First-time guest tracking is only worth doing if you measure two things afterwards: the share of guests contacted within 48 hours, and the share who came back within four weeks of their first visit. Planning Center records the first visit well. It reports neither rate.
Key points
- Planning Center is explicit that label-only visitors aren’t saved to your People database and won’t appear in People list results or Home widgets. So the fastest way to check a guest in is also the way that erases them.
- A custom report scoped to first timers only returns people who attended that event for the first time, which is not the same as attending your church for the first time.
- The Home dashboard’s first-time visitors widget shows a first check-in for a specific event or a first check-in ever, and excludes one-time visitors entirely.
- Follow-up rate is measurable only if the contact gets recorded somewhere; a phone call nobody logged is indistinguishable from a call nobody made.
- There is no credible national benchmark for guest follow-up or return rate. Three months of your own numbers is worth more than any figure you’ll read, including in this post.
Every church I’ve worked with can tell me how many guests came last Sunday. Almost none can tell me how many of last March’s guests are still around.
That’s not a discipline problem. The count is a number Planning Center hands you. The other is a question about two points in time for the same set of people, and no report is shaped like that. It’s the same gap behind every report Planning Center can’t run.
What counts as a first-time guest?
Three different things get called a first-time guest, and Planning Center gives you all three without saying which one you’re looking at.
First time at this event. The custom report scoped to first timers only generates, in Planning Center’s own words, a report of people who attended the event for the first time. A family who came to VBS in July and turns up in Sunday Kids in September is a first timer in September. Read the flag as new to this room.
First time anywhere. The Home dashboard widget is more precise than the reports are: it shows people if it’s their first check-in for a specific event (even if they’ve attended other events) or their first check-in ever. You pick. That’s the closest thing to a real first-visit flag in the product, with a caveat in the same sentence: one-time visitors are not included.
First time in the building. Nothing records this, and it’s the one everybody means. An adult who has slipped into the 11am service for two months and finally brings a child to Kids ministry reads as brand new on every report you can run. If your weekend adult attendance is a headcount rather than a check-in, a large share of your real first-timers stay invisible until they hand over a child or fill in a card. And for most churches it is, because Headcounts records numbers without names.
Pick one definition, write it down, use the same one every month. A return rate computed against a moving denominator is worse than no return rate, because it looks like data.
How do you record a first-time guest so you can find them later?
The decision that determines everything downstream is made at a station, by a volunteer, in about four seconds, and it’s close to irreversible.
Label-only mode exists for the genuinely one-off visitor: a cousin, a child dropped off by a neighbour. What it does not do is keep anything. Planning Center documents that visitors aren’t saved to your People database, that visitor contact information exists only for the event, and, in bold in the same article, that visitors will not be included in People list results or Home widgets.
The same article gives the fix in a line: if you plan to follow up with someone after the event, create a new profile and check them in with the Guest check-in type instead. At a manned station that’s typing the name, finding nothing, and selecting Add Person, after which Regular, Guest and Volunteer are the three statuses available.
Four seconds of typing is the difference between a guest you can measure and a guest who existed only in one session’s report.
| How the guest arrives | Profile saved | In lists and widgets | Can you follow up |
|---|---|---|---|
| Label-only visitor at a station | No | No | Only during the event |
| New profile, Guest check-in type | Yes | Yes | Yes |
| Connect card entered into People afterwards | Yes | Yes | Yes, but with no check-in record |
| Counted in a Headcounts total | No | No | No |
| Walked in, sat at the back, left | No | No | No |
That last row is not a joke. For most churches it’s the largest row in the table, which is why a guest number is a floor rather than a count.
Two smaller fixes for the same Sunday. Add Primary contact to your first-timers report under Include, or you get a page of four-year-olds and no way to reach a parent. It’s the most common wasted report in Check-Ins.
And decide who owns the connect cards before you print them. A card typed into People on Wednesday creates a profile with no check-in attached to it. Digital cards have their own trap: Planning Center notes that a Forms submission whose email address doesn’t match an existing profile creates a duplicate rather than updating one, which is how a returning guest becomes two first-timers.
Which two numbers matter?
Guest counts move with the weather, the school calendar and whether anyone had a baby dedication that week. Two derived numbers don’t.
Follow-up rate is the share of recorded guests who received a real contact within 48 hours. It measures your process, and it is the only one of the two you fully control.
Return rate is the share of a month’s first-time guests who checked in again within 28 days of their own first visit. Each guest’s clock starts on their own Sunday, not at month-end. It measures your church, and it is the one that tells you whether the welcome held once the music stopped.
