Bold BI

Unlocking Data Insights with Appropriate Charts

Unlocking Data Insights with Appropriate Charts

A data-driven approach is taken in many businesses. Strategic decisions are being made based on historical data analysis and interpretation. To transform raw data into actionable insights, appropriate metrics need to be showcased through the right data visuals. In this blog post, we are going to discuss charts as data visualization tool and which types are best to showcase analytics based on the KPI we need to measure.

Why charts?

A chart is a graphical representation of data that helps people easily understand large quantities of data and relationships within it. Choosing an appropriate chart is very important to visualizing data. First, you must:

  • Understand why you need a chart
  • Determine your visualization goal
  • Select the right chart to achieve that goal.

Selecting the right chart

Ask yourself the following three questions when choosing a chart type for your data:

  • Do you want to compare distinct categories?
  • Do you need to show proportions of data?
  • Are you analyzing distribution of your data?

Comparing distinct categories

To compare distinct categories (dimensions) with measurable values, the following chart types are preferable:

  • Column chart
  • Bar chart
  • Stacked column chart
  • Stacked bar chart
  • 100% stacked column chart
  • 100% stacked bar chart

Understanding relationships within data

To understand relationships among individual parts of a whole data collection, the following chart types are preferable:

  • Pie chart
  • Doughnut chart
  • Pyramid chart
  • Funnel chart

Analyzing data distribution over time

To identify distribution of data over a time period and understand the outliers, trends, and range of data values, the following chart types are preferable:

  • Area chart
  • Stacked area chart
  • 100% stacked area chart
  • Line chart
  • Spline chart

Now let’s see each of these chart types and their use cases to help you choose appropriate charts to unlock required insights from your data.

Column chart

A column chart allows you to compare different values for a set of unordered items across categories. Each categorical series is represented through vertical bars ordered horizontally. It is also suitable to display negative values.

Best practices

  • Use a column chart only to represent a smaller number of data points.
  • For longer category labels with less space to accommodate them horizontally, change the rotation angle of the label.
  • For time series distribution, if data has empty values, this chart can be used as a replacement for a line chart. A line chart will consider empty points during line rendering, which leads to a misunderstanding of trends due to discontinuity among the points.

The following example showcases a column chart visualizing the total number of support tickets resolved in each category of customer service in an IT company. This lets an IT operations team decide where they should improve to perform better.

IT customer service tickets resolved in each category
IT customer service tickets resolved in each category

Here is another example showcasing a grouped column chart. The bars are clustered into groups of more than one, each representing a measure column. In the education domain, school administrators can easily track the overall number of male and female students enrolled by grade through this chart visual.

Student enrollment details—school performance
School performance – student enrollment details

Bar chart

A bar chart allows you to compare different values for a set of unordered items across categories. It’s like a column chart, but each categorical series is represented by horizontal bars ordered vertically.

Best practices

  • Use a bar chart as a replacement for a column chart when data labels are long and need to display and compare many categories.
  • Avoid showcasing time-series distribution, as time data is best visualized rendered left to right.
  • Get the values sorted in descending order to showcase data better.

The following bar chart shows the average days taken to settle policy claims by an insurance company. With this metric, a company can see whether they’ve met or exceeded the defined average settlement time. If exceeded, they can allocate the required number of resources to handle more claims at a time.

Insurance—average days to settle a claim
Insurance – average days to settle a claim

Stacked column chart

A stacked column chart allows you to compare categories with multiple measure columns stacked one on top of the other vertically. With one series and one measure column, it looks like a column chart. To compare multiple series, you can even group the stacked bars horizontally.

Best practices

  • Use to show data changes over a time period.
  • Limit to fewer series and categories to avoid complexities during comparison.

In the following example, the stacked column chart is used to visualize the lead time to produce a drug in the pharmaceutical industry, split into different phases (planning, manufacturing, inspection).

Lead time to produce a drug in a pharmaceutical industry
Lead time to produce a drug in a pharmaceutical industry

Stacked bar chart

A stacked bar chart allows you to compare multiple measures by stacking the values in series one after the other horizontally.

Best practices

  • Don’t use too many series in each bar or it will be difficult to understand.

The following example showcases monthly expenses in each IT department, such as development, infrastructure, maintenance, support, and administration in a stacked bar chart. With this, a project manager can identify the departments that require more or less resources and decide whether to add or reduce funds to them to efficiently meet everyone’s needs.

Monthly expenses for a project in an IT industry
Monthly expenses for a project in an IT industry

100% stacked column and bar chart

100% stacked column and bar charts let you compare multiple measures through bars stacked one on top of the other, filling 100% of the chart area vertically and horizontally. It is like stacked column and bar charts, but each series bar represents a percentage of a total sum.

In the following example, a 100% stacked column chart presents profit through various channels in the pharmaceutical domain. It visualizes the profit earned as a relative percentage through multiple channels in each quarter. You can compare it with previous quarters.

Pharmaceutical sales—profit by channel
Pharmaceutical sales – profit by channel

Here is a 100% stacked bar chart showcasing claim amounts for each policy type given by an insurance company. It benefits the insurer to know how well these policies are working for the policy holders in order to make any improvements.

Claim amount comparison—insurance company
Insurance company – claim amount comparison

Pie chart

A pie chart helps you visualize the proportionality of each category to the total in the form of pie slices. If data labels are represented in percentages, then the total sum of all slices should always equal 100%.

