Tips for building a killer summary screen in Power BI

As Bernard Marr has quoted this great article for Forbes, we suffer from data overload. Information is everywhere in many different forms, and this information urgently needs to be turned into information. Data is a key factor in decision making because it is the most valuable piece of data.

The concept of data is objective, readable, easy to understand, and in this article I will provide some tips for sharing the best part of the data and converting the useful information in the PowerBI dashboard.

In the data vpersonalization tools such as Power BI, the processed data is displayed in reports or summary screens. Many areas are considered when developing a good report or panel, such as programming, design, user interface / interface, neuroscience, language. In order to guarantee the effectiveness of the information, it is necessary to establish a connection with the user, and thus the concept of a report.

If this is your first connection to PowerBI or you have no experience in this field, I recommend that you read this article, which explains in detail what PowerBI is and how to start it:

People have an innate ability to tell stories, we tell stories all the time from our childhood. Storytelling is a tool to connect with the reader and deliver information as efficiently as possible.

To create a good storytelling, we need to guarantee some basic elements, such as:

  • Beauty and design models in visualization – people are very dependent on patterns and perfect shapes. If you use some design model concepts such as Golden ratio and some color patterns suddenly enhance the beauty of the dashboard.
  • Simplicity – Sometimes less is more, limit the number of dashboard charts to present the main ones. Don’t use complicated wallpapers and a white background instead could be a good start for simplicity.
  • Ease of reading and comprehension – “A complete diagram is one that does not need to be explained further”. The chart must be very clear and easy to understand at the first inspection of each key user, if you have to spend time explaining the chart, there is a good chance that it is not a good chart.
  • Using the same language as the target audience – Sometimes you present a dashboard to a group of university students and sometimes a dashboard to the CEO of an international company, and the language they use is completely different. So paying attention to your target audience and adapting your approach to each scenario is very important.
  • Data reliability – The information displayed on the summary screen must be reliable. The data quality component is very important for each panel, check each number, KPI, filter interaction, charts that vary over time, and try to compare the figures with other company sources. The wrong number can lead to huge mistakes in the decision-making process and weaken the dashboard.

Over time and with many studies, we gained experience creating dashboards in an increasingly natural way and with good storytelling.

The question remains, how can I learn to build a good dashboard and achieve good storytelling with less experience? How can I develop my data visualization skills without reading hours of content?
In response to these questions, here are some tips to help you assemble the perfect dashboard for every situation.

It is very important to know which type of chart is best for each type of view. The wrong choice can increase the difficulty of analysis and comprehension comprehension. The purpose of graphical visualization is to present information in the simplest and most intuitive way possible.

“A complete diagram is one that doesn’t need to be explained further.”

Dr. Andrew Abela published a nice chart that helped us decide which chart best fits each scenario. Click the link below to see the full chart:

https://apandre.wordpress.com/dataviews/choiceofchart/

Dr. Andrew Abela states that:

  • Recipient To compare data we should use bar charts if we want to compare data between items

OR use a line chart to compare values ​​over time

  • Show Ratio A scatter plot for the two variables and a scatter plot that varies in point size for the three variables must be used between the values.
  • Analyze Dissemination of information, we should mainly use a histogram plot for one variable or a scatter plot for two variables.
  • Understand Composition for data, we need to use a bar chart or area chart if the variables change over time

OR a pizza chart or a waterfall chart if the variables are static.

Colors are also extremely important in your display panel. Poor color choices can make the dashboard difficult to read or visually uncomfortable. Avoid using too many colors on the summary screen if you plan to view from a distance like a PowerPoint presentation, choose colors with high contrast to improve visualization, look at the dashboard from afar and up close, show your dashboard to someone else, and look at his or her feedback.
Below is basic information on how to use a color ring to assemble combinations.

There are several themes in Power BI with predefined color palettes, but you can also find them on the Internet, otherwise I left some nice ones in the link below:

The insights can be divided into three different groups: Business visibility, performance improvement and opportunity opportunities.

These three different types of views have different views and are optimized to improve understanding in each case, and I will explain them better below:

1. Company visibility

The most common way to present the Business Visibility type of insight is a REPORT.
Reports are one-time reports designed to show answers to a specific question about a company. Reports are usually distributed via email or presented in PowerPoint.
The report is a model designed to give an idea of ​​a simple measure of the facts that have happened over time in a company. A simpler view where it does not go beyond the insights presented in the chart.

2. Improving performance

A device that is generally associated with performance improvement is SCOREBOARD.
In insights into the type of performance improvement, the charts are already more detailed and include some metrics that present data that brings values ​​to the target audience, such as objectives, average, total, maximum and minimum values, so that the team picks up the maximum value in a more practical way. This type of chart uses a lot of KPIs (indicators), calculated values ​​and statistics.

3. Finding opportunities

The device that is usually associated with Opportunity Discovery information is Dashboard.
Dashboards are information that is presented in many ways in a complex yet easy-to-understand way to the end user. Aggregate displays show data that has already been processed, including various KPIs, correlations, forecasts, data evolution over time, metrics. Normally, this type of visualization is used by analysts, managers, investors, coordinators, people who need a huge amount of information to make a decision.

Some general visualization tips for creating a complete dashboard or report.

  1. Remove the edges – Try to present your graphics in a natural way, fit it into the background of your report and interact with it, and be an environment with other graphics, don’t delete it with a border.
  2. Embrace white space – Avoid clutter too much on your dashboard, often a small space is good for overall harmony.
  3. Collapse text – Try to wrap text in graphics in the form of a caption or title. Cut a few paragraphs into short, easy-to-understand sentences. “A picture is worth a thousand words and a graphic is worth a thousand pictures”

Finally, I proposed converting the report by applying the tips in this article.
That report provides information on beer consumption in the city of São Paulo in Brazil. The first report below presents a long text with several charts and indicators of consumption. Colors are standard, text is too large, graphics are not standard, a lot of information is difficult to understand. Anyway, take a lot of time to understand everything the report is going through.

LEAVE A REPLY

Please enter your comment!
Please enter your name here