Measure what mattered, not what was easy to count

Registration numbers flatter the organiser and tell you almost nothing useful.

Registrations are the easiest number to gather and the least informative. They measure the appeal of a title and the low cost of clicking yes. Attendance is better but still shallow, because presence and attention are not the same thing online. The numbers that actually describe how a session went are the ones about participation and consequence: how many people answered when asked, how the audience held across the middle, what the questions revealed about comprehension, what people did afterwards.

Qualitative signals deserve equal weight and are usually ignored because they resist a spreadsheet. What did chat argue about. Which segment produced silence. Where did the questions cluster, and were they introductory questions that suggest the session was pitched wrongly, or applied questions that suggest it landed. A handful of specific comments read carefully will improve the next session more than a chart of attendance ever will.

Whatever is measured should be chosen before the session and tied to its stated purpose, so that the numbers answer a question rather than decorate a report. And it should be small: two or three signals reviewed honestly and acted on will do more than a comprehensive dashboard that nobody reads. The point of measurement here is to change the next design, and any measure that cannot change a decision is not worth collecting.