You can find more detail on when to use a pie chart and its best practices from another blog post.

The following example shows a patient satisfaction rating in the healthcare industry. This information helps hospital management and executives to identify areas for improvement.

Patient satisfaction—healthcare industry
Healthcare industry – patient satisfaction

Doughnut chart

A doughnut chart shows what proportion of the total each category constitutes in the form of doughnut slices. It can be used for comparing categories and groups.

Best practices

  • Display only aggregated data that can be grouped into seven or a smaller number of subcategories.
  • Represent each slice with a unique color.
  • Show data labels in percentage format.

The following example showcases average energy consumption by different sectors, such as industrial, commercial, traffic, and residential in an energy industry through a doughnut chart.

Average energy consumption by sector—energy domain
Energy domain – average energy consumption by sector

Pyramid chart

A pyramid chart allows you to make proportional comparisons among values displayed in a progressively increasing manner from top to bottom. It’s best for showcasing hierarchical structure of quantity or size.

Best practices

  • Use different colors for each segment for better clarity.
  • Don’t add too many layers and colors, for the sake of clarity.

The following example showcases the page views of a website by channel. With this, a marketing team can focus on increasing promotions in relevant channels to bring in more leads and increase website conversation rate.

Website page views by channel
Website page views by channel

Funnel chart

A funnel chart allows you make proportional comparisons among values presented in a progressively decreasing manner from top to bottom. Its shows values across different stages with different colors. So you can easily identify bottlenecks of data.

Best practices

  • Apply different colors for each section as the funnel decreases.
  • Use each section size to reflect the size of the data points.
  • Use to showcase sales opportunity pipelines and website visitor trends.

The following funnel chart shows a website conversion rate in a decreasing manner. Marketing analysts, CEOs, and executives can track website traffic such as clicks, impressions, downloads, and signups.

Website conversion rate
Website conversion rate

Area chart

An area chart allows you to compare values for a set of unordered items across categories by visualizing them with filled curves ordered vertically.

Best practices

  • Visualize trend changes over a continuous period (i.e. it’s best to present time-series data sets).
  • Avoid comparing too many items.

The following example gives you a quick understanding of the trend for planned versus completed work items in a scrum team over a week. This lets the scrum master identify whether any story is underestimated or overestimated and fill the gap with appropriate action.

Sprint management—planned vs. completed work items
Sprint management – planned vs. completed work items

Stacked area chart

A stacked area chart allows you to compare multiple measures through filled curves stacked one after the other vertically. For multiple series rendering, series are stacked on top of other series to avoid overlapping with each other.

Best practices

  • Compare total values across different categories over a time period.
  • Present positive, non-empty data points. For null or zero values, insert empty data points explicitly to represent them individually.

The following example showcases network utilization by subscribers in the telecommunications industry. With this, a telecom company sees its most preferred network category and can plan to increase its bandwidth to accommodate more users without compromising network quality.

Network utilization by subscribers—telecommunications industry
Telecommunications industry – network utilization by subscribers

Line chart

A line chart helps you analyze trends over a time period with data points connected using straight lines. It visualizes how the data changes over a time period using continuous data.

Best practices

  • Represent time-dependent data, showing data trends at equal intervals.
  • Showcase a large number of data points.
  • Avoid displaying too many lines, as it makes tracking complicated.
  • Use a different color for each category to identify them easily.

The following line chart example displays resolved and unresolved incident counts over a continuous time period at equal intervals from a customer support team. With this, an IT operations manager can easily assess the customer service performance.

Incident management—resolved and unresolved incidents
Incident management – resolved and unresolved incidents

Spline chart

A spline chart allows you to show trends for analysis over a time period with data points connected using splines.

Best practices

  • Showcase a large number of data points over a continuous period.
  • For multiple series representation, it can be used in combination with other basic chart types like column.

The following spline chart visualizes revenue comparison between months of 2016 and 2017. This lets sales teams know the ups and downs in various months of the current year and determine the influencing factors.

Revenue comparison— financial sales
Financial sales – revenue comparison

Combo chart

The combo chart is used to represent two different measurements having different ranges of values represented in two different axes.

Best practices

  • Use different chart types to represent two different data sets related to each other.
  • Use distinct colors for each chart type for better visualization.

The following example shows the product category’s sales volume and the gross profit value of a retail store. A retail sales team can easily visualize the profit percentage with respect to the increase or decrease in sales volume in each category.

Sales and gross profit by category—retail domain
Retail domain – sales and gross profit by category

The combination of line and column charts in the following example showcases a city’s weather details, such as temperature and precipitation, by month.

Average temperature and precipitation recorded in a city over a time period.
Average temperature and precipitation recorded in a city over a time period.

Conclusion

We hope this article helped you get a clear idea about chart types and their use cases in order to choose appropriate charts for your data. Use Charts and other widgets to create attractive and informative dashboards using Bold BI. If you have any questions about charts, please post them in the comments below. You can also contact us by submitting your questions through the Bold BI website or, if you already have an account, you can log in to submit your support question. Bold BI dashboards now come with a 15-day free trial with no credit card information required. Try out their data visualization capabilities for yourself!

Start your free 15-day trial today

Leave a Reply