We’ve talked about the basics of business intelligence, dashboards, and dashboard design, but what about the magic that turns raw numbers to beautifully designed charts and graphs… you guessed it: data visualizations.
As humans, we’re wired to connect with stories. Data in its rawest form is anything but a story—if anything, it feels like the exact opposite.
But when you watch the numbers transform into visualizations? That tells a story with a clear beginning, middle, and end. There’s no guessing if your protagonist survived the harrowing fall from the cliff or if the star-crossed lovers will find their happy ending. Visualizations tell you everything you need to know to understand your business story, from start to finish.
In a lightweight BI tool like PowerMetrics, you can create data visualizations even if you’re not a data analyst. After you connect your data through a pre-built or custom connector, you can watch it transform. And all you have to do is select the visualization type that effectively communicates the story or message that you want to bring to life with your data.
With the help of this article and our additional starter guides mentioned above, my hope is that you will be equipped with all of the tools you need to tell—and understand—your data story.
If you’ve got questions along the way as you design and share your dashboards, we’re here to help. We want you to be able to confidently say, I can do anything with data — and mean it.
Here’s what I’ll cover in this article:
- What is a data visualization?
- Why do we visualize data?
- What makes a good data visualization?
- How do I pick the right visualization for my dashboard?
- 12 commonly used data visualizations
- How to level-up your data visualizations
What is data visualization?
A data visualization is a graphical representation of data or information—for example, taking raw numbers from a spreadsheet and transforming that into a bar or line chart. Using data visualizations helps you to easily understand and analyze trends and outcomes within your data.
Selecting the right graphical representation depends on the story you are trying to tell. You can choose any visualization and assume it’ll do the trick, but it’s worth the time and effort to select and customize the right visualization for long-lasting impact.
Where do you put data visualizations? Data visualizations are everywhere: on the evening news, embedded in tweets, or on the front page of the newspaper. In a business context, you often find data visualizations in reports, presentations, or on a business dashboard.
Business dashboards display data visualizations to help you:
- Keep a pulse on key metrics related to the health of your business
- Integrate your data sets and data sources in one place
- Provide an at-a-glance view of business performance for key decision makers
- Get your team on the same page regarding performance and outcomes
- Intuitively and logically interpret data
Why do we visualize data?
We use data visualizations to help us tell stories. By nature, humans are visual beings—we’re drawn to colours, shapes, and patterns. So using those basic human conditions to turn numbers into visuals is an incredible skill to have in your back pocket.
You’ve likely heard the old adage: a picture is worth a thousand words. Might I suggest a slight edit: a data visualization is worth a thousand words.
When done well, a data visualization will tell the story that is hidden within the numbers in your spreadsheet without words. If you’ve had an excellent quarter and your revenue is up, your line chart will be making a slow climb to the upper right. That’s an easier story to capture than sifting through the rows of an Excel file, right?
When you put your data visualizations on a dashboard, you have the tools and information you need to make and validate your business decisions in an accessible and easy-to-share format.
You should put data visualizations on a dashboard to:
- Boost engagement
- Improve analysis
- Save time
- Tell a story
- Identify relationships
- Compare performance using numbers
Data visualization boosts engagement
Reporting on your key business metrics can be challenging. Not only do you have to get your audience’s attention, but you then have to encourage them to take action based on the available insights. Data visualization supports both of these goals.
Beautifully designed and easy to consume data visualizations capture the attention of your audience and help them analyze and action the key takeaways.
Data visualization improves analysis
Data visualization makes it easier to see trends and predict outcomes. Think about how difficult it can be to glean insights when you’re analyzing the columns and rows in a spreadsheet.
Data visualizations allow you and your team to easily compare your business performance over time or compare how your marketing campaigns perform across different platforms. And your visualizations may not always call out your wins, either—use data visualizations to quickly understand how and where you are underperforming, too, so you can quickly change your approach.
Data visualizations save you time
Looking at raw data forces you and your team to do the math. Why not have the math done for you? Proper data visualization does just that. It’s easy to get the information you need from your data when it’s presented in a visual format. Add in performance metrics or comparison values and all of a sudden you have a direct line to the health of your business on a compact dashboard that you can share with your team.
