What You Should Know Before Building Your First Dashboard?

This is a fourth of the four part series. You can find the previous blogs here , here and here. Here, I would be discussing data visualization from the perspective of various stakeholders.

Today’s blog is for you if plan to build your first interactive dashboard.

Always remember and internalize this – “Data visualization is not an art. It’s science”. There are specific graph type for each data type and you cannot make any graph as per the aesthetic requirement of the dashboard (little more on that later).

There are 4 ingredients, if used intelligently can make an effective dashboard. They are as follows:

Shape

Human mind is more adapt in deciphering some shapes better than others.

Try to figure how many times the big circle is bigger than the small circle in (Figure 1). A bit difficult right!!

Now try to figure out how many times the big rectangle is bigger than the small rectangle in (Figure 2). Slightly easier? This is because human brain is better trained to internalize a rectangle than a circle. Hence, it is important you use rectangle more and avoid using shapes like circle and oval as much as possible. (Note: The answer to the above question is 3 times if you haven’t figured that out yet.)

Size

This is pretty straight forward. The reason why I am including this point here to emphasize that if want to compare 2 or more items then showing the same shape in different size is the best way to do that comparison. However, if you put a shape with different size without the intention of comparison. It is a recipe for disaster. In this case make sure either instances of the said shape is of same size or you can go with different shape.

Color

Color plays an important role in any visualization. However, a word of caution here. You should not over use color. If you plan to use different color palette. It should always be to convey some information rather than it being a part of aesthetic. Color is normally used to show the difference in degree. Generally from most favorable to least favorable. Here, we need to be conspicuous about the culture of the user. Red is seen as unfavorable in America, Europe, Middle East and South Asia. While it is favorable in the regions like Africa, China and former Soviet Union countries. The benefits of color and how to use it effectively is explained in detail in this article.

Location

What kind of graph should be put in which part of canvas again depends on the culture of the user. In the Middle East user will see from right to left in a shape of inverted ‘Z’. In Asian countries like China, Taiwan, Singapore the information flows from top to bottom like an inverted ‘N’ and for the rest of the world information should flow from left to right in a shape of ‘Z’. Once this is ascertain, the first graph should show the highest level view like trend chart and then the subsequent charts should contain more and more detail.

Less is more…

There should also be no more than 4 graphs in one dashboard. Anything more than that will lead to clutter and will result in the loss of interest on the part of consumer. The interaction of 1 dashboard with other and the use of filter should also be minimal in order to be effective.

Types of charts and when to use it

• Bar Graph – When you want to compare 2 or more categorical field on when numeric field
• Line Graph – When you have one numeric field and one date field
• Geo Spatial – When you have latitude and longitude and one numeric field
• Tree Map – When you have many categorical field and one numeric field
• Scatter Plot – When you want to compare the relation between 2 numeric field
• Pie Chart – Don’t use it

This is high level overview on what to do while building a dashboard. Please comment below if you have any question or you can reach me here.