The windows are choices, not findings. I use 48 hours because a hard deadline forces the contact to be assigned on Sunday rather than remembered on Thursday. I use four weeks because a church that meets weekly has given a guest four chances by then, and one missed Sunday out of four is normal life rather than a verdict.
| Metric | Formula | Window | Where the data is |
|---|---|---|---|
| First-time guests | Distinct people whose first check-in falls in the period | Weekly and monthly | First-timers-only custom report; Home first-time visitors widget |
| Follow-up rate | Guests contacted ÷ guests recorded | 48 hours | The workflow CSV export in People, taken as a snapshot 48 hours after Sunday. Nothing else records the contact |
| Return rate | Guests with another check-in within 28 days of their own first visit ÷ guests recorded | 28 days from each guest’s first visit | Person-date check-in export (exact); bounded two-condition People list (approximation) |
| Connection rate | Guests in a group or on a team ÷ guests recorded | 6 months | Groups and Services, joined on Person ID |
The fourth row is one to grow into rather than start with, and its mechanics live in small group metrics.
Notice what the follow-up row depends on. A call nobody logged is indistinguishable from a call nobody made, so if follow-up lives in a group chat or in someone’s head, the rate does not exist. Choosing where the contact gets recorded is the measurement. Put it in a workflow and the rate becomes a download: a workflow’s CSV export covers the 10,000 most recent cards with the date each card was added, total days in the workflow, days in the current step, card status and days overdue, plus an optional checkbox for completed cards. One thing it does not carry is when a card moved between steps, because “days in current step” resets on every move. So a card exported after it has advanced can’t tell you whether the first contact happened on day one or day five. The fix is to make the export a snapshot rather than a history: make the first step the contact itself, and export on Tuesday evening, 48 hours after Sunday. Among that Sunday’s cards, the ones still sitting in step one are the guests nobody reached in time, and the ones that have moved on are the contacts made within the window. That division, taken every week and written down, is the 48-hour rate.
How do you compute a first-time guest return rate in Planning Center?
You can get most of the way there with built-in tools. Here’s the version that works, with the parts that don’t.
- Fix the cohort. Everyone whose first check-in fell in a single calendar month. Monthly, not weekly. A week’s guests at a 300-person church is small enough that one family swings it ten points.
- Build the first-timers list. Planning Center’s common attendee lists documents the shape: children who first attended one of a few select locations in the last five days, combined with a second rule for profiles created in the last five days. Widen the five days to your cohort month.
- Decide whether to keep the second rule. The same article notes you can run the list without it to catch people who have been part of your church a while but only just started attending that event. Keeping it means new to the church; dropping it means new to the room. The definition choice from earlier, expressed as a checkbox.
- Add the return condition, and bound it. A list cannot say “within four weeks of their first visit”, because conditions take fixed dates, not dates relative to each person. So use the nearest bounded approximation: a second check-in at that event between the first day of the cohort month and four weeks after its last day, and run the list once, on that closing date. Run it later and it counts returns that happened months on; run it as “any check-in since the month ended” and it misses the guest who came back the following Sunday. For the exact per-guest window, step 8 is the route. Lists let you require that people meet all, any or none of the conditions in a rule, which is the mechanic doing the work. Watch the date arithmetic: Planning Center’s own tip is that a since x days condition means today plus the previous x days, so an off-by-one puts last Sunday in both cohorts. That two-condition list is the nearest thing to a return report Planning Center has, and it’s documented alongside the first-timer workflow.
- Build the inverse with the same bounds. Same first condition, and the step 4 return condition negated, meaning no second check-in between the cohort start and four weeks after the cohort end. Run it on the same day. Not “no check-in since the month ended”: a guest who came back on the second Sunday and then stopped would count as a return in step 4 and still land here, and the two lists would no longer add up to the cohort. This is the list you’ll actually use, because it’s the people still worth a call.
- Export both and keep the Person ID column. A list under 500 results downloads immediately; a larger one is emailed. Match on ID, never on name. Planning Center confirms both halves of why in one place: exports include a Person ID unique to each profile, while the Household ID is not saved to the profile and exists only inside an import. Person is your join key; household is something you rebuild by hand.
- Write the two counts into a sheet, by hand, the same day each month. The step that feels beneath you is the one that matters.
- For an exact rate, or the trend rather than the month, export check-in records and pivot by person. With one row per person per date, each guest’s first date and next date sit beside each other, and “returned within 28 days” is one subtraction per row rather than a list condition. It’s the same method as any period-over-period comparison, covered in the attendance reports guide and easier to live with if the exports land in a sheet on a schedule.
Step 7 exists because of a specific limitation. People lists take dated conditions but hold no history: a list run today describes today, and there is no version of it as it stood in March. Don’t write March’s number down in April and March’s return rate is gone unless you rebuild it from raw exports. The difference between a list, a report and a dashboard is worth understanding before you design any of this, and it’s laid out product by product.
What does a first-time guest follow-up workflow look like?