Data visualizations tell a story
Visualize your data and watch a story come to life. Data visualizations tell stories, and while it may sound contradictory, meaningful storytelling is about showing, not telling. The right visualization can tell the story of a business with minimal words. Incorporate data visualization into your storytelling toolkit and you’ll be writing a happy ending.
Data visualizations allow you to identify relationships
One of the most powerful attributes of data visualization is that you can draw comparisons and identify relationships. This isn’t as easy in a static spreadsheet. Data visualizations enable you to see trends and track your KPIs so you can make decisions that are backed by data.
Data visualizations compare performance using numbers
Visualizations have an impact when you’re comparing data. Raw numbers in a spreadsheet don’t tell the story with the same impact. If your line chart is falling off a cliff, that’s pretty dramatic and will likely get your attention. A string of large numbers in a document is much more difficult to digest. Proper use of visualizations will eliminate any mystery - the comparison will be clear and obvious, so use it to your advantage in decision-making.
What makes a good data visualization?
Data visualization is an art form, and a visualization’s effectiveness is dependent on the execution. It’s only effective if it’s done right. Have you ever opened a spreadsheet or a presentation deck, only to find out-of-context numbers and thought to yourself “Everyone around me is confident and I don’t understand why these numbers are important.” In truth, everyone else is thinking the same thing about those numbers. Data visualization makes data analysis accessible. Everyone can understand complex data and by association, everyone feels confident in their decision-making and contributes value to important conversations.
We define good data visualizations as graphical representations that serve their intended purpose. If a user can interpret your visualization by asking questions about the information displayed versus how or what is displayed, then you know you’re on the right path.
Now you may be wondering, “How do I pick the right visualization to ensure it serves its intended purpose?”
Let’s break it down.
How do I pick the right data visualization for my dashboard?
It’s not as easy as picking any data visualization to present your data and information. It’s really important to match your data to the right visualization so you and your users can get the most value. If you’re not sure where to start, ask yourself these five questions.
- What relationship am I trying to understand between my data sets?
- Do I want to understand the distribution of data and identify outliers?
- Am I looking to compare multiple values or looking to analyze a single value over time?
- Am I interested in analyzing trends in my data?
- Is this visualization an important part of my overarching data story?
Answering these five questions will help you narrow down which category you should focus on, and then drill down further to which type of visualization within that category is the best fit.
Data visualization categories
There are 5 types of data visualization categories:
Temporal data visualizations
Temporal data visualizations are linear and one-dimensional, and most commonly used to represent a time series.
Temporal visualizations come with a sense of familiarity, too. Temporal data visualizations are common in newspapers to show information like housing market fluctuations quarter over quarter, or in company reports to visualize gains and losses.
The advantage to using a temporal data visualization is that we have a predisposed understanding of how and when to interpret them, so it gives your users an edge when they look at the data.
Examples of temporal data visualizations
- Bar chart
- Line chart
- Scatter plots
- Polar area diagrams
- Time series sequences
- Gantt chart
Hierarchical data visualizations
Hierarchical data visualizations order a collection of items that link back to a parent item. Hierarchical visualizations are best used to display a cluster of information, especially if it flows from a single origin point (like a tree diagram).
There is a time and place to use a hierarchical data visualization, too, as they tend to be more complex and challenging to read, but it is the simplest hierarchical visualization to follow due to it’s linear path.
Examples of hierarchical data visualizations
- Tree map or diagram
- Ring charts
- Sunburst diagrams
Network data visualizations
Network data visualizations show relationships between entities—nodes (the circles on the visualization) and links (the lines that connect to the nodes)—without the use of words.
Examples of network data visualizations
- Matrix chart
- Node-link diagrams
- Word clouds
- Alluvial diagrams
Multidimensional data visualizations
Multidimensional data visualizations have multiple dimensions. Due to the dimensionality, this type of visualization tends to be the most vibrant and eye-catching visualization type.
If you’re looking to drill down and filter your data, this is the best type of visual to use because you can break down your data in a number of ways to capture the key takeaways.
