Decision Center

Reading the reasoning trail

Every recommendation shows its inputs, its comparison window, and what would change its mind.

ENWritten by · Documentation LeadLast updated

Writes and maintains the Paid Lens documentation, from first connection to workspace administration.

The reasoning trail is the audit log of a recommendation. It answers three questions: what did the platform observe, what did it compare that against, and what would make the recommendation invalid.

Observed

The metrics that triggered the candidate, each linked back to the view it came from. Click any metric to see the same number in context.

Compared against

The baseline window, usually the trailing 28 days excluding the last 3 days to allow for conversion lag. Seasonal accounts can switch this to a year-over-year baseline in settings.

Invalidating conditions

The explicit list of things that would retract the recommendation, such as a tracking outage or a pricing change during the window. If you know one of these applies, dismiss with that reason.

Run these steps on your own accounts — start a free 14-day trial

Start 14-day free trial

No credit card required.

Related articles