What changed, why it changed, and what to do about it — with noise filtered out before anyone reads it.
Write this week’s marketing report across paid, organic, and pipeline.
Three changes matter this week; the other eleven metrics moved within their normal range and are in the appendix. CPL is up 34% because a competitor entered your two highest-intent ad groups, and the branded campaign hit its daily cap on four of seven days. Both decisions come with a recommendation and an owner.
A marketing report template that lists every channel metric side by side makes the reader do the analysis. This one leads with the two or three things that genuinely changed, explains why, separates real movement from week-to-week variance, and closes with the decisions those changes imply.
A recurring structure for reporting performance: headline change, explanation, supporting detail, and decisions. Metrics appear because they moved or because a stakeholder is accountable for them — not because the dashboard has a tile for them.
Three sections: what changed and why, what we are doing about it, and what needs a decision. Anything a reader would not act on belongs in the appendix or the dashboard.
By comparing movement against normal variance for that metric at that volume. A 12% swing in weekly conversions on 40 conversions is noise, and reporting it as a trend produces reactions that cost more than the swing.
The dashboard answers questions people already have; the report tells them which question to ask this week. A report that is a screenshot of the dashboard has skipped the only part that required judgment.
The agent pulls from Google Ads, Meta Ads, GA4, Search Console, and your CRM, reconciles the definitions that differ between them, and writes the narrative rather than pasting tables. Reports deliver to Slack, email, or Docs on a schedule, and the same run can post a short executive version alongside the detailed one.
Ads platforms, analytics, Search Console, and CRM.
Name the outcome metrics; the agent chooses which diagnostics to include based on movement.
Headline changes with causes, noise already filtered out.
Weekly to Slack, email, or Docs, with an executive summary version.
Leads with change and cause instead of a metric grid
Suppresses movement inside normal variance
Reconciles conflicting definitions across platforms
Ends with decisions, owners, and dates
What changed and why, what you did about it, and what needs a decision. If a metric would not change anyone’s behaviour this week, it belongs in the dashboard instead.
One page of narrative with detail available underneath. Length correlates with how much analysis was skipped, not with how much work was done.
Outcome metrics weekly — pipeline, cost per acquisition, qualified leads — and diagnostics only when they moved enough to explain an outcome. Reporting every diagnostic every week trains readers to skim past the ones that matter.
State the change, give the cause, and say what you are doing about it. Reports that bury declines lose credibility permanently, and the decline is usually visible in someone else’s dashboard anyway.