Examples of multidimensional data visualizations
- Scatter plots
- Pie charts
- Venn diagrams
- Stacked bar graphs
Geospatial data visualizations
One of the earliest forms of visualization, geospatial (or spatial) visualizations overlay familiar maps with data points. Geospatial data visualizations have a long history, too, as it was used for navigation before computational analysis came along.
Examples of geospatial data visualizations
- Flow map
- Density map
- Heat map
12 commonly used data visualizations
While there are numerous data visualization types available, you should choose the right one for your audience. In a business context, it’s important to choose the visualization that will help you extract the most value from your data displayed on a dashboard. It can be hard to know which visualization is best for your data set, your dashboard, and your users, but my hope is that this guide will help you do just that.
Once you pick a data visualization, it’s important to consider how it fits into your overall dashboard design, too. We have all the tips and tricks to help you design an incredible dashboard in our guide to dashboard design.
In a lightweight modern BI tool like PowerMetrics, there are a number of data visualizations to choose from, including:
- Bar or column chart
- Line or area chart
- Pie or donut chart
- Tree map
- Radar chart
- Waterfall chart
- Heat map
- Summary chart
- Scatter chart
- Bubble chart
- Combination chart
Let’s take an in-depth look at each data visualization type.
Bar or column chart
You’ve likely seen, interacted with, or built a bar chart before. I’d even be willing to bet that a bar chart is one of the first data visualizations you were introduced to! Bar charts relish in popularity because they’re easy to read and easy to understand.
Most commonly used to compare related data sets, bar charts organize data into rectangular bars proportional to the value it represents.
The x-axis of a bar chart shows the categories that are being compared and the y-axis represents the value. For example, let’s say you want to visualize real-estate market data: you could plot types of homes being sold (townhouse, condo, or detached) on the x-axis, and dollar value that the home sold for on the y-axis to help you understand the type of homes available within an allocated budget.
Use a bar chart to:
- Compare two or more values in the same category
- Compare parts of a whole
- Compare less than 10 groups of related data
Bar charts aren’t suitable for visualizing a category with only one value or visualizing continuous data.
Bar chart best practices
- Use consistent colours and labels to easily identify relationships in the data
- Simplify the length of your y-axis and start from 0 so your data is orderly
Line chart or area chart
Much like bar charts, line charts are a popular way to visualize data in a compact and precise format. Data points are represented by dots that are then connected by straight line segments.
Line charts visualize your data relative to a continuous variable, usually something like time or money, and the data points are ordered by their x-axis value.
It’s important to consider colour in a line chart. Different coloured lines make it easier to interpret the information being presented. You can read more about proper use of colour in our dashboard design guide.
Use a line chart to:
- Understand trends, patterns, and fluctuations in data
- Compare different but related data sets with multiple series
- Make projections beyond your data
Line charts aren’t the best visualization if you want to demonstrate an in-depth view of your data.
Line chart best practices
- Use a different colour for each category you’re comparing
- Use solid lines to keep the chart clear and concise
- Avoid comparing more than four categories (it can become confusing with too many!)
Pie or donut chart
Appropriately named, a pie or donut chart visualizes your data in slices within a circular graphic. Each slice of the pie represents a segment of your data. Pie charts are visualized in a full circle, whereas a donut chart, well, it looks like a donut--it has a hollow centre!
Pie charts differ in appearance from a bar chart, but serve the same purpose since they compare values in the same category. Pie charts are easy to read because the parts-of-a-whole relationship is obvious at first glance. There is a disadvantage to pie charts though; the percentage of each section isn’t obvious without adding numerical values to each slice. Keep this in mind if you’re putting a pie chart on a dashboard that is commonly used for quick analysis.
Pie charts are still a quick way to scan and gather insights as long as you follow the best practices outlined below.
Use a pie or donut chart to:
- Compare relative values
- Compare parts of a whole
- Rapidly scan your data (keeping in mind that the numerical values have to be added to each slice for deeper analysis)
Don’t use a pie chart for precise comparisons of data, not because it’s not capable of it, but there are better visualizations to choose from if precision is a requirement.
