Having more data than we know what to do with, we turn our attention to surfacing the most valuable of it. Data visualization is beginning to get its due as the premier means to give us humans the keys to the data kingdom as intuitively and quickly as possible.
Now, data visualization can be used to accomplish at least two things. On the one hand, it can provide a deep and insightful narrative about a topic by viewing a variety of measurable facets of it. On the other, it can offer exploratory or definitive analyses to drive a variety of evaluations and decision-making, low-level and high-level, enterprise-wide.
Data viz is peerless at presenting analytics visually and making complex data accessible and understandable. It helps managers and CEOs alike grasp comparisons, recognize patterns, and identify trends that would otherwise get lost in the sea of data details.
So where do you start if you want to harvest the power of data visualization?
I recommend you start with questions.
Who’s your audience?
Take some time to consider the audience for your data visualization application. What level of detail is best suited for them? How tech savvy are they? How familiar are they with the topic, the context and the information that you’ll provide them with? How comfortable are they with data visualization in the first place? And don’t forget: how much time do they have to spend with the info you’re giving them?
The answers to these questions will guide you to determine how much detail, background and context will serve your audience best.
What do you want your data viz to accomplish?
Do you want to tell a story and educate people? Or do you need to highlight conclusive insights that result in actions taken and decisions made? Design your layout based on what you want to communicate. Usually, you’ll want to present the broadest perspective chock full of conclusions and generalities and omit the details. What’s the significant trend you’re showcasing? What’s the most noteworthy aberration? What comparisons lead to conclusions your audience will want or need to act on? Focus on the highlights.
What’s the best format?
Believe it or not, data visualization is all about relationships. Whether you’re comparing, monitoring, or highlighting, you’re manipulating views of relationships between metrics. So choose the tools that are the best match for the purpose. You’ll end up with the most intuitive — and therefore most effective — visual representation of your data.
For example, simple line charts work well for charting progress over time. Bar charts work well to compare similar items across one dimension. Pie charts show proportions compared to the whole, but never use them for more than 5 elements at a time. If your variables exceed 5, use a bar chart. Then add color to simply highlight the channel, product, rate — or whatever you’re measuring — that has the most consequential story to tell.
And by the way, don’t overcrowd. And don’t use 3-D. They do more harm then good.