Monday's staff meeting shouldn't start by rebuilding Sunday.

The first twenty-five minutes

Everybody sits down. Somebody asks how yesterday went.

What follows is not a conversation, it is a reconstruction. One person has a head count they made from the platform, which is famously the worst place in the building to count from. Someone else has the children's number but only for the second service. The offering total is in a different system and the person who knows how to get it is not in the room. Two of these figures disagree and nobody can tell which one is wrong, so the group does the reasonable thing and starts discussing the discrepancy.

By the time everyone agrees on what happened, half the meeting is gone and the agenda has not started.

The instinct here is to ask for a better report. That is the wrong ask, and it is worth understanding why before you go shopping for one.

A number nobody can check is a number nobody will act on

What decides whether a summary is useful has almost nothing to do with how much is in it.

When a figure appears in front of a group, the group does one of two things with it. If they can see where it came from, they argue about what it means, which is the argument you wanted. If they cannot, they argue about the figure, which is the argument that eats the meeting.

Say the sheet reads forty-one children checked in, against a four-week average of thirty-eight. Every person at that table can reconstruct that number. Four Sundays, add them, divide by four, compare. Nobody has to trust anybody. The conversation moves immediately to whether three more children is weather, a visiting family, or the new class, and that is a conversation worth having on a Monday.

Now say the sheet reads that engagement is trending positively. Nobody can check that. Nobody knows what went into it or what it left out. So either it gets believed without examination, which is worse than useless, or somebody asks where it came from and you are back to spending the meeting on the number instead of the ministry.

The good version is arithmetic. Counts, one or two sums, an average, a sort. An algorithm plain enough that a suspicious treasurer with the church's own records and a calculator could reproduce every line of it before lunch. That sounds like a modest ambition. It is actually the whole thing, because it is what lets a number be said out loud in a room without anyone having to take it on faith.

A mirror, not a scoreboard

There is a second failure, and it is the more serious of the two because it does damage to people rather than to meetings.

A weekly summary invites comparison. Numbers in a row on a Monday morning start to feel like a report card, and the people reading it are usually tired and often the same people responsible for the numbers. It takes about three weeks for a helpful page to turn into a page somebody dreads.

The discipline that prevents it is refusing to let a figure carry a meaning it has not earned. A zero in the guest column does not mean nobody new walked in. It almost always means nobody wrote anybody down. A giving total that looks thin does not mean the church is faithless; it might mean the count was Tuesday. If your summary cannot tell the difference between a low number and an unrecorded one, it should not be permitted to imply one.

Which is why the honest framing for the whole exercise is a mirror rather than a scoreboard. It shows you what got recorded. What it does not show you is what happened, and a staff that keeps the difference in mind will get more out of a plain page than a staff that has quietly started performing for a chart.

The sections that ought to be missing

Most weekly summaries show you everything every week, because that is easier to build and it looks thorough in a demo.

It is also how a summary teaches people to skim it. If nine sections appear every Monday and six of them are usually zero, the eye learns within a month that most of the page is noise, and then it starts skipping the three that were not.

A section with nothing to say should not be there. A quiet week should produce a short page, and the shortness itself is information: you can see at a glance that this was an ordinary week, without reading a single number. And when a section does appear that has not appeared in a while, it has your attention automatically, because its presence is the signal.

The same logic argues for a length limit. If it prints on one side of one sheet of paper, somebody will bring it into the room and put it on the table. If it runs to three pages, it will live in a browser tab that two of the five people meant to open.

Watch the team, not only the congregation

Every church summary ever built counts what the congregation did. Attendance, giving, guests, groups. Almost none of them count what the staff and the volunteers did, and that is a strange omission given who is in the room reading it.

Two facts sit in your records already and nobody looks at them. Who has served four, five, six Sundays in a row without a break. Who is quietly carrying five open follow-ups while everyone else is carrying one.

Neither is a crisis. Both are the sort of thing that becomes a crisis eight weeks later, when a good volunteer resigns by email and everyone says they had no idea, and they are telling the truth, because the pile was never anywhere anyone could see it.

So a summary should notice, say it once, plainly, in the same flat voice it uses for everything else, and then do absolutely nothing about it. No alert. No red. No suggestion of what to do. The software is qualified to observe that someone has served four Sundays running; it is not remotely qualified to know whether that person is exhausted or having the best month of their life. That judgment is why the humans are in the room.