Pie or donut chart best practices
- Make sure the slices of your pie (or donut) equal 100%. To make this easier, add numerical values and percentages to your visualization to help you and your readers
- Order the pieces of your pie according to size
- Don’t compare more than 5 categories in a pie chart, otherwise you run the risk of unclear differentiation between slices
Tree maps are a hierarchical data visualization. Each category is divided into segments that represent a whole. Within a tree map, each branch is a rectangle which is then associated with smaller rectangles (or sub-branches). The rectangles, or sub-branches, are sized proportional to the data.
For example, if you have a metric that represents total house sales for three different real estate agents, the agent with the least amount of sales would have the smallest rectangle.
Tree maps are an excellent way to visualize and drill down your data into layers to show the hierarchical relationship between items.
Use a tree map to:
- See patterns in the branches due to correlation between colour and size
- Display large data sets simultaneously while making efficient use of space
Radar charts help to understand relative differences between items in your data. Due to the design of a radar map visualization, you can easily compare multiple items and identify outliers that require attention.
Think of a radar chart like a bicycle wheel. Each data point is plotted on an individual axis, or a spoke, that starts from the centre of the diagram. The diagram has circular grid lines that connect each axis. Once the data is plotted on each spoke, it is then connected together through a series of straight lines to create a polygon so you can clearly call-out any outliers or identify commonalities.
Use a radar chart to:
- Plot small to medium sized data sets (too many variables means too many axes)
- Substitute a line chart when you have limited space on a dashboard (like mobile viewing or TV displays)
If you plot too much data on a radar chart (consider how each data point gets its own separate axis), it can easily become overwhelming.
Waterfall charts visualize how an initial value is affected by positive or negative intermediate values (time or category based) over time. The final column represents the cumulative value.
Use a waterfall chart to:
- Reveal the composition or makeup of a number
- Display quantitative analysis like financial gains or losses over a period of time
Waterfall charts aren’t appropriate if you want to visualize more than one number or metric.
Waterfall chart best practices
- Use contrasting colours to call-out any differences in your data set
- Choose warm colours to indicate an increase and cool colours to indicate a decrease
Heat maps (also known as a matrix) compare information in the same category by way of colour and saturation to identify differences in the value. Warm colours typically represent high-value data points and cool colours represent low-value data points. Mousing over each square will provide you with the numeric value.
Heat maps can also be thematic, which means that the area inside the boundary is shaded in proportion to the data being represented.
Use a heat map to:
- Rate your values on a scale (for example, high to low or works well to needs improvement)
- Show a relationship between two measures
- Illustrate an important detail
Heat maps are not a good visualization to use if you want to visualize an individual metric. Heat maps don’t display the numeric value without having to mouse over, so they are best served on a dashboard that will be viewed in a browser when there is ample time and space for exploration.
Heat map best practices
- Use a simple map outline to avoid unnecessary distractions
- Use a single colour in varying shades to show changes in the data
- Avoid using multiple patterns
Summary charts display a single numeric value. If you need to simply communicate a sum or single value on your dashboard, this is a great visualization to pick. Additionally, summary charts have the capability to display a comparison between two time periods.
You can display a comparison between two time periods as a percentage change or value change (between the current value and the previous value) or you can display both values together for a comparison value. Remember, red typically indicates a negative trend and green indicates a positive trend. If a comparison isn’t relevant to the data you are visualizing, it’s best to not add one at all.
Use a summary chart to:
- Compare two values over a period of time
- Display a sum value in a compact and precise visual
A table chart displays your data in columns and rows, not unlike a spreadsheet. Tables also allow you to display numeric values and graphics like bullet charts, icons, or sparklines.
There are a few different table types to consider when you pick your data visualization.
List tables, similar in appearance to a spreadsheet, display your data in rows in the order it was returned.
Pivot tables display values as plain text. Pivot tables organize the axis data hierarchically so sub-headers are grouped under the parent header. An expert tip: Pivot tables identify blank entries. In other visualization types, like a bar or line chart, a blank entry would be hidden within the total value.
