How to Make “Cool” and “Fancy” BI Report
Your BI report isn’t “cool”? It lacks “cool” backgrounds, effects (3D, shades, gradients), and “cool” images and icons, etc.? Do you want to make your BI report look great? Here are my tips:
🛑 Do not start learning how to use Figma, Adobe Photoshop, Adobe Illustrator, or any other magic tool to create “cool” report backgrounds, images, and icons.
They are useful tools, but that’s not where the real amazingness of your BI report begins.
🛑 Do not spend your time adding decorations to your report.
🛑 Do not use the words “amazing”, “beautiful”, “design” regarding BI reports. They are often misused.
The word “design” isn’t bad (think process design, system design), but in BI it’s often overused in the wrong context and mistaken for graphic design and overall visual appearance.
🛑 One more misused term – “wireframing”. Do not design the structure of your report to make it look “cool”. Do not insert the data into the “design”.
Report design is not a visual placeholder for the data. Report design is about visualizing data and helping report users to understand the data.
🛑 Do not fall into the pitfall of thinking that data visualization equals graphic design.
🟢 Learn how to make better charts (and tables).
Better = works better, not “looks better” (although a chart that works better often looks better because of its simplicity and clarity, not because of decorations).
🟢 Make sure the report layout is based on the data visualization principles. The layout should show relationships between charts, provide faster access to what is more important, and properly use white space (shadows and gradients are not white space; they are decorations.)
🟢 Verify every element of your report for its usefulness from a data point of view:
➡️ Does it help users make better data-driven business decisions?
➡️ Does it help reveal trends, patterns, correlations in the data more clearly?
➡️ Does it help to determine if data is correct and up-to-date?
➡️ Do report slicers and filters (including cross-filtering) work correctly (rather than just look “cool”)?
You can undermine the “amazingness” of many “amazing” reports by selecting a value in a slicer. This action often result in line charts with a single dot instead of a line, bar charts with a single bar, and pie charts with no slices, among other issues.
➡️ Is it clear what data you see on each report chart? What is the source of these data? What units does it use? What does that abbreviation mean, and so on?
➡️ Is there anything potentially misleading (missing values displayed as zero, bars not starting at 0, and so on)?
➡️ Does the report design help to see the data (same things mean the same, different things mean different, colors help to see important things, and so on)?
Once you have a report that works better, ask yourself a question: Isn’t it an amazingly simple and efficient report? Is a color gradient background going to make the report better?
Collins dictionary. Synonyms of ’fanciness’: elaborateness, decorativeness.
🎯 Make your “fancy” reports fancy due to their elaborateness, not their decorativeness.
🎯 Make your “cool” reports cool because it is cool to understand the data by using these reports.
Books to read:
