Day 04 · Dashboard layout
The full checklist from my post, with a before and after for every rule. Make it clean. Make it easy to scan. Works the same in Power BI, Excel, Tableau, Looker Studio, and Google Sheets.
Start here. Most dashboards only need these six blocks, read in this order.
Decide where things go before you build a single chart. Moving boxes on paper takes seconds. Moving charts in Power BI takes an afternoon.
Can you explain every box in one sentence? If a box has no question, it doesn't get a chart.
For a monthly sales meeting. The hero answers “are we on target?”
For a content creator or brand. The hero answers “is my reach growing?”
For a service or support team. The hero answers “are we fast enough?”
Every dashboard has one main question. The chart that answers it is the hero: the biggest one, right under the KPIs.
A line chart over time.
Bars with a target mark, or a line against a target line.
A sorted bar chart, biggest on top.
A 100% stacked bar. Not a pie with seven slices.
A table with the cells that need attention highlighted.
Things that sit close together look like they belong together. Use that: proximity creates meaning.
Squint at your dashboard until it's blurry. You should see three or four clear groups, not a dozen floating boxes.
Uneven gaps make a dashboard feel messy even when the charts are good. Pick a few spacing values and repeat them.
Hold a ruler (or any straight edge) against the screen. Left edges should line up, and the gaps should all look the same.
Too many boxes create noise. Every extra box makes the others smaller and harder to read.
Name each section in three words or fewer. If you can't, it's doing too much.
People read a dashboard like a page: top left first, then across and down. Put things in the order you want them read.
Don't shrink the desktop layout. Put the KPIs in a two by two grid, then the main chart, then the insight, then the details. Power BI has a separate mobile layout for this: View › Mobile layout.
Show your dashboard to someone for five seconds, then ask what they looked at first. It should be your headline.
Dashboards should explain the takeaway, not just display the data. Leave a box for it in your wireframe from day one.
• Sales are up 8%, but margin is down to 28.6%.
• Check the discount-heavy products first.
What changed+why+what to do next
One to three bullets, each under 20 words. Use numbers that are on the dashboard, and end with an action. Insight = action.
A sample dashboard that follows all seven rules. The small numbers show where each rule is at work.
Sample Store Sample data
Revenue in millions. Last 12 months.
KPIs 3
Above target for the last three months.
RevenueTarget
September, share of new users.
| Product | Revenue | Discount | Margin |
|---|---|---|---|
| Cold Brew Bundle | 1.9M | 25% | 12% ▼ |
| Espresso Beans 1kg | 1.6M | 20% | 18% ▼ |
| Hand Grinder | 1.1M | 15% | 21% |
| Drip Kettle | 0.9M | 10% | 27% |
▼ Margin below 20%.
All numbers are made up for practice.
Clean structure makes dashboards easier to read and easier to trust. Tick these off on your next dashboard; your ticks are saved in this browser.
0 of 7 done
Three prompts that use these rules. Fill in your details once and they drop into every prompt. Works with ChatGPT, Claude, or Gemini.
Saved only in this browser.
AI can suggest a layout, but it can't see how your team uses the dashboard. Keep what fits, and check every number it writes into an insight against your data.