Real-time dashboards: what to measure so people actually use them
A dashboard only gets used when it answers a specific decision someone makes regularly. Abandoned dashboards almost always share the same flaw: they show everything the tool offers instead of the few numbers that change what the team will do this week. Fewer indicators, with an owner and a target, beat thirty charts.
30-second summary
- A dashboard exists to decide, not to prove data exists.
- Start from the question someone needs to answer every week.
- An indicator without an owner and a target becomes wallpaper.
- Real time only matters where decisions are real time. Elsewhere it gets in the way.
- A dashboard nobody opens should move to chat or email, not keep waiting for visitors.
Why most dashboards die
Because they were built backwards: someone looked at everything the tool exported and put it on screen.
The result is a dashboard with thirty numbers where none is clearly the most important. Whoever opens it does not know where to look, and within two weeks stops opening it.
Start from the decision, not the data
The right question is not "what can we measure", it is "what decision does someone need to make every week?".
Real examples:
- Do I increase or reduce this campaign's budget?
- Which location needs attention this week?
- Are we on pace for the monthly target?
- Did any channel stop delivering?
Each of those defines two or three numbers. Those are the dashboard. Everything else is an appendix.
How many indicators should a dashboard have?
Few. If you need to scroll to see everything, there are already too many.
A structure that works:
- Three headline numbers, large, compared against the previous period
- One alert block, showing only what is off track
- One detail view for whoever wants to investigate
The detail exists to answer "why" after a number caught attention, not to compete with it.
Every indicator needs an owner and a target
A number without an owner produces no action. A number without a target produces no judgement: is 4.2% good or bad?
Before adding an indicator, answer two things: who acts when it worsens, and at what value it is considered bad. If there is no answer, it should not be there.
Does real time actually matter?
Almost never. Real time serves real-time decisions, like a campaign on a peak day or a system in operation.
For marketing and sales management, daily data is usually enough and more reliable, because partial numbers swing and lead people to react to noise. A dashboard refreshing every minute invites changes to campaigns before they have enough data to judge.
The dashboard nobody opens
There is an uncomfortable truth: even a good dashboard gets few visits.
The fix is not a better dashboard, it is inverting the flow: send the summary to where people already are. A short message with three numbers and what changed, in chat or email, gets read by far more people than any link.
The dashboard still exists for whoever wants depth, and the summary guarantees the information arrives. We showed that format in report automation: from raw data to chat.
The sign that it is working
It is not the number of visits. It is someone saying, in a meeting: "I saw on the dashboard that this dropped, so we changed X".
Until that happens, the dashboard is still a pretty report. On reading numbers without fooling yourself, see how to read media reports.
Building this, connecting data sources and delivering the summary where the team already is, is the work of area next.
Frequently asked questions
Which tool should I use to build a dashboard?
The tool is the least important part. A spreadsheet connected to your sources solves most cases at low cost. Dedicated tools make sense when there are many sources, many users or a need for access control.
How often should the data refresh?
Daily covers almost every marketing and sales case. Refreshing every minute tends to trigger reactions to normal fluctuation, which hurts more than it helps, especially for campaigns still in learning.
How many indicators should a dashboard have?
Few, and ranked. Three headline numbers, an alert block and an optional detail view cover most needs. If every number has the same visual weight, none of them does.
How do I get the team to actually use the dashboard?
By taking the summary to where the team already is instead of waiting for them to visit a link. A short recurring message with the main numbers and what changed gets far higher readership, and the dashboard stays for whoever wants depth.
An agency gives you a generic team.
A hub gives you a specialist per front.
Four domains, one direction, united by method. The difference between executing and solving.