Class check-ins: methodology

How the check-in is collected, what each answer means, which answers are counted, when a figure is published at all, and what this data cannot tell you.

Both mood answers are collected after class; the before-class answer is the woman's recollection of how she felt before the session.

Collection point and timing

The check-in is part of the wrap-up a woman completes at the end of a live class, in the member app. It is voluntary: a woman may finish a class and answer nothing, and many do.

The scales

• Mood, before and after (both answered after class): 1–10
• How hard the session felt: 1–10
• How much of the class she rated herself as keeping up with: 0–100

The kept-up figure published on these pages is the MEAN of that self-rating across the check-ins that carried it. It is not a share of women, and it is never printed with a percent sign — a woman rating herself high and a woman rating herself low both move the mean, and no threshold anywhere decides who did or did not keep up.

Which check-ins are counted

A check-in is counted only when it comes from the wrap-up, falls inside the reporting window, carries both mood answers as valid numbers on the scale, and belongs to a live class that is publicly listed, not deleted, not a test session, and not a one-to-one appointment. A check-in that cannot positively prove it belongs is dropped rather than counted.

Valid answers per measure

Effort and the kept-up self-rating are optional questions, so fewer check-ins carry them than carry the two mood answers. Each measure is counted on its own, and published beside its own count of valid answers and its own count of distinct women.

Window (IST calendar days, inclusive): 2026-08-15 to 2026-09-13 — Release wb-20260913-9ce7836f0d5ae85d

The unit is a check-in, not a woman

Each row is one check-in after one class. A woman who attends eight classes in the window and answers each time contributes eight check-ins. The distinct-woman counts beside each measure are how many women those check-ins came from.

Publication floors

Overall figures are published only above both of these:

• 50 valid check-ins
• 30 distinct women

A single format is published only above both of these:

• 30 valid check-ins
• 15 distinct women

Effort and the kept-up self-rating are gated separately, on their own valid-answer and distinct-woman counts, against the same floor as the scope they sit in. A measure below its own floor is withheld while the rest of the scope publishes — its counts are still reported, so how thin it was is visible.

Releases

Every publication carries an identifier derived from the window and from the published numbers themselves. Two pages printing the same identifier are printing the same figures. The window rolls daily, so a page loaded on a later day may carry a later release.

What this does not show

These findings describe what women reported. They do not establish that a class caused the difference: every check-in comes from a woman who chose to attend and chose to answer, and there is no comparison group.

Two further limits are worth saying out loud. The before-class answer is a recollection given minutes after the class ended, so it carries whatever the class itself did to her memory of it. And answering at all is voluntary: a woman who had a poor session is under no obligation to fill the wrap-up in, so these figures describe the women who answered, not everyone who attended.

Back to the current release