The common advice is simple: the more numbers your dashboard shows, the better. Pack in the charts. Add another row of tiles. Wire up every API you have. If a metric exists, surface it.
That advice is wrong. Or at least, it is incomplete in a way that quietly costs businesses time, focus, and money.
A dashboard is not a display of everything you can measure. It is a decision-making surface. And most dashboards I am asked to fix have the same problem — they are full of numbers nobody acts on.
A dashboard is a question, not a museum
Every chart on the screen should answer a question somebody actually asks. "Are we on track this month?" "Is this campaign working?" "Which customers need a call this week?"
If a tile does not map to a real question — one a real person asks in a real meeting — it is decoration. Take it off. The screen gets calmer, the eye finds the important things faster, and the dashboard starts pulling its weight.
Vanity numbers feel productive and are not
Page views. Total signups. Followers. Total revenue all-time. These numbers go up and to the right by default, which is exactly why they are dangerous. They reward the team for existing, not for improving.
The numbers that matter are the ones that can move in either direction this week — conversion rate, qualified leads, refunds, churn, time to first response. Those are the metrics that change behaviour, because a bad week shows up immediately and somebody has to do something about it.
One screen, one job
The most useful dashboards I have built do one thing well. A sales dashboard tells the founder whether the month is on track and where the gap is. An operations dashboard tells the team what to handle today. A finance dashboard tells the owner whether the business is healthy this quarter.
Trying to do all three on one screen produces a screen that does none of them. If you genuinely need three views, build three views. Do not staple them together and call it comprehensive.
Trends beat snapshots
A single number on its own is almost never useful. "Revenue this month: $42,300." Compared to what? Last month? The same month last year? The target?
Good dashboards almost always show movement — a sparkline, a delta, a comparison to a previous period. The eye is trained to spot direction faster than magnitude. A flat line and a falling line tell two different stories, and the story is usually what the viewer needs.
Show the few things that drive the many
Most businesses have three or four numbers that, if they move, everything else moves with them. For a service studio it might be qualified leads, close rate, average project value, and delivery time. For an ecommerce brand it might be sessions, conversion rate, average order value, and repeat rate.
Find those numbers. Put them at the top. Make them large. Everything else is supporting evidence. The owner should be able to glance at the first row and know how the business is doing in under five seconds. If that is not true, the dashboard is not finished.
A good dashboard makes the next action obvious
The point of looking at a dashboard is to do something different afterwards. If the page is green, the action is "keep going." If something is red, the action should be obvious — which campaign to pause, which customer to call, which queue to clear.
The best dashboards I have shipped include a small "what to do now" panel beside the numbers. Not a chart. A short list. The data has already been interpreted for the person who has to act on it. That is the difference between a dashboard and a report.
When more data genuinely helps
There are real cases where density is the right answer. A trading desk, an operations control room, a fraud team — environments where a trained operator is paid to scan many signals at once and where missing one has a serious cost. In those settings, more numbers, denser layouts, and faster refresh rates are correct.
Almost no small or medium business is in that situation. If your team is not staffed to watch a screen for eight hours a day, you do not need a screen designed for someone who is.
The better rule of thumb
Build the dashboard backwards. Start with the decisions the business actually makes — weekly, monthly, quarterly. Work out the smallest set of numbers needed to make each of those decisions well. Put those numbers on the screen, with comparisons, in the order they get used. Stop there.
A useful dashboard is small, calm, and honest. It tells the owner the truth about the business in a glance, and it makes the next move obvious. Everything else is noise dressed up as insight.
If your current dashboard is loud, crowded, or quietly ignored, that is usually a design problem rather than a data problem — and it is the kind of problem worth a conversation.