What We Decided Not to Count

Every dashboard is an argument about what matters. That is not a side effect of building one. It is the primary thing a dashboard does — the numbers are secondary to the claim, implicit in the layout, that these are the quantities worth your attention on a Monday morning.

Which is why the hardest product decisions in church software are not about what to measure. They are about what to leave out.

The easy numbers are easy for a reason

Attendance. Giving. App opens. Messages sent. Volunteer signups. Every one of these is cheap to collect, moves visibly week over week, and produces a chart that looks like feedback. They are the first things any competent engineering team would instrument, because they are already sitting in the database as a side effect of the software working at all.

So they get surfaced. Not out of any philosophy — out of availability. And then something happens that nobody chose and nobody notices for about a year.

A pastor opens the software on Monday and sees a line going up or down. Then again the next Monday. Then again. And gradually, without a single decision being made, that line becomes the scoreboard. It is the thing that shows up. Things that show up every week become the definition of the job, and by the time anyone examines the assumption, a great deal of Tuesday has been organized around moving a number that was instrumented because it was convenient.

I do not think this is a church problem. It is the oldest failure mode in management software, and I have been on the wrong side of it in other industries. But it lands harder here, because the work being measured is not throughput. It is people.

Three things we decided not to build

We do not rank congregations against each other. This would be straightforward to build, and pastors do ask for it. Every SaaS product eventually ships benchmarking, because it is engaging and it makes the data feel more valuable.

Here is why we have not. A church of two hundred in a shrinking town, thirty years into a faithful ministry, compared against a church of two hundred in a growing suburb, four years old, is not receiving information. The two are not doing the same thing under different conditions; they are doing different things. A percentile is an answer to a question that was never well-formed, and its practical effect on a pastor at eleven at night is not insight. It is shame with a chart attached. I have watched that specific dynamic play out in other industries and I will not build the version of it that operates on someone's calling.

We do not put an engagement score on an individual person. We could compute one. The data is there, it would be reasonably accurate, and every product instinct I have says it would demo beautifully.

But think about what happens the week after it ships. A number exists that sorts a congregation from most to least engaged. It is right there, sorted, on a screen belonging to someone with limited hours and genuine care for everyone on the list. That number will decide who gets a call. Not because anyone decided it should — because it is the only ordering available, and an ordering that exists gets used.

I have not met a pastor who wants an algorithm making that choice. Several have told me they would want to know who has gone quiet, which sounds like the same request and is not. "Who has gone quiet" is an observation that returns the decision to the pastor. "Who is least engaged" is a verdict that has already made it.

We do not report on prayer. A prayer list is the most sensitive data in a church building — addictions, diagnoses, failing marriages, immigration status, things people have told exactly one other person. The instinct to summarize it is understandable and, I think, exactly wrong. Some things lose their nature when they become a trend line. Whatever a chart of prayer requests over time is, it is not pastoral care, and its existence would change what people are willing to write down. The moment a congregation suspects their prayers are being counted, they will start filing the safe ones.

What we surface instead

The alternative is less impressive-looking, which is exactly the problem with it commercially. It does not chart. It does not demo. It is mostly a list.

Who has gone quiet — not scored, just noticed, with the interpretation left where it belongs. What is waiting on a decision only the pastor can make. Which volunteer asked a question four days ago and has not heard back. What is scheduled this week and what is not staffed. Where the last thing someone tried to do in the software failed and they gave up.

None of that is analytics. It is closer to a well-kept desk than to a business intelligence tool. And the honest reason we can build it that way is that we are not selling to a board that wants a growth curve. That is a real advantage and I am aware it is partly circumstantial.

The counterweight I have to hold

I want to be careful, because the argument I am making has a lazy version that I do not believe.

The lazy version says numbers are worldly and the church should not measure anything. That is not humility; it is a refusal to be accountable dressed up as reverence. Churches that count nothing are not more spiritual than churches that count carefully — they are simply unable to notice that giving has fallen for six months, or that the young families who came last spring all quietly stopped coming, or that the same four people are doing everything and are about to break. Refusing to look does not protect anyone. It just delays the discovery until it is expensive.

There is also a version of my position that is just cowardice about hard conversations. Some numbers are uncomfortable because they are true. A pastor who does not want to see the giving trend may need to see the giving trend. Declining to build something is not automatically wisdom, and I have to keep testing my own restraint against the possibility that it is really just avoidance.

So the standard I try to hold is narrower than "measure less." It is: does this number return a judgment to the person, or make one on their behalf? Does it describe a situation, or rank a soul? Would I want it on a screen at eleven at night when someone is tired and discouraged and looking for evidence about their own worth?

Every dashboard teaches. Ours will teach something whether we are deliberate about it or not. The only real choice is whether we picked the lesson on purpose.