VIDEO
HIGHLIGHT
Log InSign up
Power BI visualization best practices by Marco Russo.
SummaryTranscriptChat

Power BI visualization best practices by Marco Russo.

Introduction and Background

In this section, Charles Stirling introduces himself and Marco Russo, who is calling in from Italy. They discuss the purpose of the session and encourage participants to ask questions in the chat window.

Introduction

  • Charles Stirling introduces himself and Marco Russo.
  • Marco Russo is calling in from Italy.
  • Participants are encouraged to ask questions in the chat window.

Training Overview

Marco Russo provides an overview of the training session on improving dashboards in Power BI. He explains that the content was designed by Daniele, who works with them. The goal is to provide best practices for creating effective and visually appealing dashboards.

Training Overview

  • The training focuses on improving dashboards in Power BI.
  • The content was designed by Daniele.
  • The goal is to provide best practices for creating effective and visually appealing dashboards.

Definition of a Dashboard

Marco Russo discusses the definition of a dashboard in Power BI. He explains that a dashboard is a visual display of important information consolidated into a single screen for quick monitoring and analysis.

Definition of a Dashboard

  • A dashboard is a visual display of important information.
  • It is consolidated into a single screen for quick monitoring and analysis.
  • The goal is to provide information at a glance without scrolling.

Characteristics of a Good Dashboard

Marco Russo discusses the characteristics of a good dashboard. He emphasizes the importance of utilizing space effectively, presenting clear numbers, using appropriate colors, and focusing on user goals.

Characteristics of a Good Dashboard

  • Utilize space effectively to present clear numbers.
  • Use appropriate colors that convey meaningful information.
  • Focus on user goals and provide quick analysis of data.
  • Avoid clutter and unnecessary elements.

Mistakes in Dashboard Design

Marco Russo discusses common mistakes in dashboard design. He presents an example of a poorly designed dashboard that lacks effective use of space, clear numbers, and appropriate color choices.

Mistakes in Dashboard Design

  • Poor use of available space.
  • Small and confusing numbers.
  • Inappropriate color choices.
  • Lack of focus on user goals.

Conclusion

Marco Russo concludes the session by highlighting the importance of creating visually appealing and effective dashboards. He encourages participants to apply best practices to improve their own dashboards.

Conclusion

  • Creating visually appealing and effective dashboards is crucial.
  • Applying best practices can help improve dashboards.
  • Encouragement to apply the concepts learned in the session.

Dashboard Design Best Practices

In this section, the speaker discusses the flaws in a dashboard design and highlights the importance of following best practices for creating effective dashboards.

Common Mistakes in Dashboard Design

  • The first dashboard shown violates several design rules out of the 15 defined rules for creating good dashboards.
  • One of the violated rules is keeping everything at a glance, which means avoiding the need to scroll within a dashboard.
  • Another mistake is not keeping the elements aligned properly, which affects the overall aesthetics and message conveyed by the dashboard.
  • The use of symbols without clear explanations or legends can be confusing, especially when presenting to stakeholders who may not be familiar with business acronyms.
  • Some gauges on the dashboard appear similar but have different values, making it difficult to compare them side by side.
  • Including excessive detail in metrics can make them less useful. Simplifying numbers and using appropriate units can enhance understanding.
  • Choosing colors carefully is crucial as some people may have colorblindness or perceive colors differently. Ensuring color choices are distinguishable is important for conveying information effectively.
  • Selecting appropriate chart types is essential. For example, using line charts instead of bar charts for showing trends over time can improve visualization.

Principles for Designing Effective Dashboards

  • The speaker introduces a set of 15 principles that should be followed when designing dashboards. These principles aim to create visually appealing and informative dashboards that convey data effectively.

Timestamps are provided where available to help locate specific points in the video transcript.

Principles of Effective Dashboard Design

In this section, the speaker discusses key principles for creating effective dashboards.

