7.3. Visualizing Distribution in Looker Studio
Visualizing the distribution of data can help us gain insights into various aspects of the data, such as understanding how different channels contribute to overall performance.
In this lesson, we will discuss different chart types for displaying distribution and their respective best practices.
Pie charts are generally considered the least effective way of displaying data distributions, especially when there are more than two or three categories.
However, when there are only two categories, such as new vs. returning visitors or male vs. female, pie charts can be acceptable. Ensure you use contrasting colors for different segments and display percentage labels to improve comprehension.
Stacked Bar Charts
Stacked bar charts, particularly 100% stacked bar charts, are a better alternative for showing distribution across multiple categories. These charts can be presented horizontally or vertically and allow for easier comparison of category contributions than pie charts.
Distribution Over Time
To visualize the changes in distribution over time, we can use time series charts, such as area charts. Two examples of this include:
- 100% Stacked Bar Chart by Days: By breaking down the distribution by day, week, or month, we can uncover trends and patterns that might not be visible in a single snapshot of the data. This chart type can help identify trends, such as different device usage on weekends versus weekdays.
- Stacked Area Chart: This chart type is a line chart with the area under each line shaded in a color. It can show how the distribution of a certain metric changes over time, revealing trends or significant events that affected the data distribution.
Visualizing distribution can provide valuable insights into your data. Choose the appropriate chart type based on the number of categories and the insights you want to convey. Pie charts can be suitable for two-category distributions, while stacked bar charts and area charts are more versatile and informative for displaying data across multiple categories and over time.