Ranked tables allow you to display your data in ascending or descending order. Ranked tables also allow you to add a comparison value between equivalent time periods. Comparison values have a time and place. For an in-depth look at comparison values, check out our dashboard design guide.
Use a table to:
- Categorically display two-dimensional data sets
- Drill-down large data sets with a natural drill-down path
Tables aren’t the best option to display a large data set. For clarity, we don’t recommend more than 10 different rows. Keep in mind, if your dashboard isn’t displayed on a browser, your users probably won’t have the means to scroll through multiple rows in a table. It’s best to reserve tables for exploratory dashboards.
Table chart best practices
- Be mindful of the order of your data, ensure that labels, categories, and numbers take priority over graphics
- Try to avoid more than 10 different rows in a table to avoid clutter (dashboards should be easy to read at-a-glance)
Do you have a lot of different data points that you want to visualize in the same set? Scatter charts are a great way to do this.
Like a radar chart, scatter charts are helpful in identifying outliers or understanding the distribution of data.
If the values on your chart form a band that extends from the lower left to upper right, this usually indicates a positive correlation between variables. Alternatively, if the band extends from upper left to lower right, it indicates a negative correlation. No pattern? No correlation!
Use a scatter chart to:
- Show a relationship between two variables
- Display a compact data visualization
Scatter charts aren’t appropriate if you want to quickly scan your dashboard for information or if you need precise data points.
Scatter chart best practices
- Trend lines are an excellent way to analyze the data on a scatter chart but we recommend you stick to 1 or 2 trend lines to avoid any confusion
- Start at 0 for the y-axis so your data is orderly
Most charts have a value axis and a date or category axis. Where a bubble chart differs is that it intersects three values—on the x-axis, on the y-axis, and the third displays as the bubble size (the bigger the value, the bigger the bubble).
For example, you could plot the data for the North American real estate market and the European real estate market to compare average house price (y-axis) versus average income (x-axis) versus house category—detached, townhome, condo (bubble size). A bubble chart enables you to compare three values about a particular geographic area against another.
Like the scatter chart, bubble charts allow you to quickly identify outliers and correlations in your data set.
Use a bubble chart to:
- Show a relationship between three variables
- Display a compact data visualization
- Call-out outliers and correlations in your data
Bubble chart best practices
- Use a smaller set of data—the more values, the more bubbles which can be means for misinterpretation
Aptly named, combination charts compare two data sets over the same dimension (date or country, for example). Combination charts display a set of values as columns on the left axis, and a set of values as a line on the right axis, appropriately combining both into a combination bar/line chart.
Combination charts are unique from other data visualizations because you can compare two data sets that have different numeric scales or formats—like comparing dollar values and percentages in the same data visualization.
Use a combination chart to:
- Compare to dissimilar data sets in the same visualization (i.e. comparing currency and percentage)
How to level-up your data visualizations
When it comes to dashboard and data visualization design, you should seek input from the people who will be using it. Develop iteratively. And know that you probably won’t get it right the first time. Make sure you invest time to ask for feedback, understand what’s working, and what needs improvement. Data visualization is a skill, and like any other, it has to be practiced.
The best piece of advice I can give? Have fun with your data visualizations. If you’re looking to level-up your skills outside of business context, create a data visualization for your fantasy hockey or football league, or the stock market, or even your favourite film genre. One of the greatest corners of the Internet to get inspiration and understand the limitless possibilities with data visualization is r/dataisbeautiful. Or show off your visualization by tweeting at us! Be creative, explore, and push the limits. That’s what data visualization is all about.
Start building your PowerMetrics dashboard today
- Sign up for a free PowerMetrics account. You can get started here.
- Check out our latest PowerMetrics how-to series on YouTube: How to build a dashboard, how to visualize your data with different chart types, how to segment and filter your data, and how to apply dashboard filters.
- Explore MetricHQ, find the metrics that matter most to you, and try Instant Metrics.
- Subscribe to our weekly Metric Stack newsletter to get a new benchmark in your inbox every week, in addition to what's trending in the world of metrics, expert interviews, and tips and tricks to help you level-up your analytics.
Topics: Data Visualization
Originally published October 6, 2021, updated Oct, 07 2021