How to Track Church Attendance: Five Methods, Ranked

How you track church attendance decides which questions you can ever answer. A clicker gives you a number. A check-in station gives you names. Everything downstream depends on that one choice: guest return rates, seasonal trends, who quietly stopped coming.

Key points

• Five methods are in practical use: a tally clicker, the Headcounts app, Check-Ins stations, Groups attendance, and an analytics layer over the top. Only the last three record who.
• Headcounts records the number of attendees without names, ages or grades, and Planning Center’s own guidance is to use Check-Ins or Groups instead if you need to track individuals.
• Check-Ins is priced by the number of unique check-ins on your busiest day, so the cost scales with how many people you record, not how many stations you run.
• Kids and volunteers get recorded by name almost everywhere. Adults walking into a sanctuary are the gap, and no station setup closes it on its own.
• A total you cannot turn into a list of names will answer “how many” forever and “who” never.

Every church I’ve worked with tracks attendance. Almost none of them chose how.

The method arrived with the building, or with whoever set up the kids’ check-in five years ago, and the weekly number has been produced the same way ever since. That’s fine until someone asks a question the method can’t answer, and the honest reply is that the data was never collected in a shape that could.

So it’s worth being deliberate about it, once.

What are the ways to track church attendance?

Planning Center’s own help documentation splits the question the way I’d split it: some events only need a general attendance number, and other events need to track specific attendance so you can follow up with certain people. Everything below follows from which of those two you’re doing.

The ranking below is by what a method can answer, not by how little work it is. A method that costs you nothing and tells you nothing is not a bargain.

The five methods, ranked

5. A clicker and a notebook

An usher counts the room during the second song and writes the number on a card. It costs nothing, works during a power cut, and is genuinely accurate for a room you can see.

Its problem is not accuracy. It’s that the number goes into a notebook, or a spreadsheet on somebody’s own laptop, and the record has no structure: no split by service, no separation of adults from kids, no way to compare this September against last. Two years later you have 104 numbers and nothing to say about them beyond the average.

The cheapest upgrade here isn’t a new method. It’s putting the same numbers somewhere shaped, which is what the next option is for.

4. The Headcounts app

Headcounts is Planning Center’s free counting app, and it is better than most churches assume. You take a count from the Headcounts tab of any Check-Ins event or through the mobile app for Android or iOS, tapping plus and minus on a clicker interface. The total updates in real time even when several people are counting at once, so two ushers working opposite sides of a balcony don’t have to reconcile afterwards. You can set a capacity, and the screen turns yellow at 90% of it and red at the maximum.

What it doesn’t do is deliberate. Planning Center’s documentation says it plainly: Headcounts records the number of attendees without tracking individual details like names, ages, grades, and if you need to track individuals you should use Check-Ins or Groups instead.

The reason to prefer it over the notebook is that the counts land in the same place as your check-in data. They feed the Check-Ins event chart, which plots the number of unique people checked in or counted in Headcounts over a timeframe you choose, and exports as CSV. Structured totals, at zero cost, with no volunteer training. For the main auditorium, this is the realistic answer for most churches, and it will stay the realistic answer.

Just don’t mistake it for attendance tracking. It’s attendance counting.

3. Groups attendance

Midweek is where person-level data is easiest to get, because the group leader already knows everyone’s name.

After each meeting, the leader gets an email with a Take Attendance button that opens the form in Church Center; they can also take it from the group’s Events tab. The result is the best per-person attendance view Planning Center ships out of the box: a % column showing each member’s attendance relative to the number of events in the timeframe, a green check per date, and greyed-out columns where the leader didn’t submit anything at all. That is a useful piece of honesty, since it distinguishes “nobody came” from “nobody recorded it”.

Two things to know before you lean on it. Cancelled events and events with no submitted attendance are excluded from the percentage, so a leader who takes attendance half the time produces a flattering number rather than a wrong one. And visitors are removed from the roster after they miss one event, which means the casual attender you most want to notice is the one who quietly falls off the list.

Groups covers a narrow slice of your church, but it’s the slice where somebody already knows the names. What it records and what it doesn’t is worked through in small group metrics.

2. Check-Ins stations

This is the method that changes what you can ask.

