A daily word ladder takes about two minutes, which means most players never choose a time limit; the game chooses one for them. Choosing it deliberately is a training tool.
There are two useful settings. The first is a hard stop: two minutes, then you take the hint or lose the ladder. That setting trains decisiveness, because you learn to commit to a route rather than holding three plans open. The second is no limit: play until it is solved, however long it takes. That setting trains depth, because you can afford to explore the wrong corridor properly and learn why it was wrong.
Most players accidentally use a fuzzy version of the hard stop. They do not decide when to stop; they stop when they get bored or when the solve stops being easy. That is the worst of both settings, because it neither trains decisiveness nor depth.
The rule I use is to pick the setting before I look at the puzzle. If I have three minutes and a fresh mind, no limit. If I have a busy morning, hard stop and hints allowed. The decision is made in one second and it removes the ambiguity that makes the morning play feel worse than it should.
Pick the setting on the puzzle page before you see the board, and use practice mode for the deep sessions where you actually want to explore.
A last note on the fuzzy stop. The reason it feels productive is that it always ends in a solve, either by luck or by hint. But a solve produced by an undeclared stop is not evidence of anything, and players who run on fuzzy stops tend to plateau while believing they are improving. Declaring the setting takes one second and restores the signal.