The alarm goes off at 7:30. By 9:15, Daryna is already three sessions deep into Hotjar recordings, watching anonymous users abandon a checkout form at the same field for the fourth time this week. This is not glamorous work. This is pattern recognition with a cup of tea going lukewarm beside the keyboard.
What the Numbers Look Like at Breakfast
A typical UX audit session for a student researcher starts with quantitative data: bounce rates, average session duration, scroll depth. According to the Nielsen Norman Group, users spend an average of 10-20 seconds on a webpage before deciding whether to stay or leave. That window is brutally short. Daryna marks the timestamp where drop-off spikes and cross-references it with the page element visible at that scroll position.
The number that keeps appearing this morning is 68%. That is the percentage of users who abandon the registration form after encountering the date-of-birth field. Not because the field is broken. Because it uses a three-dropdown format instead of a simple text input. One structural decision, three months of data, one obvious answer.
Afternoon: From Spreadsheet to Sketch
After lunch, the data shifts from passive observation to active hypothesis. Daryna maps the session recordings against the click-map overlay and notices something the aggregate stats missed: mobile users tap the submit button twice before the page responds. The server response time is 2.3 seconds. Google research from 2023 puts the acceptable mobile load threshold at under 1 second before abandonment probability rises sharply.
Data does not tell you what users feel. It tells you where they stopped.
That distinction matters. By 4 PM, the hypothesis is written, the wireframe is adjusted, and the A/B test parameters are set. Tomorrow the data collection starts again. Same ritual, different variables.