Why did interfaces get faster while waiting got longer?
Loading metrics have improved for a decade. The feeling that the site is thinking has not gone anywhere.
Why the question came up
We moved three projects to static delivery and got honest numbers: time to render halved. In interviews people described the new version with the same words as the old one.
So the thing being measured is not the thing a person experiences as waiting.
What is already known
- Core Web Vitals describe largest element paint, interaction response and layout shifts — three different events, not one "fast".
- Skeletons, spinners and transitions add real time after the data has already arrived.
- A delay before the first visual response to a tap is experienced differently from the same delay mid-load.
What is not known yet
- Where the line is after which animation stops explaining a transition and starts stretching it.
- How to measure perceived waiting outside a lab without asking the person directly.
- How much the feeling of speed depends on layout predictability rather than on time.
Current hypothesis
Hypothesis
Perceived speed is set not by load duration but by the gap between a person acting and the first sign that the interface heard them.
Next step
Measure two intervals separately across twenty interfaces: to first visual response and to screen readiness. Compare with blind ratings of the same screens.
Sources and notes
Links lead to the authors’ own materials. QVORIX publishes original texts, summaries and analysis; third-party materials are not republished in full.