Skip to content
Tutorials Center
My learning

Topics

Lesson 26 of 38 ยท Analytics and Attribution

Marketing Dashboards and Reporting

Learn to design marketing dashboards and reports: choosing metrics, comparing periods, visualizing data honestly, and telling a story that leads to decisions.

A report that nobody acts on is wasted effort. Good marketing reporting answers three questions in order: what happened, why did it happen, and what should we do next? This lesson covers how to design dashboards and write reports that people actually read.

Dashboard or report?

  • A dashboard is a live, glanceable view of a few KPIs for ongoing monitoring (daily or weekly).
  • A report is a written or presented analysis for a period (monthly, campaign wrap-up), with interpretation and recommendations.

A dashboard shows the numbers. A report says what they mean. Teams that only send dashboards get stuck in "interesting" instead of "so what?".

Designing a dashboard

  1. Start with the audience. An executive needs revenue, CAC and payback. A channel owner needs spend, CPA and creative performance. Build different views.
  2. Limit it to about five to eight headline numbers, each with a comparison (previous period or target) and a trend.
  3. Put the most important number top left; people read in that order.
  4. Use consistent definitions and label units and date ranges clearly.
  5. Make problems visible: use a simple traffic-light or target line, so a missed target stands out.

Common tools include Looker Studio, Power BI, Tableau, Metabase, Grafana and the built-in dashboards of analytics and ad platforms. A spreadsheet connected to your data is a fine start.

Choosing visuals honestly

To show Use
Trend over time Line chart
Compare categories Bar chart, sorted
Share of a whole (few parts) Stacked bar or simple donut
Relationship of two numbers Scatter plot
A single number with context Big number with change and sparkline
Funnel Horizontal bars by step

Start axes at zero for bar charts, keep scales consistent between charts you compare, and avoid 3D effects. A truncated axis can turn a 2 percent change into a dramatic cliff.

Compare to something

A number alone has no meaning. Always include a comparison:

  • Previous period (week over week, month over month).
  • Same period last year, which cancels seasonality.
  • Target or forecast.
  • Benchmark from your own history.
JS Period-over-period summary

Press Run (or Ctrl + Enter) to see the console output here.

No output.

The result already tells a story: traffic is up, leads are down and cost per lead is higher, so something in the funnel got worse. That is where the investigation starts.

Writing the report: a simple structure

  1. Headline summary: three or four sentences that state the result and the decision needed.
  2. Key metrics versus targets (the dashboard).
  3. What drove the change: by channel, campaign, device, audience.
  4. What we learned: tests run, what worked, what failed.
  5. Next steps: specific actions with owners and dates.

A report that gets read

Summary first Result and decision in a few lines
Few key numbers With targets and comparisons
Explain why Slice by channel and segment
Next steps Owner and date for each action
1 / 2

Spotting anomalies

An unexpected jump or drop is often a tracking problem before it is a marketing result. A simple check compares each day with a recent average and flags big deviations.

JS Flag unusual days

Press Run (or Ctrl + Enter) to see the console output here.

No output.

A drop to 120 may be a broken tag; a spike to 980 may be bot traffic or a viral post. Investigate before celebrating or panicking.

Automate the routine

Schedule a weekly email of the dashboard, set alerts for large deviations (spend spikes, zero conversions, site down) and keep a changelog of campaign and site changes so that you can explain jumps. See marketing automation for more on automating routine work.

Common mistakes

  • Dashboards with dozens of tiles and no priorities.
  • Reporting activity (posts published, emails sent) instead of results.
  • No comparison, so nobody knows if a number is good.
  • Cherry-picking a time window that flatters the result.
  • Dropping the data-quality caveats: say which numbers are estimates.

For more on defining numbers, see analytics fundamentals.

Check your understanding

Quick check ยท 4 questions

of 4 answered You scored / 4 ยท perfect!

  1. 1. What is the main difference between a dashboard and a report?

    Dashboards monitor; reports explain and recommend action.
  2. 2. Which chart best shows a trend over time?

    Line charts make change across time easy to see.
  3. 3. Why include a comparison with every metric?

    Targets, previous periods and benchmarks give numbers meaning.
  4. 4. A sudden 70% traffic drop for one day is most likely first to be checked for:

    Sudden changes are often broken tags, outages or bots before real behavior change.

Summary

  • Dashboards monitor a handful of KPIs; reports explain and recommend.
  • Design for the audience and always show a comparison or target.
  • Use honest, simple charts: lines for trends, sorted bars for categories.
  • Structure reports as summary, metrics, drivers, learnings, next steps.
  • Investigate anomalies for tracking problems first.

Next: Conversion Rate Optimization