Scaling in splits the intended size across several entries: a third at the level, a third on confirmation, a third on the retest. The full risk budget is decided first, then divided.
The advantage is that it reduces timing error. Few entries are precise, and three fills around a zone usually beat one guess. The discipline is that the tranches must be planned; adding "because it went against me" without a pre-set plan is averaging-down, which is a different and more dangerous thing.
Worked example: $600 of dollar-risk, stop $2 below the zone. Instead of 300 shares at once, buy 100 at $52.00, 100 at $51.40 and 100 at $50.80, all stopping at $49.50. The blended entry is $51.40 and total risk is roughly $570 - close to plan, provided the third tranche is skipped if the stop is hit first.
Related: scaling-out, pyramiding, averaging-down, dollar-risk