Approach by data visualisation

It is evening and already dark. I feel forced to stay at home because of the cold October weather. In order not to be bored I decided to make a post. This post will be the special one because today I'm not going to describe how to do any graphics with tableau. Today I'll share with you my approach by data visualization.

As a data analyst with 1,5 year of work experience I have to create a lot of dashboards. My challenge is to create a reasonable dashboard with unknown data. So this is my approach:

The first thing I do is ask the questions

Who is my audience? (Management, Marketing, HR etc. -with this information you can decide what kind of KPI are important)
What is relevant for my audience? (strategically, analytic or operative KPIs)
Which data I need in order to create desired dashboards?
Which data I miss?
What is my deadline?

In the second step I begin to study the data.

While my first year as a data analyst I was always overdone with data, when I opened an unknown file. In order not to drown in data float I learned to swim. I touched the "water" slowly before I submerge in this "water". I concentrate on one data set and ask again some certain question. (This question should be based on the expectations of your audience)

- Can I calculate all relevant KPIs?
- Can I compare all relevant KPIs?
etc.
If you cannot answer your questions with data then you need to obtained additional data.


Third step is one of my favorite steps: Draw you ideas

After data exploration I know that I have all data I need and I know from my audience what they expect from dashboards. So in the third step I do some visualization drafts. Be careful with different forms and colors for your graphs, as it is very important. You can find a lot of useful advises in this whitepapers.

Usually I am succeed in implementations of my drafts. But sometimes I  have to experiment with data, as some drafts appears not concise enough after visualize them with tableau. For more information I advise you this books: Einstieg in die Demonstrativeness's or Visualize this

The last step - my absolutely favorite one: data visualization

Have you ever heard about "10 sec rule"? It means that the dashboard should be understood within 10 second. In order to reach this, here are some advices: 

1. Choose a meaningful title

Watch on headline (fat and large)/subheading (less fat and less large). Also experiment with font size and font art.

2. Create a visual hierarchy

Choose light color or make the elements, which are irrelevant, weaker, in order to push them to the background. Use arrows or lines in order to focus on essentials.

3. Readability

Think about what you audience graphical or statistical cannot understand and add you graph with 

description /hints or comments. With the short description your story will be easier to understand. Graphs and visual hints increase the readability and help to interpret data easier. 

Don't forget to add some trends. You can also comment them, as comment can clarify your story. Try to answer such questions like: what does this deviation mean? What does this trend mean? etc.

4. Useful comparisons

Comparisons helps to understand data better because if you see any equivalent data you can estimate and evaluate them. Think about which data can be compared and visualize them side by side.

5. Negative spaces

Choose the distances between graph and also choose the spaces inside the graph carefully. It is important if you build some groups or if you want to represent some graphs side by side

6. Highlighting

Highlighting can guide the reader through your data and route the eye of reader toward the important information. It enhances what the reader maybe already seen or they guide the reader to the information which should be also established. On this point use you creativity, choose strong colors and discover the best color combination for you.

And last but not least: No matter how good your visualization is, the most important thing that your calculation and analysis are correct and understandable for you audience. You have to earn the trust from audience! Explore your data carefully. Note, that your audience make important decisions based on your dashboards.

If you have any questions feel free to contact me.