What an attendance percentage hides when drop-ins and comps share a roster

Published November 3, 2026

The number that looks clean and explains little

A studio owner asks a simple question: how full were our classes last month? The platform reports an attendance percentage by dividing bodies in the room by capacity. The number looks authoritative because it comes from a dashboard. It can still be misleading if the denominator counts seats that were never sellable, or if the numerator includes people who did not occupy a paying spot.

Drop-ins, complimentary passes, staff training spots, instructor guests, and member make-up sessions all change what the percentage means. A class can show 92 percent attendance while two of the twelve bodies are staff and one is a guest who will never be billed. Another class can show 60 percent while every sellable spot was taken and the remaining capacity was held for equipment or safety spacing.

This note describes a measurement review a fitness virtual assistant can prepare for a studio manager. It does not propose an industry standard. It proposes a way to define the denominator, classify the numerator, and publish a result that a manager can defend when comparing one class, one coach, or one month with another.

Start with the decision the number will support

Before exporting attendance data, write down what decision the number will inform. A decision to add a class needs a different measure than a decision to change a coach's schedule. A decision to market a slow time slot needs a different denominator than a decision to evaluate whether a waitlist policy is working.

If the decision is capacity planning, count sellable spots and distinguish them from physical capacity. If the decision is member engagement, count unique members who attended rather than total check-ins. If the decision is revenue, attendance is only a proxy and should be paired with the billing record. A single percentage cannot answer all three.

Write the question in one sentence and name the population. For example: among reformer classes at the downtown studio in October, what share of sellable spots were occupied by members or paying drop-ins at the scheduled start time? That question already excludes staff spots, comps, and late arrivals, which changes the result.

Separate physical capacity from sellable capacity

Physical capacity is the number of people the room can safely hold under the studio's layout and supervision rules. Sellable capacity is the number of spots the studio actually offers for booking after subtracting equipment that is out of service, instructor needs, and any blocked spots.

A reformer studio may have twelve machines but only ten bookable if two are reserved for a beginner class format or a coach demonstration. A cycling studio may have thirty bikes but block two near the instructor's platform. A yoga studio may reduce capacity when the room is configured for a workshop. These are not accounting tricks; they are operational facts that belong in the denominator.

Record the capacity version and the date it applied. If the studio changed the layout in the middle of the month, split the reporting period at the change date instead of averaging across two different rooms. A capacity change is a legitimate reason for a step change in the percentage, and hiding it inside a monthly average makes the trend unreadable.

Classify every body in the room

A usable attendance measure needs a small set of mutually exclusive categories. At minimum, distinguish members attending against their plan, paying drop-ins, complimentary guests, staff or contractor training, and instructors. Add a category for make-up sessions if the studio allows them, because a make-up can occupy a sellable spot without producing a new sale.

The classification should come from the booking record, not from a guess at the door. If the platform cannot distinguish a comp from a paid drop-in, the studio has a data dictionary problem before it has a measurement problem. Fix the booking type or add a required field at check-in rather than asking the front desk to remember.

Avoid counting the same person twice when they leave and re-enter, or when a booking is transferred to another member. Use the booking identifier or the check-in event, and state which one the count relies on. A count of bodies at a moment in time and a count of check-in events are different measures; both can be useful, but they should not be mixed without a note.

Decide how to treat no-shows and late cancels

A seat that was booked and not used is not the same as a seat that was never sold. If the studio wants to measure demand, a no-show still represents a booking attempt. If the studio wants to measure room utilization, the seat was empty. These are two different numerators and should not be collapsed into one attendance rate.

A late cancellation that frees a spot which is then rebooked is a third case. The original booking did not attend, but the seat was used. The class may show full attendance while the original member's behavior is a retention signal. Report the rebooking separately if the studio wants to understand both.

Define the cutoff for "late" before running the report. A cancellation two hours before class and a cancellation two minutes before class can have different operational meanings. Record the cutoff and the time zone. If the platform stores times in UTC, convert them to the studio's local time before classifying a late cancel near midnight.

Use the scheduled start, not the upload time

Attendance exports often carry a timestamp for when the record was created or last updated, not when the class started. A check-in recorded at 6:17 p.m. may belong to a 6:15 p.m. class, or to a 6:30 p.m. class if the member arrived early. The class session identifier is the reliable link; the timestamp is supporting evidence.

Decide whether the measure counts attendance at the scheduled start or by the end of the class. A member who arrives ten minutes late still occupied a spot, but they may have missed the warm-up and the coach's planned progression. For capacity planning, end-of-class attendance is often enough. For engagement or safety, the start-time measure may be more meaningful.

Record the rule in the data dictionary. If the studio changes from start-time to end-of-class counting, the percentage will move even if behavior did not. A trend line without a rule-change marker invites the wrong conclusion.

