By Michael Nocito, data analyst · Published August 8, 2026
By the end of this page you can turn a table of events into a funnel in one query, work out what fraction of people get from each step to the next, tell the difference between a step people pass and a step people skip, and produce the one summary a funnel is actually for: where everybody stops. It is about twenty-five minutes, and every result below was run.
Here is what to do today, on the funnel you already report. Look for any step converting at or near 100%. That is almost never a step working perfectly; it is a step that can be bypassed, or one that is logged as a side effect of the next one. Find the sessions that reached the later step without the earlier one, and you will have found it in a single query.
The short version: count distinct sessions per step with conditional aggregation, then check whether reaching a step actually required the one before it.
A step that can be bypassed is the idea, so it gets the picture.
