Attendance

Comparing this year's attendance to last year, the right way

Easter moves, months have four or five Sundays, and a raw year-over-year number can lie to you without meaning to.

8 min read

Ask AI · in the $19/mo plan

Ask your own records a question. Is attendance actually trending down, or was that one week?” — answered from the records you already keep. It reads your church and no other, and it can't invent a number.

10 questions a month included · no AI add-on to buy

Someone on your board asks how attendance this year compares to last year, and the honest answer is more complicated than pulling two numbers off a chart and subtracting. The calendar itself moves under your feet. Easter is not on the same date twice in a row. Some months hand you four Sundays and some hand you five. A year-over-year comparison that ignores this will tell you a story that has nothing to do with how your church is actually doing.

None of this means the comparison is useless — it means it needs a little care before you trust it. Here is what actually moves the number, and how to line things up so what is left over is signal instead of noise.

Easter is the biggest variable you are not accounting for

Easter can fall as early as March 22 or as late as April 25, a range of more than a month. Most churches see their single highest attendance of the year on Easter Sunday, so where it lands changes which month gets the bump. Compare April of one year to April of the next, and you might be comparing a month that contained Easter to a month that did not — or the reverse. The difference you are measuring is the calendar, not your congregation.

The fix is simple once you see the problem: compare Easter Sunday to Easter Sunday, and compare the Sunday count for the season around it rather than the specific month name. If you track attendance by season — Advent, Lent and Easter, ordinary time in summer, fall launch — the comparison holds up regardless of which weekend the date lands on.

Some months have five Sundays and some have four

A month with five Sundays will almost always show higher total attendance than a four-Sunday month, even if the church is exactly as full each week. If you are comparing monthly totals instead of weekly averages, a five-Sunday March this year against a four-Sunday March last year will look like eight to ten percent growth that never actually happened. It is arithmetic, not momentum.

The safer unit is the per-Sunday average for the period, not the sum. A month that averaged 140 people across four Sundays and a month that averaged 140 people across five Sundays represent the same church. Only the total looks different. Whenever you catch yourself comparing raw monthly totals year over year, check the Sunday count first.

Holidays land on different days of the week

Christmas, Independence Day, Labor Day weekend — these shift across the days of the week from year to year, and each one pulls attendance down a little when it lands adjacent to a Sunday and a family takes the whole weekend for travel. A year where Christmas fell on a Sunday itself is not comparable to a year where it fell on a Tuesday. Neither is wrong. They are just different years to be a church trying to have a normal service.

When you build your comparison, mark the Sundays that sat next to a major holiday in each year and treat them as a separate category rather than folding them into the general average. That keeps one unusual weekend from dragging the whole year’s baseline in a direction that has nothing to do with your ministry.

A single bad Sunday is not a trend, even repeated across years

Every congregation has had the Sunday that a snowstorm, a stomach bug going through the school, or a scheduling conflict with a community event cut attendance by a third. If that happened once last year and not this year, comparing those two specific Sundays will make this year look like it grew by a third, when really one year had bad luck and the other did not.

This is exactly why a rolling average across several weeks is more honest than any single Sunday, in either year you are comparing. Reading attendance well starts with resisting the pull of any one number, whether it is unusually high or unusually low, and looking instead at where the average has settled over a run of weeks.

Build the comparison around a stable unit, not a stable date

The most reliable way to compare years is to stop anchoring on the calendar date and anchor on something that does not move: the number of Sundays in a defined season, the average per Sunday within that season, and a consistent definition of what counts as “attended” from year to year. If you changed how you count — started including a second service, began counting a Sunday school hour separately — note the change and treat the years on either side of it as not directly comparable until enough time has passed to smooth it out.

A church of 150 that averaged 145 per Sunday across Lent one year and 138 per Sunday across Lent the next year has a real, if modest, question to ask. A church that compared a five-Sunday March to a four-Sunday March and saw the same ten-person swing has an arithmetic artifact, not a question. The difference matters, and it is only visible once you have controlled for the calendar.