Build a denominator table before the numerator

A denominator table lists every class session in the window with its sellable capacity, capacity version, and any blocked spots. It is the frame for the study. Build it from the schedule, not from the attendance export, because the attendance export only contains classes that someone booked.

Include cancelled classes and classes with zero attendance in the frame, then decide whether to exclude them. A cancelled class has no meaningful attendance percentage, but it may still matter for capacity planning. A class with zero attendance is a real result, not a missing value. Excluding empty classes makes the studio look fuller than it was.

If the schedule changed after the class occurred, preserve the schedule version that applied on the class date. A later edit that merges two sessions or changes the room should not rewrite the denominator retroactively. The same discipline applies to capacity: use the value that was in force when the class ran.

Report the numerator in parts

Publish the total attendance alongside the categories that make it up. A single percentage can hide a shift from paying drop-ins to comps, or from members to staff. The parts explain why the percentage moved.

A compact table works: class session, sellable spots, member attendees, paying drop-ins, comps, staff, no-shows, late cancels, rebooked seats, and the resulting utilization rate. The rate can be calculated as occupied sellable spots divided by sellable spots, with the category breakdown available beside it. Do not force every category into the rate if the decision does not need it.

For a manager review, a short narrative is more useful than a wall of numbers. State the measure, the window, the frame count, the exclusions, and the two or three largest sources of difference from the prior period. A percentage without that context is a conversation starter, not a finding.

Check the data before trusting the pattern

Run a few basic checks before comparing months. Confirm that every class in the denominator has a matching session identifier in the attendance export. Look for duplicate check-ins, check-ins before the class was created, and attendance recorded after the class ended by more than the studio's grace period.

Check the capacity values for zeros and blanks. A zero-capacity class will either drop out of the denominator or produce an undefined rate. A blank capacity may mean the room was not configured, not that the class was empty. Treat both as data quality findings and report them rather than silently filling the gap.

Check the member identifiers for duplicates and shared accounts. A family account can produce two attendees under one billing record, which is fine for attendance but misleading for unique-member counts. State which identifier the unique count uses and whether it is the member or the household.

Keep personal data out of the trend report

The trend report does not need names, contact details, or payment information. Use aggregate counts and, when a case-level example is necessary, a pseudonymous case key. The NIST Privacy Framework is a useful structure for limiting data collection to what the review needs, but the studio's own privacy obligations and retention rules still govern.

If the review requires checking a specific member's attendance, keep that case file separate from the aggregate report and restrict it to the roles that need it. Do not paste attendance exports into a general chat channel or a shared drive with broad access. Delete temporary exports when the review is complete unless the studio's retention policy requires keeping them.

Interpret the result without overreach

A local attendance percentage describes the classes in the frame during the window. It does not describe the fitness industry, the studio's market, or the quality of a coach. A high percentage can reflect a small room, a popular time slot, or a denominator that excludes comps. A low percentage can reflect a new class, a weather event, or a capacity change.

Compare like with like. A 6 a.m. class and a 7 p.m. class may have different demand patterns. A reformer class and a mat class may have different sellable capacity. If the studio wants to compare coaches, use the same class format, time slot, and capacity rule, and report the number of sessions behind each percentage.

Name the largest uncertainty. With a small number of sessions, one cancellation or one staff spot can move the percentage by several points. Say so instead of presenting a precise figure that implies more confidence than the data supports.

Turn the review into one operational change

End the review with a decision. If the data shows that comps are crowding a peak class, the studio may change the comp policy or block a separate spot. If drop-ins are filling a class that members cannot book, the studio may adjust the booking window. If no-shows are high but rebooking is also high, the problem may be notification timing rather than demand.

Assign an owner, a due date, and a follow-up measure. Keep the baseline definition with the action so the next review compares the same thing. A measurement review that produces a chart but no owned change is just another report in the inbox.

The value of this work is not a perfect percentage. It is a shared definition that lets the studio manager, the front desk, and the virtual assistant look at the same class and reach the same conclusion about what happened. That shared definition is what makes the next schedule change, staffing decision, or retention conversation easier to defend.

Limits and sources

This note is a measurement method, not a claim about typical studio performance. The American College of Sports Medicine publishes professional resources on exercise science and health fitness, but it does not provide a universal attendance benchmark for a single studio's classes. The Centers for Disease Control and Prevention publishes physical activity guidance for the population, not class utilization rates for gyms. The Small Business Administration offers general guidance on managing business finances and records, which supports the bookkeeping discipline described here but does not define a fitness metric.

Use the method as a local protocol. Adapt the categories to the studio's booking system, write down the version you used, and keep the evidence with the case. A defensible number is one whose definition can survive a change in staff, a platform migration, or a member's question six months later.