Keep it Simple and Avoid Distractions

  • Simplify your dashboard design to avoid clutter and unnecessary decorations.
  • Avoid using background images or black backgrounds that do not add value to the information displayed.

Align Elements and Be Consistent

  • Arrange charts and text in a logical order for better readability.
  • Use consistent colors and chart types across the dashboard, unless there is a valid reason to differentiate them.

Highlight Relevant Information

  • Emphasize important data by giving them more prominence on the screen.
  • Positioning matters, as humans tend to assign more importance to information placed in prominent areas of the dashboard.

Be Clear and Provide Context

  • Clearly explain what is being shown on the dashboard.
  • Avoid using acronyms without providing explanations or legends when necessary.
  • Display data in context by comparing it with relevant benchmarks or previous periods.

Choose Appropriate Visualization Tools

  • Select the right visualization tools that effectively convey the intended message.

Axis Start from Zero, Shorten Numbers, Show Variations

This section covers additional principles for effective dashboard design.

Start Axis from Zero (for some charts)

  • Some charts require starting their axis from zero, while others may not have this requirement. Consider the nature of your data when deciding whether to start from zero or not.

Shorten Numbers for Better Readability

  • Reduce the number of digits displayed in metrics to improve readability. For example, representing one million as "1M" can be more effective than writing out all six zeros.

Show Variations Appropriately

  • When highlighting changes over time, choose appropriate ways to display variations.
  • Differentiate between significant differences and minor fluctuations based on how you measure and present the data.

Avoid Noise and Suggestive Relationships

  • Avoid suggesting relationships between unrelated metrics by providing adequate spacing or separation.
  • Be mindful of how the proximity of numbers or charts can imply connections that do not exist.

Use the Right Visualization

  • Select the appropriate visualization type to effectively communicate the desired information to users.

The remaining sections of the transcript will be summarized in subsequent chapters.

Dashboard Comparison: Left vs Right

In this section, the speaker discusses two different dashboards and compares their effectiveness in presenting data to users. The left dashboard is preferred by many investors due to its visual appeal, while the right dashboard has been improved based on certain design rules.

Left Dashboard

  • Investors prefer the left dashboard due to its visual appeal.
  • It provides a comprehensive view of data that can be easily accessed throughout the day.

Right Dashboard

  • The right dashboard has been improved based on design rules.
  • It eliminates unnecessary elements such as borders, gridlines, and images for simplicity.
  • A white background is preferred for better printing results.
  • Avoid excessive use of typography and stick to a simple font style.

Issues with the Real Dashboard

In this section, the speaker encounters issues with a real dashboard that violates design rules. The dashboard requires scrolling and does not provide an optimal user experience.

  • The real dashboard violates design rules and requires scrolling.
  • Scrolling causes delays in screen updates during streaming sessions.
  • Clicking on different visualizations changes the entire display.
  • Enlarging everything reduces screen space and wastes valuable real estate.

Improved Dashboard Design

In this section, the speaker showcases an improved version of the previous dashboard that adheres to design rules. This version provides more information in a readable format.

  • The improved dashboard offers more numbers, information, and data visibility.
  • Unnecessary elements have been removed for better readability and usability.
  • Users can clearly see all changes when interacting with the dashboard.

Rule 1: Keep it Simple

This section focuses on the first rule of effective dashboard design - simplicity. The speaker emphasizes removing visual effects, borders, gridlines, and excessive typography.

  • Simplicity is the foundation of effective dashboard design.
  • Remove visual effects, borders, gridlines, and unnecessary images.
  • Use a white background for better printing results.
  • Avoid excessive use of different fonts and emphasize simplicity.

Rule 2: Avoid Unnecessary Additions

This section discusses the second rule of effective dashboard design - avoiding unnecessary additions. The speaker highlights the importance of minimizing graphic elements, larger fonts, and logos that do not add value to the information presented.

  • Minimize graphic elements that do not contribute to data understanding.
  • Larger fonts should be used sparingly for emphasis.
  • Logos should only be included if they serve a purpose in the dashboard.

