Every tester keeps a short log each morning: how they slept, anything that woke them, any pain, and a one-to-ten comfort figure. Across the current round that is 271 entries, and the useful thing about them is the shape of the curve rather than the average.
Week one tells you almost nothing
Comfort scores in the first seven nights correlate weakly with the final score. Two of our nine mattresses were rated below 7 on night one and finished above 8.8; one was rated 9 on night two and finished at 8.1.
The first week is mostly novelty and adaptation. Anyone moving from a sagging mattress to a supportive one is going to be uncomfortable while their body relearns a position it has not held in years.
What appears in weeks two to four
Five failure modes showed up consistently in the second half of every test and never in the first.
Perimeter softening. The edge of a mattress does not collapse on day one. It gives way progressively as the foam in the perimeter fatigues, and on two of our nine mattresses the ten-minute seated compression reading was measurably worse at week four than at week one.
Heat that only matters when the weather changes. A mattress tested only in a cool week will pass a cooling test it should fail. This is why our room is held at 68°F year-round, and why we still note the season.
Motion transfer you learn to notice. Our light-sleeping tester was woken twice in week one on a mattress and six times in week four on the same one. Nothing changed about the mattress. She had stopped sleeping through the novelty of a new bed and started registering the movement.
Position drift. Several testers reported unconsciously changing their sleeping position on firmer mattresses — moving from side to back over the first fortnight to escape shoulder pressure. That is the mattress failing at what it was bought for, and a three-night trial would have recorded it as a success.
The pillow mismatch. Changing a mattress changes how far your shoulder sinks, which changes the pillow loft you need. Three of our testers reported new neck pain in week two that resolved when they changed pillows. It is easy to blame the mattress.
What does not change after week one
Two things are apparent immediately and do not improve:
Shoulder pressure on your side. If you are waking with a numb arm on night four, the surface is too firm for your shoulder. Thirty more nights will not change it. A topper will.
A surface you actively dislike. Latex bounce and memory foam sink are polarising and the reaction is stable. Nobody in our group has ever grown to like a feel they disliked on night one.
The practical version
Judge a mattress at week four. Not week one, when you are adapting, and not week twelve, when you have adapted to something wrong.
Keep a two-line log — comfort out of ten, and anything that woke you. It takes fifteen seconds and it is the difference between "I think it's fine?" and knowing whether to return it before the window closes. Our trial and warranty guide covers when those windows actually close.