Drop Rules

Drop rules are rules that decide which incoming events are discarded before storage — such as bot traffic, test data or malformed events.

Drop rules are a quality tool but with a sharp edge: any event dropped here does not come back, so an over-broad drop rule can quietly sacrifice real data too. The key distinction is between dropping and filtering: a filter keeps the data and merely sets it aside in the display, so it is reversible; a drop does not store the data at all, so it is irreversible. Hence the healthy rule is to drop only what you are sure will never be needed (an obvious bot, a broken event), and to filter the rest rather than drop it. The most common use of dropping in web analytics is discarding non-human traffic so the stats reflect the real visitor.