Data Visualisation Tips

The other day I was at a meeting where data was presented as 8 bar charts on the same slide. Despite all the hard work that was done to extract the data and mine insights, I couldn't understand a thing. Too many numbers, too many comparisons. The idea was to show totals, then break it down into products, then show before and after. It is all logical, but just doesn't work on one slide.

This situation made me think that all the analysis we do has little to no impact if we can't communicate our findings in an easily digestible way.

Inspired by true story, here are several data visualisation tips from me:

Embrace Negative Space 

Don't fear white space in your visualisations. Proper spacing between elements reduces cognitive load and helps people focus on the most important information. Overloaded visuals are harder to understand and please, no 8 bar charts!

Highlight Cells in Tables 

If you use tables, understand that people naturally read them and check every cell (if you are presenting, you might notice that people go quiet, as they are digesting the visual). Therefore, if you want them to concentrate on a specific number only, use a heatmap or highlight that specific cell.

Also, if you are displaying percentages in the table, double-check that they add up to 100%. Because of rounding, a column often adds up to 101% or 99%, and there is always that one person (!) in the room who will make sure to point it out.

Follow Natural Reading Patterns 

In Western cultures, people typically scan in an F-pattern (top-left to right, then down) or Z-pattern. Place your most important data points where the eye naturally begins. I work in the UAE, where half of the audience scans data from right to left and half from left to right, so I am testing what works best.

Limit Colour Usage 

Use colour strategically and sparingly. Aim for 3-5 colours maximum in a single visualisation. Colours should serve a purpose—highlighting trends, distinguishing categories, or drawing attention to outliers.

And another untold rule—I just found out that I should never use blue and green colours in presentations at my new work. Just the rule in the company. Several years ago, I worked for a guy who couldn't stand purple. No matter how good the work was, if it had purple, it would make him angry and he would ask you to redo everything. This is something you would never be able to predict.

Choose the Right Chart Type

Match your visualisation type to your data story. Line charts show trends over time, bar charts compare quantities, scatter plots reveal relationships, and pie charts work only for showing proportional parts of a whole (and only with few segments). And even then, most people prefer stacked bar charts to pie charts.

Simplify for Clarity 

Remove all non-essential elements (gridlines, borders, decorative elements) unless they directly contribute to understanding. If you are using stacked bar charts, you don't need to add % to every single bar if it doesn't bring any value.

In the following example, we don’t have % for ‘Don’t know’ and ‘Right track‘, as we want the audience to pay attention to ‘Wrong track‘.

Also, it is a good idea to adjust the starting point of the chart to make change more prominent.

Include Context 

Always provide sufficient context through proper labelling, legends, and annotations.

There are several resources I like to use to get some inspiration for my visuals:

Our World in Data

Resource that taught me I can use nice skinny lines instead of usual bar charts, and arrows to add legend that doesn’t fit.

Tableau Public

Users share their reports at Tableau Public, and some of them are true masterpieces. I sometimes check out those dashboards to get some ideas.

Envato

Not sponsored. This is a website where you can buy various digital assets, like stock photos, videos and PowerPoint presentations. I buy presentation templates, but I also sometimes just look through presentation previews just to see how else I can present my data.

Not many of us can become Leonardo da Vinci of Tableau reports, however there is always an opportunity to grow and improve.

Keep pushing 💪

Karina

Need more help?

Just starting with Python? Wondering if programming is for you?

Master key data analysis tasks like cleaning, filtering, pivot and grouping data using Pandas, and learn how to present your insights visually with Matplotlib with ‘Data Analysis with Python’ masterclass.

Get CV, Portfolio and LinkedIn Review.

More from me: YouTube | TikTok | Instagram | Threads | LinkedIn