A dashboard can show dozens of measures and still leave the team unsure what to do next. More charts do not solve that problem. The report needs an organising question: which decisions should this meeting help the business make?
For a service company, five recurring decisions provide a useful structure. The exact measures depend on the business and the reliability of its data.

1. Where should the next budget go?
Show spend alongside appropriately defined outcomes and their maturity. Separate raw enquiries from qualified opportunities and completed business. Include the relevant cost basis so that media-only numbers are not mistaken for fully loaded acquisition costs.
Flag limited evidence. A campaign with one attractive result does not necessarily deserve the entire budget. The dashboard should help the team judge uncertainty, not conceal it behind a green arrow.
2. Which creative question comes next?
Group assets by meaningful concepts, not only filenames. Explain the audience problem or persuasion hypothesis each concept addresses. Show the available response evidence and note important differences in distribution.
The decision might be to develop a promising angle, improve an unclear execution or investigate a decline. A table that merely ranks advertisements by clicks rarely provides enough context for the next brief.
3. Which part of the website needs attention?
Identify the pages receiving relevant traffic and the steps where the intended task becomes difficult. Combine quantitative signals with actual inspection. If a mobile form is unusable, the team should fix it rather than wait for a larger sample to confirm the obvious failure.
Record recent website changes next to the trend. Without that context, the report may attribute a technical release to a change in customer demand.
4. What does sales need to resolve?
Show response ownership, qualification reasons and the status of recent enquiry groups. Keep personal details out of the general dashboard. The purpose is to identify patterns and responsibilities, not expose customer records unnecessarily.
A repeated wrong-service enquiry may require clearer marketing. An accumulation of unanswered enquiries may require operational action. The dashboard should preserve the distinction.
5. What still needs investigation?
Create a visible section for unexplained differences, missing data and assumptions. Assign an owner and a review date to each meaningful issue. This prevents uncertain interpretations from becoming accepted facts simply because they appear in a polished report.
Finish the meeting with a short decision log: action, reason, owner and expected review point. At the next meeting, begin with those decisions rather than rebuilding the discussion from the latest charts.
Keep the display restrained
Use a consistent date range, readable labels and a clear comparison period. Avoid decorative gauges that imply precision the data does not support. Show definitions close enough that a new reader can understand the numbers.
A successful dashboard reduces the time between observing a problem and assigning the next useful action. Its quality is measured by the decisions it clarifies, not the number of widgets it contains.
EBJ FIELD NOTES / THE WEEKLY REVIEW
| 01 What changed? | 02 Why might it matter? | 03 What next? |
Every dashboard section should support a named decision and show enough context to make it responsibly.
Put it to work
Every dashboard section should support a named decision and show enough context to make it responsibly.
Explore EBJ’s measurement and reporting services, or tell us what your business is trying to improve.