A check-in station records the person, the event, the session time, the location, the check-in type and who processed it. Planning Center offers three kinds of station, and picking the wrong one for a room is the usual reason a rollout stalls:

• A manned station is run by a trusted volunteer who can check people in, add new visitors, and edit profile information.
• A self station lets a person check in members of their own household by entering a phone number or scanning a code. No volunteer required.
• A roster station is a digital class list that lets a teacher tick off who arrives, populated from people who previously checked into that location or event.

Roster stations are the underrated one for classrooms, with a caveat worth reading twice: the settings control how many past sessions, up to six, are used to fill the Ready tab. A child who’s been away for seven weeks won’t be on the roster the week they come back, which is precisely the week you want them recorded.

The cost model matters before you plan a rollout. Check-Ins is priced by the number of unique check-ins on your busiest day of the week, usually Sunday, with all plans including unlimited stations, events and administrators, and room to exceed the limit occasionally for spikes like Christmas or Easter. Tiers start at a free plan capped at 10 daily check-ins, then $15 a month for 30, up to $239 for unlimited. So a fourth iPad to shorten the queue costs nothing; checking in every adult multiplies your bill.

That pricing shape is why almost nobody checks in adults, and it’s the single biggest constraint on what your attendance data can tell you.

1. An analytics layer over the top

The top of this list isn’t a fifth way to collect attendance. It’s what you put over the four above so the collection is worth something.

Once you have person-level records, the questions that matter are all comparisons: who came weekly last autumn and monthly now, which first-time guests came back, whether the family that stopped bringing their kids is still giving. Planning Center records all of it faithfully and reports on each product separately, so none of the built-in attendance reports compare a person’s frequency between two periods, and none roll attendance up to a household.

Close that gap with exports and a spreadsheet, or with church analytics software that reads the products nightly and joins them on a person key. Neither route works at all unless the collection underneath records names.

How accurate is your attendance number?

Less accurate than you’d like, and the reason isn’t sloppy ushers. It’s that “attendance” isn’t one thing until somebody defines it.

Start with the gap between what people say and what gets counted. For years Gallup found a steady 40% of Americans saying they had attended worship the previous weekend; when Kirk Hadaway and Penny Marler physically counted attenders at a representative sample of churches, only 20.4% of the population had turned up. Gallup’s own number has since fallen to 21% in a 2024 poll, so the gap has largely closed. But the lesson stands. That’s a survey finding, not a comment on your ushers, but the lesson transfers: a number’s method decides what it’s worth, and two numbers produced by different methods should never sit next to each other.

Then the definition problem, which is bigger than most churches realise. Lifeway Research asked pastors what makes someone a regular attender and found no consensus at all: a plurality (30%) said twice a month, 24% said once a month, 15% said three times a month, and 13% said weekly. They’re split down the middle on whether a stream counts, too: 46% require in-person attendance of some kind while 48% say online qualifies.

So write your own definition down, put it next to the number, and don’t quietly change it. “Unique people who checked in or were counted, all services, adults and children” is a fine definition. So is a different one. What ruins a trend line is redefining it in year three and forgetting you did.

Online is the softest number of the lot, and it matters more than it used to: Hartford’s 2024 attender survey found 26% of participants regularly engage in online worship and 19% move between online and in-person, while noting the frustration leaders feel at knowing so little about viewers on YouTube, Facebook and Vimeo. Lifeway’s own guidance says why in one line: count viewing devices and multiply by average household size, “because no platform does this for you”. Your figure should never be the raw view count.

One honest gap while we’re here: there is no reliable national benchmark for what share of your roster ought to be in the room on a given Sunday. Hartford’s 2025 Faith Communities Today survey puts the median US congregation at 70 regular worship participants, which tells you whether you’re large or small and nothing about whether you’re healthy. The only comparison carrying real information is your church against itself.

How do you set up attendance tracking in Planning Center Check-Ins?

If you’re starting from a notebook, this is the week’s work. Nine steps, in the order Planning Center’s own setup walkthrough takes them.