Follow-up rate is measurable at all because Planning Center gives the contact somewhere to live. A workflow, in its own description, tracks people’s progress toward multi-step goals or keeps tabs on your communication history with them. That second half is the point here. Both the common attendee lists and common lists articles point at the same pattern: build the list, set it to refresh nightly, add an automation that drops the results into a workflow where someone follows up.
| When | What happens | Who | Recorded where |
|---|---|---|---|
| Sunday morning | Guest gets a profile and a Guest check-in | Station volunteer | Check-Ins |
| Sunday night | List refreshes, automation opens a workflow card | Nobody. It’s automatic | People workflow |
| Within 48 hours | A named person makes contact, in their own words | Assigned staff or volunteer | Workflow card |
| Week 2 | Second touch, only if the first one got a reply | Same person | Workflow card |
| Week 4 | Check whether they returned; close the card either way | Workflow owner | Workflow card |
| Week 6 | Two visits or more, hand to groups | Groups lead | Groups |
Three notes on running it.
Set the overdue timeframe to two days. Building a step lets you choose Mark overdue after and pick a timeframe, and cards that exceed it get an overdue badge you can filter the step view by. Set it to 48 hours and your follow-up standard stops being a value and becomes a filter. The assignee is notified in their admin inbox and by email, so the reminder doesn’t depend on anyone remembering. Assign to a person, too. A card assigned to “the welcome team” is assigned to nobody.
Close the card either way. An open card is not a follow-up in progress; after three weeks it’s a follow-up that didn’t happen. Leaving cards open is how a follow-up rate quietly reports 90 percent while the real number is half that.
Automate the list, not the message. The nightly refresh and the automation belong to a computer. The message doesn’t. A guest can tell a person from a template in about one sentence, and the second visit turns on that sentence more than on anything else on this page.
Is there a benchmark for church visitor follow-up rate?
No. And I looked properly, because it would be a useful thing to publish.
Nobody collects it. Lifeway Research, Barna, Pew, Gallup, the Hartford Institute and the National Congregations Study all publish serious church data, and none of them measure what share of first-time guests return, or what share of churches contact a guest afterwards. Lifeway’s 2024 survey of 1,003 Protestant pastors gets closest by measuring the mechanism: 80% have printed cards guests are asked to complete to enable follow-up, while 38% have an online form. It never asks whether the card produced a phone call.
The nearest proxy is Lifeway’s 2023 study of 1,001 churchgoers who had changed churches: more than 7 in 10 (71%) decide to regularly attend their current church after visiting worship services three times or less, 20% after a single visit and 50% after two or three. Good argument for a four-week window, bad benchmark, for a structural reason: everyone in that sample joined a church. It says nothing about the guests who never came back, because none of them are in it.
You’ll also see a 24-hour rule quoted as though it were a finding. It isn’t. Lifeway’s own advice piece on the subject states that a personal follow-up within 24 hours will separate your church from the majority and attaches no figure, sample or study to it. Treat 24 hours or 48 hours as an operating standard somebody chose, and pick the one your team will hit.
What the research does give you is scale, and it’s smaller than most guest-tracking advice assumes. The National Congregations Study’s 2018-19 wave puts the median congregation at 70 regularly participating people, with a median of 50 regularly participating adults, and the Hartford Institute’s 2025 survey of 24,165 attenders cites earlier congregational research estimating the average congregation gains about 5% of new attendees per year. Multiply the two and the median church gains three or four new attenders a year. That’s my arithmetic, not theirs. A return rate with a denominator that small is a story rather than a statistic, and should be read as one.
So build your own. Three months of your own return rates, computed the same way each month, beat any national figure, because they’re measured against the only comparison that matters: you, last quarter. Make the first month the baseline. If the second is five points better, whatever you changed worked.
Two sanity checks. Read it monthly rather than weekly, for the same reason you’d read attendance as a trend rather than a week. And read guests beside the people already there, because a church can post a healthy return rate and still shrink if it’s losing regulars faster than it gains guests. That’s the argument for tracking a small set of metrics together rather than one at a time.
What guest tracking still can’t tell you
Four gaps, and they’re the four that turn a monthly report into a monthly argument.
The household, not the person. A family of four who visit once produce up to four first-timer rows and one decision. Nothing rolls them up, and household ID doesn’t survive an export, so collapsing them is a manual pass with a surname sort.
The guest you never recorded. Every number here is computed over guests who got a profile. Improving capture makes your return rate look worse before it looks better. Worth saying out loud before the chart appears.
The rate over time. No report compares this month’s cohort against last month’s, so the trend exists only in whatever sheet you’ve been maintaining by hand.
What happened after week four. A guest who returned twice in April and vanished in June looks identical to one who stayed, because the window closed before the interesting part.
That last set is why we built Parable the way we did. It reads Planning Center every night into one place where check-ins, giving, groups and serving share a person key, so a guest cohort is something you can follow forwards: March’s first-timers, how many came back, how many were in a group by September, with names attached rather than percentages. The recurring version sits on a church metrics dashboard beside the numbers a board asks for monthly, and which ones those should be is worked through in church health metrics for your board.
Under a couple of hundred people you don’t need any of this. You’ll remember the family from three weeks ago, and remembering is better than a report.
Past that, you won’t, and neither will anyone else. The guest count will keep telling you the door worked. Only the second visit tells you the room did.