Data visualization is the process of translating data and metrics into charts, graphs and other visual reports. These visualizations let viewers discover patterns and relationships in the data that they otherwise might not see — helping to turn the information into a cohesive story. Data visualization enables organizations and individuals to gain a clearer understanding of their performance and goals.
There are many types of data visualizations such as the familiar column, line, bar and pie charts. But there are many others including:
Histograms – display data with a single variable using bars of different heights
Combo charts – integrate multiple sets of data using two different chart types
Waterfall charts – visualize positive and negative values over time
Stacked charts – show how the volume of one group relates to another
Scatter plots – used when there are many data points
Bubble charts – demonstrate larger quantities of data by increasing bubble size
Heat maps – showcase large datasets by location
Data visualization software tools can either create these graphics automatically or help users create their own. Many popular data analysis packages like Microsoft Excel have built-in visualization capabilities. Higher-end tools can perform sophisticated queries without requiring users to understand Structured Query Language (SQL) coding — and then render analyses in many types of visual formats. Data visualization can also be used to build dashboards -- usually with two or more data visualizations that help viewers make better business decisions.
Why is data visualization important?
Visualizing information in graphical ways can give users insights into their data. By enabling users to look at and explore data from different perspectives, visualizations can help you identify patterns, connections and relationships within that data, as well as understand large amounts of information very quickly. Other reasons why data visualization is important:
Create a message that’s bigger and more memorable. A picture is worth a thousand words, because visuals are more strongly tied to memory. Visuals help commit important concepts to memory, helping to make sure the “story” is passed on to consumers or employees. Interesting visuals keep an audience focused and attentive, which is difficult in a “short-attention-span” world.
Get a clearer understanding of the “big picture.” It’s difficult to process large amounts of information at once, but data visualization can help. Users may get a more accurate idea of what is really going on in the organization by using graphics to gain better insight. They’re able to better monitor key performance indicators (KPIs) and identify important patterns and trends — as well as outliers. Visuals may bring into focus subtle correlations and relationships happening within the organization that a user might not see looking at the data alone.
Make better decisions faster and act quickly on them. With data visualization, users can better understand an organization’s best next steps and spend less time performing data analysis. They can quickly review strategies, make updates efficiently and act on decisions faster — achieving success with greater speed.
Allow everyone to participate in decision making. Traditionally, the only people who could understand company data worked in the IT department. Now, with data visualization software, finance, sales and marketing teams can create their own graphics and clearly communicate business strategies to their teams. Visualizations can be shared among teams, and a broader spectrum of people can act on the information, instead of attempting to decipher a massive spreadsheet.
Using software for data visualization
Data visualization software can track data in real-time and visualize an organization’s metrics, goals and KPIs. Data visualizations help to communicate the company’s “story” and focus teams on prioritizing goals that need the most attention. But with so many tools available, how do organizations decide?
Pam Baker, in her blog, The best data visualization tools for 2019 (1), provides five criteria for choosing tools for data visualization:
What kind of visualization does the tool support? — Review what kinds of data your organization collects and how it uses that data. Most vendors offer free trials. Start by experimenting with various tools to see what works best for your organization.
Which data formats does the query function support? — Look for SQL and NoSQL databases, and specific apps such as Oracle or SAP Financials, as well as sales tools like customer relationship management (CRM) apps and email marketing software. They should cover business platforms that your organization is actively using.
How well can the tool drill down into source data? — How hard is it to drill down beyond first-tier queries? Can it drill down on a live data visualization? Your organization may find sophisticated drill-down capabilities critical, allowing you to change the visualization data immediately without completely starting over.
How comprehensively can the tool export visualizations? — What are your options for exporting it to other people for their use? Look for a variety of flat graphic formats (CVS, JPEG, PDF). Also, can the tool work with code snippets that can be dropped directly onto webpages, incorporated into other apps through open application programming interfaces (APIs), or rendered on desktops and mobile devices?
What are the product’s advanced processing capabilities? — If your business is collecting big data or developing an IoT offering, advanced processing is an important function. Some tools simply rely on back-end data warehouses to do most of the query processing.
Choosing the right data visualization tools and how to use them
Graham Turney in his blog, How to choose the right data visualization tools for your applications (2) provides helpful tips to finding a solution that meets the needs of end users — without high costs or requiring significant development resources.
Make navigation easy. Tools that require coding or understanding of relational databases are generally poorly suited to business users. Simple drag-and-drop functionality helps them get the most out of the visualization tool.
Save time. Having visualization capabilities embedded within the application itself provides a convenient and seamless user experience.
Allow customization. It’s critical to use dashboards with a live connection to the data and dynamic rendering capabilities that users can customize. Executives need the ability to sort, filter and drill down into the data to address questions on the fly.
Avoid hidden costs. Factor things like contract requirements, data integration needs, implementation time and ongoing maintenance into the price. Look for a solution that’s quick and easy to implement up front and that won’t require hours of development time.
Once the tool is deployed, how do organizations create memorable and effective visualizations? IBM Design provides a step-by-step strategy:
Define your intent with users. Identify what teams would benefit most from data visualization and what are their issues. Rely on simple charts up front and don’t be intimidated by working with huge amounts of data.
Understand and clean data. Look at your data set for its structure and the typologies of data included (strings, numbers, dates). Analyze rows and columns by observing their relationship and implications. Clean data by looking for inconsistencies, remove duplicates and check for character encoding.
Model data and check for visual validity. Rely on simple visual representations of data to identify patterns and trends or to verify a hypothesis. Use basic visual models to actually see the data, especially when the data set is too big to be understood directly.
Experiment with structure and style. Represent the information in new and imaginative ways. Draft different versions of your design and try a variety of charts — while staying within established organizational graphic standards.
Test and iterate. Gather impressions and opinions from users, or from people who have never seen the project. Always use research to influence iterations and justify necessary changes.
Refine and implement. Look for bugs and functional errors. Check your visualizations for the following common human perception pitfalls:
We don’t see first what stands out. Instead, we look for differences.
We only see a few things at once.
We seek meaning and make connections.
We rely on connections, metaphors and visual signals to understand what is expected.
IBM Cognos Analytics
Easily create your own compelling visualizations or use those recommended by the system and get additional data insights presented in natural language.