A marketing dashboard with too many metrics goes unwatched. A dashboard should be judged not by how much it shows but by whether the viewer can decide their next action from it. Try building separate dashboards for each viewer’s role.
Executives: one page, five metrics or fewer
What executives want to know is “Is it working?” and “Can we spend more?” Condense to revenue contribution, new customers, customer acquisition cost, and the conversion flow, and show month-over-month and year-over-year change alongside.
What you must settle before building a dashboard is metric definitions. If people disagree on what “lead” refers to, or at what point a “conversion” is counted, they end up telling different stories over the same number. Make a one-page metric definition sheet and keep it next to the dashboard.
Team leads: look at channels and the funnel
What a team lead needs is which channels are efficient and which funnel stage is leaking.
A dashboard that shows only numbers gets interpreted differently by different people. Put change from last month, a target line, and a one-line explanation beside each metric, and even a first-time viewer understands the situation right away. Let the dashboard explain the metrics that need explaining.
- Traffic, conversion rate, and acquisition cost by channel
- Conversion rate by stage and change from last month
- Progress against goal by campaign
A dashboard’s value isn’t the number of widgets filled in, but the number of behaviors changed after looking at it.
The Story Light editorial team
Practitioners: numbers that move daily
For practitioners, what matters is metrics they can check daily and act on right away. A layout that shows “what to fix today,” such as send performance, response by ad creative, and error or anomaly alerts, works well.
A dashboard starts going stale the moment it’s finished. Once a quarter, check “Did any decision come from this metric?”, remove the widgets that went unused, and add metrics that fit the new questions. The dashboard that stays light is the one that survives.
What to do after you build the dashboard
A dashboard isn’t a deliverable you finish once. Metrics you thought were necessary at first often go unchecked by anyone after a few months. About once a quarter, go through every widget and ask whether looking at this number changed anyone’s behavior, and boldly take down the ones with no answer.
Also note on the screen where the data comes from and how often it refreshes. When the refresh time isn’t clear, trust wobbles every time a number looks off. Recording in the metric definition sheet which source to follow when ad platform and analytics tool figures differ also cuts needless arguments.
- Note an owner and refresh cycle for each widget
- Clean up unused metrics every quarter
- Decide the reference source to follow when numbers differ
- Set up a way inside the dashboard to collect user feedback
Meeting habits that turn numbers into stories
Even a good dashboard is just decoration if it isn’t used in meetings. Make the first five minutes of your regular meeting a time to look at the dashboard together, and name one number that changed from last week and why. When everyone talks from the same screen, you also save the work of building separate reports.
You also need the discipline not to jump to conclusions when a strange number appears. First check whether the way things are counted changed and whether a one-off event happened, then discuss causes. When this check becomes a team habit, trust in the numbers and use of the dashboard rise together.
When first designing a dashboard, meet one or two of the people who will use it and ask what question they open this screen to answer. Rather than listing requested metrics as they are, keep only the numbers truly needed to answer that question, and the screen becomes much tidier. Even for a metric someone says is needed, asking what decision they make with that number reveals what’s really necessary.
Check the finished screen by showing it to a colleague seeing it for the first time, without explanation, and asking what they can tell from it. If they don’t get it at once, that signals the layout, not the metrics, needs work. Small tests like this produce dashboards that last.
Operating a dashboard takes as much time as building one. Assign an owner and put a schedule of regular reviews and revisions into the work. A dashboard nobody maintains is easily abandoned quietly after a few months.
Summary
Boldly remove metrics that haven’t been opened once in the past three months. If deleting a metric feels hard, first ask yourself whether any decision was ever made from it.
Add your first comment to this post