What to do once you trust the number

Once you have accounted for Easter’s date, the Sunday count, and the holiday placement, a real year-over-year change is worth taking seriously — but it is still a starting question, not a verdict. A softening trend is worth pairing with what else is happening: are fewer guests showing up in the first place, or are the guests who show up not returning? SundayBridge charts attendance against its own history so the comparison is there to look at, but the reading still has to account for the calendar quirks above before you draw a conclusion from it.

If the softening coincides with fewer guests turning into established next steps, the story might be less about the Sunday count and more about what happens after a big Sunday — whether the people who visited once ever heard from anyone again. Attendance is the visible symptom; the follow-up process is often where the real answer is hiding.

Keep the underlying count honest all year

Every comparison is only as good as the numbers feeding it, and those numbers come from whatever happens at the door each week. A consistent check-in process matters more for year-over-year comparisons than it seems like it should, because a change in how carefully attendance is recorded — a new volunteer at the table, a busier lobby, a season where nobody quite got around to it — can create a swing in the data that has nothing to do with who actually walked in. If this year’s count came from a tighter process than last year’s, the improvement in your chart may be an improvement in your record-keeping, not your church.

The same caution applies to big single-day events. A well-planned Easter service or a fall kickoff Sunday can be genuinely larger because of real effort put into planning the event well, and that is a legitimate reason for growth — just keep it labeled as an event effect rather than folding it silently into the ordinary-Sunday average for the year.

A year-over-year chart is only as trustworthy as the notes sitting next to it. If the Wednesday night gathering moved to Sunday afternoon, if a satellite campus opened or closed, if the church added a second service and started counting it separately — any of these will move the number in a way that has nothing to do with whether people are showing up more or less. Keep a short, plain list of what changed each year, right alongside the attendance figures. It takes five minutes and it saves you from arguing about a number that was never really comparable in the first place.

The same goes for staffing changes on the volunteer side. A church that lost its greeter team lead for a season and picked the role back up eight months later may see a dip in guest attendance that looks like disengagement but is really a gap in who was watching the front door. Attendance numbers do not explain themselves — the context around them does, and that context lives in whoever remembers what actually happened that year, not in the chart.

The comparison that actually holds up

None of this is complicated once it is named. Compare seasons, not months. Compare Sunday averages, not Sunday totals. Note the holidays and the outlier weeks in both years before you subtract one from the other. Do that, and a year-over-year number stops being a source of false alarm or false comfort and becomes what it was supposed to be all along: a reasonably honest picture of whether your church is gathering more people, fewer people, or about the same as it was a year ago.

Frequently asked questions

Should we compare Easter to Easter, or the same calendar date every year?
Compare Easter to Easter, and every other Sunday to the Sunday nearest the same point in the church calendar, not the same date on a wall calendar. Easter can land anywhere between March 22 and April 25. Lining up a March 30 service one year with an April 20 service the next just because both fell on a fifth Sunday tells you nothing useful.
How many years of history do we actually need before a comparison means anything?
Two years will show you a direction, but three or four is where the picture steadies. One prior year cannot tell you whether last year was the anomaly or this year is. If you are just starting to track attendance carefully, keep going — the comparison gets more honest every year you add, not less.
What about a year with a building closure, a pastoral transition, or a weather disaster?
Note it plainly next to the number and do not let it silently drag your baseline. A single unusual year — a flood, an interim season, a snowstorm that closed three Sundays in a row — should be flagged as an outlier rather than averaged in as if it were ordinary. Otherwise you end up comparing a normal year to a damaged one and drawing the wrong conclusion.
Is a five percent year-over-year drop something to worry about?
On its own, no. A five percent swing sits well inside the range that a moving holiday, a five-Sunday month, or a single bad-weather week can produce by itself. Worry when the same direction shows up two or three years running, or when it shows up alongside other signals — fewer new guests, fewer people serving, giving trending the same way.