Rule 3: Emphasize with Exception

This section covers the third rule of effective dashboard design - emphasizing with exception. The speaker explains when and how to use bold font styles to draw attention to specific elements in a dashboard.

  • Emphasizing with bold font styles should be an exception rather than a norm.
  • Limit the use of bold font styles to one or two instances in a dashboard.
  • Zero emphasis is often better than excessive emphasis.

Conclusion

In this final section, the speaker concludes by highlighting the importance of following design rules for creating effective dashboards. The goal is to provide valuable information in an easily accessible format for business users.

  • Effective dashboards prioritize simplicity and remove unnecessary elements.
  • Improved designs offer more data visibility and readability.
  • Users can better appreciate changes when interacting with well-designed dashboards.

New Section

In this section, the speaker discusses the issues with the design of a dashboard and how it affects readability and usability.

Problems with the Design

  • The picture in the background takes up unnecessary space and is distracting. Timestamp: 0:28:22
  • Due to wasted space, a smaller font had to be used for other information on the dashboard. Timestamp: 0:28:45
  • Unnecessary borders around numbers and bar charts make them crowded and difficult to read. Timestamp: 0:29:12
  • Two numbers in the bar chart are not helpful for understanding trends and should be displayed separately. Timestamp: 0:29:32

Improving Readability

  • By removing pointless graphics, more space is created, allowing for better separation of visuals and more data to be included in the report. Timestamp: 0:30:00
  • A clean design can result in a more readable and visually appealing dashboard. Timestamp: 0:30:52
  • Grouping related charts together while keeping unrelated parts separated improves organization. Timestamp: 0:31:43
  • Using clear space instead of lines helps create separation between different elements on the dashboard. Timestamp: 0:32:05

New Section

In this section, the speaker highlights some examples of dashboards that may look visually appealing but have usability issues.

Usability Issues with Dashboards

  • Overlapping areas in charts make them difficult to read and understand. Timestamp: 0:32:32
  • Inconsistent color choices can lead to confusion and misinterpretation of data. Timestamp: 0:33:15

New Section

In this section, the speaker continues discussing usability issues with dashboards.

More Usability Issues

  • Overlapping areas and transparency in charts make it hard to understand the meaning of data. Timestamp: 0:33:37
  • Poor color choices can create confusion and misinterpretation of metrics. Timestamp: 0:34:19

Note that these summaries are based on the provided transcript and may not cover all the details or context from the video.

Dashboard Design Issues

In this section, the speaker discusses various design issues with dashboards and suggests improvements.

Ambiguity in Numbers and Lack of Reference

  • The numbers displayed on the dashboard are unclear and do not indicate whether they are cumulative or ratios against something else.
  • There is a lack of reference to understand the percentage values shown. It is important to know what these percentages represent.

Unnecessary Decoration and Hiding Information

  • The background image and unnecessary decoration can be distracting and add no value to the dashboard.
  • Important information is hidden under certain areas, making it difficult to interpret the data accurately.
  • The numbers displayed on the dashboard are not easily readable due to the absence of separators or appropriate formatting.
  • The meaning of colors used in the dashboard is unclear, making it difficult to understand their significance.

Improved Dashboard Design

  • A better-designed dashboard with a lighter background has fewer issues compared to previous examples.
  • However, there are still some challenges in comparing numbers provided in tables and gauges, which could be improved with better visualization techniques such as trend lines.

Dark Background vs White Background

This section explores the preference between dark backgrounds and white backgrounds for dashboards.

Preference for White Background

  • While some people may prefer dark backgrounds, a white background is generally easier to print and avoids certain trade-offs associated with using a black background.

Chart Reference Categories

This section introduces categories for classifying different types of charts used in Power BI and other visualization tools.

