“This is my First Dashboard. Review it.”
Every first dashboard is a good one if it’s used to learn.
It’s rarely “amazing” or “beautiful” (despite all the “amazing” and “beautiful” comments). But it’s still good. Even if it falls short from a data visualization perspective, it’s valuable for your learning. You’ve started. Don’t stop.
Just make sure your next dashboards aren’t just about repeating and reformatting. Focus on:
➡️ Better understanding of data modeling.
A star schema is a good practice, even for a simple dashboard based on one fact table and a 3-4 dimensions.
I’m also guilty of talking about chart types and colors when asked to review a dashboard. The first technical question should be: “Show me a screenshot of the data model.”
➡️ Don’t use implicit measures. Start using simple DAX.
Do not drag and drop anything from the fact table. Use dimension tables (for slicers and for grouping in a visual) and explicit measures (measures you have created by writing DAX) only.
Write your first simple measures such as Sales Amount := SUM(‘Sales'[Amount])
➡️ The reasons for including each visual.
Why is there a card for measure A (And what does this large number mean? Is it good or bad? How has it changed?) and a detailed long table with a scrollbar for measure B? Maybe a bar chart is a good compromise for both measures?
First dashboards often include tables, pie charts, donut charts, column charts, bar charts, line charts and cards just because you want to try all visuals. Think twice about why “Sales by country” is a pie chart, “Sales by category” is a bar chart, and “Sales by salesperson” is a column chart. There’s nothing wrong with having three bar charts to show the same type of data relationship.
➡️ The reasons each chart was formatted the way it was.
Why does this chart use so many colors, and why do the same colors mean something different on another chart? There should be a purpose for every color, and every chart element. Less is better.
➡️ Forget about colorful backgrounds, images, and other decorations.
Learn how to make meaningful dashboards instead of just “beautiful” and “colorful” ones.
➡️ Test interactivity (slicers, cross-filtering, cross-highlighting).
If, when you selected a year in a slicer, your line charts look like a dot, your pie charts look like a circle with no slices, and your bar charts look like a single bar with no other bars for comparison, then likely you don’t need this slicer at all (or you need different charts).
Verify cross-filtering and cross-highlighting. If it creates confusing results, then disable it. Enable only meaningful cross-filtering and cross-highlighting.
➡️ Edit the titles. Don’t keep the default ones.
Make sure a person who knows nothing about your dashboard can understand what each visual represents (measure, period, unit) just by reading the titles (subtitles).