A report is three choices
Not a dashboard someone else designed for a gym that is not yours. You build the question yourself, from three parts: a measure (revenue, active members, attendance, check-ins, trials converted, ten in all), a split (by plan, discipline, location, month, belt, source and more, fourteen in all), and any filters you want. The answer renders as a chart or a table, whichever reads better.
Swap any slot and the question changes. Revenue by discipline becomes revenue by location. Trials converted by source becomes trials converted by month. Attendance by belt tells you whether your blue belts are drifting. Same three moves every time.
Ten measures, fourteen ways to slice them
Ten measures against fourteen dimensions is 140 base questions before you have touched a filter. That sounds like a lot until you notice each one is something an owner has actually asked out loud: which plan brings in the money, which source sends members who stay, which location fills its mats.
Keep the good ones
Most owners end up with three or four reports they actually live by. So the builder does two things with a finished report: pin it, and it waits for you at the top of the screen every time you come back, or schedule it, and it lands in your inbox every Monday. The question you asked once becomes a habit you never have to remember.
Rates that do not lie
Here is where most reporting tools quietly cheat. Ask for a conversion rate split by source and the lazy way to compute it is to average the per-source rates. That number is wrong whenever the groups are different sizes, which is always. MatPilot computes every rate from the underlying counts: total converted over total trials, weighted the way arithmetic actually works.
Why this is worth a whole section. A wrong rate does not look wrong. It looks like a dashboard, with a confident number on it, and you make a marketing decision off it. Computing from the counts is invisible when it works, which is exactly the point.