Categories for Chart Classification

  • Eight categories are defined for classifying visuals: comparison, change over time, whole flow, ranking, spatial distribution, correlation, and others.
  • Having too many categories can lead to confusion, so it is important to choose the most relevant category for a particular metric.
  • Sometimes visuals may belong to multiple categories, in which case it is recommended to create separate charts for each category.

The transcript provided does not cover the entire video.

New Section

In this section, the speaker discusses the process of identifying categories and choosing visualizations for a dashboard.

Identifying Categories and Choosing Visualizations

  • The first step is to determine the percentage of each category being sold compared to the entire volume. This helps in easily identifying the category.
  • Once the category is identified, the next step is to choose a chart or visualization for that particular category.
  • The speaker provides a PDF visual reference that can be downloaded for free. It contains high-resolution vector format charts that can be zoomed in, zoomed out, and printed with good quality.
  • The PDF is regularly updated and available at a specific URL provided by the speaker.

New Section

In this section, the speaker addresses a question about examples shown in Power BI and other tools.

Examples in Power BI

  • The examples shown during the presentation were found on the internet and may have been created using various tools.
  • However, all the examples used to demonstrate applying rules are created in Power BI.
  • The speaker showcases a Power BI dashboard as an example, highlighting its functionality and real-time data updates.

New Section

In this section, the speaker explains how they obtained real examples from public websites to showcase potential mistakes in dashboard design.

Real Examples vs. Bad Examples

  • The bad examples shown during the presentation were not created using Power BI but were actual dashboards found on public websites.
  • These dashboards were used to demonstrate common mistakes or areas for improvement in dashboard design.
  • The speaker emphasizes that they did not steal any content but rather used publicly available dashboards as illustrations.

New Section

In this section, the speaker provides a downloadable PDF visual reference for choosing chart types and explains its contents.

Downloadable PDF Visual Reference

  • The speaker shares a URL where the audience can download a PDF visual reference.
  • The PDF contains various chart types categorized by their suitability, with explanations for each category.
  • Charts with yellow backgrounds are preferred choices, those with white backgrounds are considered good but not the best, and those with gray backgrounds are not recommended.
  • The speaker suggests that studying this visual reference can help in making informed decisions when selecting chart types.

New Section

In this section, the speaker addresses a question about when to split a dashboard into separate ones due to information overload.

Breaking Up Dashboards

  • When a dashboard becomes too crowded or difficult to follow due to excessive information, it may be necessary to consider splitting it into separate dashboards.
  • The speaker suggests that each dashboard should focus on one area of the company or specific metrics.
  • However, it is important to note that high-level executives may still want to see an overview of everything but at a top-level summary rather than detailed information.
  • If drilling down into specific areas or metrics is required, creating additional dashboards can be beneficial.

New Section

The importance of dashboards and the different metrics that can be included for CFOs and CEOs.

Dashboard Metrics for CFOs and CEOs

  • Dashboards can show the top 10 customers, cash flow trends, credit trends, and cash flow forecasts. These are important metrics for CFOs.
  • CEOs may also be interested in four or five additional metrics from different areas of the company.
  • The level of detail needed in a dashboard depends on its purpose. Higher-level data can include information from various parts of the company, while more detailed dashboards focus on specific departments.
  • It is possible to provide more metrics about a specific department without overwhelming the user with excessive information.

New Section

Discussion about using dynamic dependent validation lists to choose visualizations in Power BI.

Using Dynamic Dependent Validation Lists for Visualizations

  • The idea of using dynamic dependent validation lists to choose visualizations is suggested.
  • In Power BI, there could be a wizard or feature that helps users select visuals based on their data.
  • However, it is debated whether allowing users to dynamically change visualizations within a dashboard is a good idea.
  • Dashboards should be consumed as they are designed, providing an overview rather than advanced analytics capabilities.
  • Power BI offers many features beyond designing dashboards, such as creating detailed reports for data exploration.

New Section

Differentiating between dashboards and reports in Power BI and understanding their goals.

