 min read

Two women at a folding table in a church hall pressing rubber bands around thick stacks of cards, a gray tote and a paper form beside them

Church growth

What church direct mail costs: $0.26 at retail, $0.136 with the nonprofit permit

EDDM and nonprofit bulk rates read off USPS Notice 123 on 29 July 2026, the two $390 fees, the 3,145-piece break-even, the size rule that disqualifies a 6 by 9 postcard, and when to mail nothing.

Jul 12, 2026 · 25 min read

Back to top