How-to: Power BI Analytics Dashboard
Simon Nuss and I recently hosted two Power BI workshops online: a beginner level workshop to create an interactive resume (see mine), and an intermediate level demo to create a COVID-19 dashboard using public data. Simon also shares his expert Power BI development tips below the videos. Happy Dashboarding!
Power BI Development Tips by Simon Nuss, Mar 24 2020
Spend more effort creating “good data”. It results in less expended time over the entire project.
What is “good data”? This comes with experience but is largely defined in the literature as a “star schema”.
When developing the initial mode, aim to spend 40% of your time in Power Query, 40% in DAX, and 20% designing reports.
Avoid the trap of adding physical columns; instead, add measures. Rule of thumb: physical columns for filters or axis’; measures for everything else.
Do not make wide frankentables. Keep data in separate tables. However, avoid daisy-chaining (snowflaking) tables. Respect the “star schema” design.
“Good design is as little design as possible” - Dieter Rams.
Enable “Snap to Grid” for easy visual alignment
Begin with highly detailed operational reports that have a daily or weekly cadence, e.g., start with three COVID-19 pages: Cases, Deaths, & Recoveries. Lastly, develop the executive summary on a weekly or monthly cadence.
Maximum of 4 visuals per page (excluding cards and slicers).
Whitespace is your friend.
Avoid slicers in favor of horizontal bar charts.