In my first week at one company I was shown the marketing dashboard. More than 40 charts, 6 tabs, numbers refreshed every morning. I asked: what was the last decision you made after opening it? The answer was silence.
The problem: 40 charts, zero decisions
Dashboards usually come from wanting to "see everything". Numbers multiply and attention scatters. A manager opens the dashboard, glances at a few green and red arrows and closes it. There is data, but no action.
A simple test: for every chart ask, if this number changes, who will do what? If there is no answer, the chart should not be there.
The rule: one panel, one question
Every panel should answer one business question. "Should we raise the ad budget?" "Which channel is behind this month?" "Is the new product page working?" Once the question is clear, so is the metric you need.
The decision map
Before building a dashboard I prepare a small table with four columns: question, metric, threshold and action.
- Raise the budget? If ROAS is above [target], increase the budget.
- Where does the funnel leak? If step-to-step conversion drops week on week, start a UX audit.
- Is SEO working? If the organic share of revenue does not grow quarter on quarter, change the content plan.
Once this table is ready, the dashboard itself is 30 minutes of work. The real work is agreeing the table, not drawing charts.
Role-based views
At AzInTelecom we built separate views for the CEO, marketing and product teams. Same data, different questions: the CEO sees revenue against target, marketing sees channels, product sees the funnel.
"A good dashboard tells you what to do, not what happened."
Three steps to start
- Write down the 5 decisions your team makes every week.
- Pick one metric and one threshold for each decision.
- Move every other chart to a separate "research" tab.
Check after a month: how often was the dashboard opened and how many decisions did it lead to? That is the dashboard’s own KPI.