Data Visualization

Visualizing data is more than transforming it into bar graphs and pie charts. When you choose the right interactive data visualization to highlight the most important aspects of your data, you can illuminate new insights and communicate them more persuasively. And that data storytelling can result in smarter actions and bigger outcomes for your business.

1. Faster decision-making: By viewing and manipulating big data in visual and animation formats, you can understand the story your data tells you at a quick glance, rather than poring over piles of numbers and tables for hours or weeks.

2. More data exploration: The best tools allow you to interact with all your data, directly on the chart to discover hidden patterns, see data relationships and uncover actionable insights — all without IT support.

3. Better track business initiatives: Dashboards help you easily track the performance of business initiatives by allowing you to quickly see how everyday operations affect key performance indicators (KPIs).

4. Extend your analytics investment: Because visuals make it easier to understand data, everyone in an organization — including business users — can explore data and find insights that improve company growth and effectiveness.

Form follows function. In other words, before you choose a visualization type, ask yourself, what is the function of your chart? What relationships in your data do you want to show?

Here we describe and give examples for 9 functions of visualizations and the corresponding chart types for each.

Change Over Time

The function of time series charts is to show how data is trending over any amount of time. For example, change over time charts can give insight into a product's sales over the past five years or a stock’s price over the past five hours.

Chart types:

Distribution

The function of distribution charts is to show how data is spread across a group. This helps you spot outliers and commonalities, as well as see the shape of your data. For example, public policy officials might want to see the demographic or income characteristics of a certain population.

Chart types:

Part-to-Whole

This category of charts is best for showing how a single KPIs or metrics can be broken down into component parts. A good example would be if a marketing leader wanted to see all new leads broken out by their source.

Chart types:

Correlation

The function of correlation charts is to convey relationships between variables. Some charts, like a bubble chart, can represent three dimensions of data, where the size of the bubble is the third value on top of the XY axis values. For example, an executive could see in one chart how sales volume and profit are correlated by country.

Chart types:

Flow/Movement

The function here is to show movement data or the flow of data between conditions. For example, in data science, flow maps show how something like migration happens from one location to another.

Chart types:

• Chord

• Network

• Sankey

• Waterfall

Ranking

The function for ranking charts is to show how a list of data points relate to each other. For example, a bar chart in descending order would highlight which salesperson is driving the most sales.

Chart types:

Deviation

The function here is to highlight variation of data points from a given baseline. For example, a finance leader might want to visualize an organization’s budget surplus vs deficit.

Chart types:

• Bar Diverging

• Bar Diverging Stacked

• Line Surplus/Deficit Filled

• Spine Chart

Magnitude

The function of magnitude charts is to convey relative or absolute comparisons in quantity. An example would be if an executive wanted to compare domestic vs international revenues.

Chart types:

Spatial

Spatial, or geospatial, charts serve the function of communicating geographical locations and patterns in data. For example, police might want to evaluate crime statistics in different parts of a city.

Chart types:

As we stated above, first consider the function of your chart. The Harvard Business Review helps frame this question, suggesting that you determine whether your goal is to declare or explore information and whether this information is conceptual or data-driven.

Using these two parameters as axes, we get the 2x2 matrix below, showing the four main purposes for your visual communication: idea illustration, everyday dataviz, visual discovery, and idea generation.

Image credit: Harvard Business Review

Before you get started with your own visual analytics initiatives, get inspired by today’s best visualization examples. And learn from the grandfather of data visualization, Edward Tufte. In his groundbreaking book, The Visual Display of Quantitative Information, he emphasized clarity, simplicity, and precision to communicate complex information. His tenets hold true today.

And, here are 6 modern visualization techniques:

Titles Matter

Compare the visualizations below and see how much more the charts with effective titles help the viewer quickly understand the information.

Generic Chart Titles

Descriptive Chart Titles

From single spreadsheets to complete business intelligence platforms, you have many choices when it comes to data visualization tools.

Data scientists and data analysts often use open source JavaScript libraries and tools such as D3.js, ECharts, deck.gl, and Vega or programming languages such as Python or R for complex data analysis. Open-source or custom data visualization options offer advantages in cost-effectiveness and flexibility but they pose challenges such as a steep learning curve, limited official support, limited templates, integration complexities, and varying feature availability.

For modern data viz tools, the following capabilities provide the most value:

Suggest visuals with AI

Data analytics tools use artificial intelligence to recommend visualizations that can help even novice users build their own analytics views and discover hidden insights.

Present data in the optimal context

The best tools offer the flexibility to visualize data in the most relevant and intuitive formats and can combine data from multiple data sources to give users the full picture.

Facilitate data discovery

Static charts and linear drill-downs stop short of answering user questions about their data. Tools should allow users to freely explore data in whatever direction their intuition leads them.

Embed anywhere

People need access to visual representations of data and insights no matter which application they’re using. The best tools make it easy to embed analytics wherever people are working.

While some tools can help you make better decisions and track business performance, there are some significant challenges you need to look out for.