Skip to content

Data Window

The Data Window is the number of days of history a rule analyses. It is the single setting that most often explains why a rule reached a different conclusion than you expected.

Value
Default Data Window 90 days
Hard maximum 180 days, regardless of plan
Your account ceiling Set by your Automation Window, shown in the App
Default recency exclusion The most recent 2 days are ignored

The hard maximum is 180 days. No plan and no add-on raises it above that.

What a 90-day Data Window actually reads A timeline running left to right from 92 days ago to today. A wide shaded band covers the 90 days the rule analyses. A narrow band at the right marks the two most recent days, which are skipped. A dashed line marks the anchor where the analysed window ends. 92 days ago Today Anchor Data Window · 90 days 2d Everything the rule evaluates its thresholds against. Hard maximum 180 days, whatever the plan allows. Skipped by default: Amazon attribution is still arriving, so these days understate orders and sales.

A 90-day Data Window with the default two-day recency exclusion. The anchor is where the analysed window ends — everything to the right of it is what the rule reads.

Amazon advertising data is not final when it first appears. Attribution keeps arriving after the click, so the last day or two of any report understates orders and sales. A rule that reads those days at face value will conclude that recent spend produced nothing.

By default a rule ignores the most recent 2 days for that reason. You can change how many days are skipped.

Two names for the same underlying limit, used in different places:

  • Data Window is the field in the rule editor. It is what a specific rule looks at.
  • Automation Window is what your account is entitled to. It is the ceiling every Data Window on the account has to fit inside.

Your account’s Automation Window ceiling is shown in the App.

A longer window is not automatically better. It is a trade between confidence and responsiveness:

  • Short window — reacts quickly, but a seasonal dip or a slow week can look like failure.
  • Long window — settles a question properly, but takes longer to notice that something has genuinely changed.

If you want both, that is what Momentum Guard does: it holds a long baseline and watches shorter checkpoints against it.

See also Evidence and periods.