A dashboard is a one-screen summary of the numbers that matter most for a specific purpose. A good one answers questions at a glance. A poor one shows twenty charts and answers none. Here is how to build a useful first dashboard.
1. Define the audience and purpose
Ask: who will look at this, how often, and what decision will it help them make? A weekly sales dashboard for a store owner is very different from a monthly marketing report for a leadership team.
2. Choose three to six key metrics
Pick metrics that directly support the purpose. For an online store, that might be:
- Revenue
- Number of orders
- Average order value
- Conversion rate
- Top products by revenue
If you are unsure which metrics to use, read important business KPIs every manager should know.
3. Prepare clean data
Dashboards are only as reliable as their source. Use a tidy table: one header row, one record per row, consistent date formats and no merged cells. Google Sheets works well as a source for beginners.
4. Sketch the layout first
On paper, draw a simple grid:
- Top row: scorecards with headline numbers and comparison to the previous period.
- Middle: a trend chart (line) and a comparison chart (bar).
- Bottom: a detailed table for people who want to dig deeper.
People read from top-left, so put the most important number there.
5. Pick the right chart for each question
| Question | Chart |
|---|---|
| How is it changing over time? | Line chart |
| Which is bigger? | Bar chart |
| What share of the whole? | Stacked bar or a simple pie with few slices |
| What is the exact value? | Scorecard or table |
6. Add a few useful filters
A date range control and one or two dimension filters (region, channel) are usually enough. Too many filters make a dashboard harder to use.
7. Test it with a real person
Show it to someone in the intended audience and ask them to answer a question using it. Watch where they hesitate. Simplify accordingly.
Keep it honest
Start axes at zero for bar charts, label time periods clearly, and note any data gaps. Misleading visuals erode trust quickly; see how to explain insights without jargon.
In Week 3 of our course you build a dashboard like this in Looker Studio using a prepared dataset.
