Suchen...
Generic filters
Exact matches only
Search in title
Search in excerpt
Search in content
Bissantz Bixel

The twelve most common mistakes in data visualization


Graphics and diagrams in business intelligence

Data visualization (DataViz) within companies is highly demanding when it comes to clarity and comprehensibility. When data is visualized, graphical code is created. Readers and recipients of data visualizations are typically in a hurry. Their attention span is limited. This means it is important for them to be able to decode the visualization flawlessly. For companies, incorrect conclusions and misinterpretations go hand in hand with costs and risks.

Data visualization for business intelligence is not aimed at the teasing comments or outright manipulative tricks of opinion-formers in the media, nor does it seek to entertain executives with journalistic data storytelling. Its primary aim is to provide orientation as to where targets are being achieved or missed, to enable cause analysis, and to send out signals. All this requires is a small number of consistently used visual tools and the right colors.

THE TWELVE MOST COMMON MISTAKES IN VISUALIZATION

A lot of mistakes are made in data visualization. For companies, errors in business intelligence go hand in hand with costs and risks. Here is a brief presentation of twelve typical mistakes. You can find out how to do things right in our white paper – so that you can avoid the pitfalls as a reader and an author alike.

1. Samba for the eyes – Stacked columns

We often encounter bar charts as stacked columns. When it comes to the clear and simple communication of business-critical values, these are as unsuitable as a spaghetti diagram.

2. The discrepancy problem – Linear formats for different scales

Linear formats require linear scaling. But that alone is not enough to present KPIs such as revenue and profit with the degree of reliability that is elementary to business intelligence.

3. Decograms – Decoration instead of information

Graphic designers with no understanding of data should be out of a job in the age of data science – or so you might think. Decoration instead of information remains a notorious problem in data visualization, but it is not something that should be allowed anywhere near business intelligence.

4. Deception by cropping – Unsuitable squiggle breaks

Data visualization loves cropping and cropping until the knife is blunt. Authors believe that squiggle breaks free them from all responsibility for data visualization leading to incorrect conclusions. But they are wrong.

5. Iconitis – Meaningless symbols

Pretty icons and pictagrams look good and are useful? Most of them are anything but – and they end up being meaningless instead. Iconitis cuts people into pieces, squeezes meaning into half-dogs, and disrespectfully stretches national flags. In business intelligence? No thanks.

6. Chartjunk – Props with no purpose

Edward Tufte came up with the term “chartjunk” in 1984. You would think that might have been enough time to eliminate the phenomenon from our visual culture. But props with no purpose remain prevalent. Even annual reports still feature moiré.

7. Deceptive graphics – Truncated columns

Data is hailed as the new gold. But all too often, the supposed benefits are skewed by deceptively truncated columns. And Excel, for many the favored tool when it comes to DIY business intelligence, is far from guilt-free.

8. Hard work for the eyes – Incorrect grouping

If you need a legend for your labels, you’re doing something wrong. How can a diagram be expected to replace words when the user has to take the time to read it? The ability to correctly group data is a fundamental tool of every data scientist.

9. Misleading grids – Equal spacing for unequal time periods

Cropping is not the only way to make graphics deceptive. They can cause mischief lengthwise, too. Using equal spacing for unequal time periods takes quite some graphical force – but it happens anyway.

10. Pie chart carnage – Too many segments

Cutting away at pies until everyone has a face full of cream is nothing new – and nothing to strive for. Pies can do more than people think. We show how to breathe new life into the pie chart.

11. Magical axes – Scaling comparable data incomparably

One mistake that irritates us, not least because it is done by people who ought to know better and who have good role models, is when comparable data is scaled so as to be incomparable.

12. Percentage circles – Adding circles not segments

Even the prettiest formats can be skewed. Like when circles are used for infotainment instead of information. A no-go in business intelligence. Both for circles and generally.


Download the white paper “The twelve most common mistakes in data visualization” now

What are the mistakes you should look out for when using graphics and diagrams in business intelligence, performance management, data science, and controlling? We have compiled the most important and most common mistakes in data visualization for you – so that you can avoid the pitfalls as a reader and an author alike. Request white paper now!






* required fields