Skip to main content
Use this pattern for a weekly review, a stakeholder check-in, or the first task after connecting an account. The goal is a scoped analysis, not an automatic account change.

Before you start

  • Connect the source that contains the metrics you want to analyze.
  • Choose a date range and comparison period.
  • Decide which outcome matters: revenue, conversions, efficiency, spend, or another named metric.
  • Confirm the workspace currency and reporting timezone if they affect the comparison.

Starting request

What useful output looks like

A reviewable result should make the scope visible and distinguish four things:
  1. The metrics that changed.
  2. The entities that contributed most to the change.
  3. Explanations supported by the available data.
  4. Questions that need more context or another source.

Verify the result

  • Match the date window and comparison to your request.
  • Check that each metric comes from the expected connected account.
  • Look for attribution, currency, and freshness notes.
  • Open the cited campaigns, creatives, or products before accepting a recommendation.
  • Treat a correlation as a hypothesis unless the evidence supports a causal explanation.
You are done when you can state what changed, where it changed, and what you will investigate next without relying on an unsupported conclusion.
For a deeper investigation, continue with Diagnose a performance change. To understand analytics context, see Analytics.
Last modified on August 7, 2026