Dashboards vs Reports

  • A dashboard serves as an overview tool for multiple users to get a big picture of the company's performance.
  • On the other hand, reports allow deeper exploration and analysis of data with more advanced analytics capabilities.
  • The ratio of users between dashboards and reports is typically 10 to 1 or even 100 to 1.
  • While dashboards are consumed by many users, reports are used by a smaller number of users who require more detailed insights.
  • It is important to understand the goals and purposes of dashboards and reports when designing them in Power BI.

New Section

Design considerations for separating visuals in a dashboard using backgrounds.

Separating Visuals with Backgrounds

  • The suggestion is made to use slight backgrounds behind visuals to enhance separation in a dashboard design.
  • Custom visuals were created because the default visualizations in Power BI did not provide the desired level of separation.
  • Using backgrounds as a way to create reference points or achieve specific visualization features can be considered as a workaround.
  • However, it is acknowledged that this approach is due to current limitations of the product and may not be the ideal solution.
  • The design patterns used in dashboards should prioritize effective separation and clarity without relying solely on background usage.

Designing a Background for the Dashboard

The speaker discusses their approach to designing a background for the dashboard and explains why they choose not to spend time aligning visuals with the background due to potential changes in future versions of Power BI.

Design Considerations

  • The speaker believes that workarounds for aligning visuals with the background would be time-consuming and unnecessary.
  • They mention that small changes in new versions of Power BI can sometimes disrupt alignment, making it impractical to spend time on precise alignment.

Avoiding Pitfalls in Dashboard Design

The speaker mentions that they have strategies to avoid common pitfalls in dashboard design and suggests watching their webinar for more details. They also address questions about accessing slides from the presentation.

Strategies to Avoid Pitfalls

  • The speaker mentions cheating techniques they use to avoid common pitfalls in dashboard design.
  • They recommend watching their webinar for more information on these strategies.

Accessing Slides

  • Viewers can access the slides from the presentation through a provided link.
  • The slides are available as a PDF document, which will be shared after the session.

Adding Content and Links

The speaker suggests adding content or links to make it easier for people visiting the page later. They discuss providing a link or additional content in the comments section.

Providing Links or Content

  • The speaker suggests adding a link or relevant content to make it easier for viewers who visit the page later.
  • They propose adding this information either as a comment or within the description of the video.

Sharing Slides and Closing Remarks

The speaker confirms that they will provide a link to the PDF version of the slides and discusses closing remarks for the webinar.

Sharing Slides

  • The speaker will share a PDF version of the slides and provide a link for viewers to access them.
  • They mention that they will fix any issues with sharing the slides after the session.

Closing Remarks

  • The host thanks the speaker for conducting the webinar and expresses excitement about their upcoming session.
  • The speaker appreciates the opportunity to present their training course and invites viewers to provide feedback on their design principles.

Previewing Power BI Dashboard Design Course

The speaker provides an overview of their Power BI dashboard design course, including a short URL where viewers can access more information and download a PDF. They encourage feedback on their design principles.

Course Overview

  • The speaker introduces their Power BI dashboard design course, which was launched recently.
  • Viewers can find more information about the course at a provided short URL.
  • They invite viewers to download the accompanying PDF and share feedback on their design principles.
Video description

What is a dashboard? It is, or it should be, a visual display of the most important information needed to achieve one or more objectives, consolidated and arranged in a single screen. Does this definition corresponds to a Power BI dashboard? Not necessarily, also a report can be a dashboard in this definition. Regardless of its technical implementation, many dashboards fail to reach their goals and are merely decorative, even if they end up being ugly. And this latter issue, alone, is a good reason why a dashboard could be ignored. This session introduces the principles to design useful and good looking dashboard in a page of a Power BI report. You will see some of the common issues and the possible solutions, with an overview of the main rules that you should apply to create a useful, consistent, and nice-looking dashboard. The link Marco referenced was: http://sql.bi/dashboard http://sql.bi/visual-reference (Today they are the same link but in two weeks they will point to different content)