5 things wrong with data visualization dashboards.

Many companies are shifting from Excel spreadsheets towards dashboards, built with data visualization tools like Tableau, Microsoft BI and Cliq. The objective of this shift is to make data-driven insights more widely accessible across an organisation and, to encourage data-driven decision making by managers. Unfortunately, in many cases, a company's investment in data analytics fails to unleash the potential of data across an organization.

Simply put: the industry-standard data visualisation dashboard is failing to win over management. Here are 5 of the many reasons why -

1. Dashboards increase noise and diminish signals.

I have seen dashboards with 12 or more competing charts included on a single page! When dashboards display too much information on one page, we don't know where to look, and as a result of the visual confusion we may overlook the most important messages.

2. Signals are hidden behind click paths making executives search for important information.

Busy executives don't have time to waste clicking around an interface blindly searching for the information they need right now. One VP told me "I don't want another tool to interact with; that's what my team of ten business analysts are paid to do. I want the insights immediately - without clicking around to find them".

3. Confusing use of colour.

Using colour properly is difficult. Each new colour that's added to a dashboard increases the complexity of colour relationships exponentially. Even formally educated artists and designers struggle to use colour properly. It takes years to learn how to control colour relationships.

When colour is applied without proper consideration it can confuse people and work against comprehension of the signal.

4. Dashboards hit completely different datasets with the same hammer.

Dashboard developers typically resort to a small set of familiar charts - pie charts, bar graphs, line charts, maps - to explain completely different datasets. This is like hitting every data problem with a same-same chart hammer.

Repetitive use of the same chart formats works against an intuitive understanding of differences. While template charts are quick and easy to work with, different datasets should be given proper design consideration, in isolation, in order that their signals are described clearly.

5. Sometimes a chart is not even the answer.

Unsurpisingly an organisation's business analysts are numerate, and numbers speak to them, but many of their colleagues are not. People have different strengths and waeknesses and different learning preferences.

For some recipients, a chart is never the right way to communicate a signal to them.

Conclusion:

Dashboard templates are quick and dirty, but they aren't effective communication tools in many situations. Therefore, a company's investment in data analytics may fall flat. However, by engaging a professional information design specialist you can transform the communication potential of data-driven insights across your organisation.


Otomatek: