There are a lot of different ways to visualize data. Of course there are preferred methods for doing so. This post is not necessarily about those practices, but more on the alternatives. I expect some comments on this post because of that ;). With this post I want to talk about time-series and trending alternatives.
From our traditional practices most of the times we visualize trends with a line chart. That way it is easy to spot fluctuations and spikes.
This is exactly the way it is easy and simple to follow how a measure is developing over the year. Now if I wanted to highlight a certain relationship in this chart with color, this chart doesn’t give to much options to use color in an effective manner. I could use markers or use a secondary axis, but that doesn’t always simplify my chart, right?
Looking for information design options
Maybe a bar chart might also be interesting to use. This example shows an intuitive addition of color for the time of year. Not necessarily a very relevant insight. But your intuition probably tells you what the color could mean, even without my explanation.
Interesting enough, without knowing, we created a silent legend for time with this chart. This means that we now know that the more intense the color becomes the further we are in the year. Lets use some of this concept to show why this matters.
Ok, lets think about a somewhat more difficult question for which we want to create some insights.
Our manager has asked us to give insight into weekly trends and how our product groups are performing regarding quantity sold and margin % for those dimensions.
With this we know that there is a need to visualize relationships, which requires multiple variables. My suggestion for this would be to use a scatter chart. A scatter chart can contain at least 4 variables. y-axis, x-axis, bubble size, bubble color. And then I’m not even talking about using composite variables or dimensions. Which could help to differentiate even further.
Remember our silent legend?
Here you can see that the more intense the color of the bubble the further in time we are. This means we are providing some indication for a relation with time. With that we now could spot patterns and/or clusters that potentially could suggest seasonality or trends.
Combining things together
Of course with alternatives like these you could consider not using stand alone data visualizations for that. It is more powerful and comprehensive when you combine multiple visualizations onto one user interface. Have a look:
Within one view we now show the silent legend and the trend for at least one variable, quantity sold. And the insight our manager was asking for now is solved with an alternative visualization which helps us to understand more of the time of year while still including the relationship that was important to get insight into.
Hopefully this gives you some ammo for once in a while trying some alternatives to discover options. Making mistakes, experimenting and diving into alternative solutions is a good way to learn and invent new ways of thinking. Go ahead, do it and see what it will bring you!
Happy designing and may the decision making process be inspired by you.