Method
How a study runs
The method is slow on purpose. Session analytics here means a locked sample, two passes, and a memo that can be read aloud.
This page is the spine of a full session study. Path mapping borrows the first two steps and then draws. Practice days use the same two-pass rule with your staff holding the pen.
1
Intake
We ask how consent was obtained, which screens are in bounds, and what you already believe. Beliefs are written down so they do not sneak into the log as facts.
2
Sample lock
A list of session identifiers is frozen. Late arrivals go to a parking lot. Testers we recognise are removed and named in an appendix so nobody pretends the sample was cleaner than it was.
3
First pass
Each visit is watched or read once. The log captures time, screen, visible action. No interpretation. If someone in the room starts a story, Jia Wen stops the clock.
4
Second pass
We return to pauses, repeated fields, language switches, and exits. We still describe before we guess. Guesses, if any, are labelled in the memo as questions.
5
Memo and appendix
The memo is short enough for a meeting. The appendix is long enough to replay. Counts appear only when the sample can honestly carry them.
6
Clarification
A call or a [readout](/services/stakeholder-readout/) stays inside the freeze. New clips are a new conversation, priced as follow-on reading or a new study.
What we will not do in the method
We will not score a session. We will not call a visit “healthy.” We will not hide a small sample behind a round number. If eighteen people looped through an address screen, the memo says eighteen of the locked sample, not a percentage dressed as a law of nature.
Ask to use this method on your app
If the steps above match how you want the work done, request a session review. If you need a map or practice days instead, say which, and we will answer against that brief.