1. Create the event, then a time inside it. An event is the recurring thing: Sunday Kids, Wednesday Youth. A session is one occurrence. Times sit inside the session; get this right first, because every report you’ll ever run is scoped to an event.
2. Create your locations before your stations. Locations are how the data gets grouped later, and renaming them after six months of history is the kind of chore that makes people stop reporting altogether.
3. Decide the station type per room. Manned at the main entrance where new families arrive and someone needs to add them, self stations for the returning-family queue, roster stations in classrooms.
4. Put the devices in place and set up the printers. Associate each printer with its device and check there are actually labels in it. This is the step that fails on the Sunday, not in the planning meeting.
5. Test every lookup path. People arrive by phone number, name search, barcode, mobile pass, or a Church Center QR code, and choosing a person from search results checks in their whole household. Try each one before a volunteer has to.
6. Set the check-in types you’ll actually use. Regular, Guest and Volunteer. The split is what makes your numbers readable later. A guest count you can trust beats a total you can’t decompose.
7. Stop using label-only mode for guests you intend to follow up. Planning Center is explicit that a visitor added this way will not be added to the database. They get a label and vanish. Use a profile and the Guest type instead, or your guest tracking has a hole in it from day one.
8. Turn on the Services integration for volunteers. When a scheduled person checks in, the integration triggers attendance in both products, so serving attendance costs you nothing extra.
9. Enable Church Center for the event. Families pre-check themselves and their household in the app, which generates a QR code they scan at a station to have labels printed, and an optional Allow Complete Check-In setting skips the station scan entirely. It’s the cheapest way to shorten a queue families have started arriving late to avoid.

Budget a Saturday for setup and expect the second Sunday to go better than the first. Then leave it alone for a quarter before you change anything, because your first month of data is mostly people learning where to stand.

How do you count the adults who never check in?

This is the question every church hits, and there’s no clever answer, only a workable one.

You count the auditorium with Headcounts, because it’s the only realistic method for a room where people walk in and sit down, and you accept that this number has no names attached. Then you treat the named data you already have as your person-level sample:

• Kids checked into children’s ministry. A checked-in four-year-old implies a household in the building. Usually. It does not tell you which adult, and sometimes none: a drop-off, a grandparent, a friend’s parent. Treat it as a household-presence signal, never as a named adult check-in, and don’t write an adult’s attendance record from it. With that limit, it is the most reliable household attendance signal most churches own, and it’s already being collected for safeguarding reasons.
• Volunteers checked in to serve. With the Services integration on, every scheduled person who shows up is recorded by name.
• Group members marked present. Midweek, by their leader.

Between the three, most churches know far more adults by name than they think, week after week, which is what reading a rhythm needs. The blind spot is real, though: empty-nesters and single adults who don’t serve and aren’t in a group. If those people matter to your follow-up, a lobby self station is the only route, and it will cost you on the check-in tier.

Online attendance is the other half of this question, and the accuracy section above covers why those numbers are soft. Report it separately; never fold it into the in-room count. A composite figure is the one that gets quoted for years and means nothing.

What do you do with the numbers once you have them?

Collection is the boring half. Here’s what the data is for once the tracking is running, and what each answer costs.

Three habits are worth forming early.

Count unique people, not check-ins. A family of four generates four rows and one household. A volunteer who serves at 9am and attends at 11am can appear twice. Decide which you’re reporting and keep it consistent, because the day someone changes it quietly is the day your history stops being comparable.

Compare to the same week last year, not to last month. A month is mostly weather, school holidays and one long weekend. Reading a twelve-month attendance line properly is a skill: the summer dip, the September return, the fifth-Sunday months that make a monthly total look wrong. I’ve written it up separately in church attendance trends.

Read it by household, not by head. Eleven families leaving and eleven visitors arriving is a flat total and two very different churches. That arithmetic is why a steady attendance number can hide a slow departure for months.

If you want to build these views yourself, exporting Planning Center data to Google Sheets is the honest starting point, and first-time guest tracking covers the one metric that’s worth setting up even if you do nothing else on this list.

Where attendance tracking stops

Good tracking answers how many completely and who partially. It never answers what changed, because that’s a comparison between two windows for every person at once, and nothing in Planning Center is arranged to run it.

That’s the specific gap Parable fills. It reads every Planning Center product nightly into one place where check-ins, giving, groups and serving share a person key, so a household is one row instead of four reports, and “who came weekly last autumn and hasn’t checked in since Easter” is a question you type. The recurring figures sit on a church metrics dashboard that a board can read without anyone rebuilding a spreadsheet the night before, alongside the rest of the metrics that actually matter.

But the order here is not negotiable, and it’s the reason this post ranks collection above analysis. No layer invents data that was never collected. If your church has counted heads for ten years, the analytics you buy tomorrow can tell you the totals moved and nothing whatsoever about who moved them.

Start recording names in the rooms where you can. The reports can wait; the history can’t be backfilled.