Visualization best practices

Effective reports are much more than simply answering documented business questions with the available measures and columns of the dataset. Reports also need to be visually appealing and provide a logical structure that aids in navigation and readability. Business users of all backgrounds appreciate a report that is clear, concise, and aesthetically pleasing.

Now that the report-planning phase described earlier is complete, the following list of 15 visualization practices can guide the report development process: 

  1. Avoid clutter and minimize nonessential details:
    • Each visual should align with the purpose of the report—to gain insight into a business question:
      • Visualizations should not represent wild guesses or functionality that the author finds interesting
    • Eliminate report elements that aren't essential for gaining understanding:
      • Gridlines, legends, axis labels, text boxes, and images can often be limited or removed
    • The report should be understandable at a glance, without supporting documentation or explanation.
A simple but helpful test is to view a Power BI report on a laptop screen from a distance of 12 to 15 feet, such as from the opposite end of a conference room. At this distance, it will be impossible to read any small text and only the shapes, curves, and colors will be useful for deriving meaning. If the report is still meaningful, this suggests the report is effectively designed visually.
  1. Provide simple, clear titles on report pages and visuals:
    • Text boxes can be used to name or describe the report, report page, and provide the last-refreshed date
  2. For chart visuals, use the length of lines and the two-dimensional position of points to aid visual comprehension:
    • In line charts, users can easily perceive trends and the divergence of lines relative to each other
    • In column or bar charts, users can easily distinguish relative differences in the length of bars
    • In scatter charts, users can quickly interpret the two-dimensional position of data points relative to each other
The purpose of these two attributes (line length, 2-D position) as the primary communication mechanism is to guide the user to an accurate assessment with minimal effort. Other visual attributes such as color, shape, and size can also be beneficial, particularly when these properties are driven by the data, such as with conditional formatting and KPIs. However, line length and 2-D position (X, Y coordinates) have a natural advantage in visual perception. For example, the differences between three items on a clustered column chart is much more obvious than the same three items presented on a pie chart.
  1. Position and group visuals to provide a logical navigation across the canvas:
    • The most important visuals should be positioned in the top-left corner of each report page
    • If multiple visuals are closely related, consider grouping them within a shape object
  2. Use soft, natural colors for most visuals:
    • Avoid overwhelming users with highly saturated bright or dark colors
    • Only use more pronounced colors when it's necessary to make an item stand out, such as conditional formatting
  1. Only apply distinct colors to items in chart visuals when the colors convey meaning:
    • For example, three colors might be useful for the data points of three separate product categories
  2. Align visuals to common and X and Y pixel positions:
    • For example, if a visual in the top-left corner of a page has X and Y position values of 20 and 40, respectively, then other visuals on the left side of the canvas should also have an X Position of 20
    • Likewise, the top visual(s) on the right side of the canvas should align with the left visuals at a Y position of 40 
  3. Distribute visuals vertically and horizontally to create an equal amount of space between visuals:
    • The amount of spacing should be adequate to clearly distinguish the visuals as separate entities

With one or multiple visuals selected in Power BI Desktop, a Format tab will appear on the ribbon per the following image:

Alignment, distribution, and Z-order format options
The three format options (Align, Distribute, and Bring forward and Send backward (Z-order)) are consistent with common MS Office applications, such as Excel and PowerPoint. Between these formatting options and the four properties available under the general formatting card for all visuals (X Position, Y Position, Width, and Height). Report authors can ensure that visuals are properly aligned and spaced. The Show gridlines and Snap objects to grid options under the View tab also support alignment. 
  1. Choose a page background color that will naturally contrast with visuals, such as the default white or a very light gray.

 

  1. For column and bar charts, sort visuals by their measure to provide an implicit ranking by the given measure:
    • This sorting is only applicable to nominal categories, such as product categories, when the individual items in the category don't need to follow a custom sort order
  2. Fill the available report canvas space; avoid large blank spaces in report pages.
  3. Provide supporting context via tooltips and additional lines in charts, such as target values and the min, max, and average:
    • Several measures related to a given visual can be displayed via tooltips without incurring performance penalties
    • The Power BI Analytics pane provides several support lines, including a trend line and a predictive forecast line
  4. All report pages should follow a common design theme and color palette:
    • Preferably all reports in a project and even for an organization should follow the same basic design guidelines

A Switch Theme icon on the Home tab of Power BI Desktop in Report View exposes options for importing a report theme and thus overriding the default color and formatting properties:

Import report theme
Custom Report Themes are a preview feature as of the November 2017 release of Power BI Desktop and allow organizations to apply a custom set of formatting properties to Power BI reports. For example, an organization can embed its corporate colors into a report theme (a JSON file) to apply this set of colors to all Power BI reports. Additionally, more elaborate formatting properties can be specified in report themes to standardize report development, such as the font family and font sizes. Existing report themes are available for download from the Power BI Report Theme Gallery (http://bit.ly/2pyUKpl). Additionally, tools and documentation are available for easily generating report themes, such as the Power BI Tips Color Theme Generator (https://powerbi.tips/tools/color-theme-generator/). 
  1. The quantitative scale for column and bar charts must start at zero:
    • Custom quantitative scales, such as from 12% to 15%, can be applied to line, scatter, and bubble charts to emphasize specific ranges of values
Consider two items, Product A and Product B, of a clustered column chart with margin percentage values of 32% and 34%, respectively. With a base of zero, the two items would correctly appear similar for the given measure. However, if the base value of the visual starts at 31% and the max value of the scale is set to 35%, Product B would visually appear as a dramatically higher value. This distortion is the reason that quantitative scales for column and bar charts must start at zero.
  1. Lines should only be used to connect interval scale data points, such as time series and ordered bins of values:
    • A line should not, for example, represent the sales for different product categories
    • A line should, however, represent the sales of products by unit price bins (for example, $0 to $10, $10 to $20, and so forth) 
..................Content has been hidden....................

You can't read the all page of ebook, please click here login for view all page.
Reset
3.21.43.26