AI Campaign Performance Review Workflow

5 steps to a reviewable draft

1. 01

   Freeze the review window

   Record the campaign dates, reporting timezone, metric definitions and data source. Do not compare periods with different definitions as if they were identical.

   Result: A review scope that others can reproduce from the same export.

2. 02

   List meaningful changes

   Note changes to audience, offer, creative, targeting, budget, delivery or tracking. Mark unknown changes instead of filling the gap with a theory.

   Result: A dated change log beside the metric table.

3. 03

   Draft the performance review

   Ask for a narrative that stays close to the supplied data and labels possible explanations as questions.

   Result: A review that distinguishes the numbers from the story people want to tell about them.

   Your prompt

   Copy prompt

   Draft a campaign performance review from the supplied records.
   
   Objective: {{objective}}
   Data period and metric definitions: {{data_context}}
   Metric export: {{metrics}}
   Dated campaign changes: {{changes}}
   
   Use sections for objective, observed results, comparison basis, changes during the period, data limits, possible explanations, questions to test and next decision. Cite only supplied figures. Label an explanation as [TO TEST]. Do not claim causality, lift, incrementality, attribution certainty or business impact without supporting evidence.

   Example input

   Review a two-week workshop campaign with email sends, landing-page visits and registrations; creative changed on day eight.

   Example output

   Observed result: registrations rose after day eight. Explanation: [TO TEST]. The period also included a new email subject line.

4. 04

   Verify the figures and language

   Reconcile every figure with the export and challenge causal wording. Ask the data owner about missing tracking or delayed conversions.

   Result: A narrative that keeps factual reporting separate from interpretation.

5. 05

   Choose one next test or decision

   Convert the most useful uncertainty into a test card or an explicit decision to stop, continue or collect more data.

   Result: A performance review that leads to an accountable next move.