1. Utilizing Complex Graphs That Stakeholders May Misinterpret

The interns often tried to utilize every available chart type in Power BI, which led to confusion. Their preference for intricate plots over straightforward bar charts highlighted two key issues:

• They failed to consider their audience, who were the imaginary board members.
• They prioritized showcasing their technical abilities rather than addressing business inquiries.

Always remember to tailor your visualizations to your audience. In this case, CEOs and board members usually have limited time and technical expertise. Simple charts like bar graphs or KPI cards are more effective for presenting insights to them.

2. Overloading a Single Page with Visuals

While some interns created insightful metrics and visuals, they cluttered them all onto one page. This made it difficult to:

• Grasp the narrative they aimed to convey.
• Focus on key trends and insights.
• Read the visuals clearly due to the sheer number of elements.

A crowded dashboard can overwhelm stakeholders and lessen the impact of individual insights. As a guideline, try to limit yourself to four visuals per page (excluding KPI cards). If you find it necessary to include more, consider adding an additional page.

3. Omitting Slicers When Needed

I observed that a couple of interns didn’t use any slicers throughout their reports. While not every page requires a slicer, they are crucial in many cases.

When dealing with extensive related datasets, slicers help stakeholders hone in on specific data segments effortlessly. For example, if you have sales data from various regions, slicers allow users to view sales from one region at a time.

Think of your Medium stats dashboard; the dropdown for monthly views is a slicer. Slicers are just as vital as the charts and metrics in your reports.

4. Including Visuals That Don’t Address Business Questions

Typically, business questions guide visualization, but in this instance, the interns were encouraged to formulate their own inquiries. Some visuals fell short in this regard:

• Certain visuals addressed broad business questions but lacked relevance to the specific business context.
• Some insights were applicable to the marketing or sales teams but not to the board members.
• Others lacked trends or patterns, such as displaying nearly equal ratios of homeowners versus non-homeowners using a pie chart, which would have been more effective if the ratio showed a clearer distinction.

Always ensure that each visual serves a purpose and relates to a pertinent business question.

5. Including Unnecessary Data or Failing to Remove Redundant Columns

While this issue was minimal due to the small dataset, it’s a poor practice at the industry level. Including irrelevant data increases file sizes and processing time, which can become problematic as data volumes rise.

Aim to eliminate unnecessary columns and avoid importing any data from the database just because it’s accessible. Your Power BI reports should reflect a systematic approach.

6. Rushing the Pre-Visualization Process

This error was evident when certain plots didn’t update with slicers, or when currency symbols were missing from KPI cards. Understanding that Power BI involves more than just placing visuals is crucial.

Take the time to thoroughly analyze your data through profiling, addressing missing values, identifying outliers, and ensuring proper relationships and schemas. This diligence will enhance the quality and robustness of your reports as new data comes in.

7. Employing Excessive Colors

This problem stemmed from improper legend use and a lack of cohesive color themes. Remember, in data visualization, less is often more.

Best practices for color use include:

• Use color sparingly to highlight key information.
• Choose colors that harmonize well together.
• Ensure high contrast ratios for accessibility.
• Use colors with representative meanings (e.g., red for decline, green for growth).
• Maintain color consistency across visuals.

8. Misusing Data Labels

I observed several errors related to data labels across different plot types. Here are some best practices:

• Bar Chart: If you’re including data labels on bars, remove the y-axis values.
• Line Chart: For wavy lines with many peaks and troughs, consider excluding data labels.
• Combo Chart: Avoid data labels, as it can be unclear which label belongs to which plot. Rely on multiple y-axes for clarity.

9. Neglecting to Adjust Y-Axis Values

Often, we overlook extreme values in line and area charts because the graphs didn’t account for the full range of the y-axis. This oversight can lead to misleading interpretations. Always verify and adjust the maximum and minimum y-axis values accordingly.

10. Choosing Inappropriate Plots for Specific Data

This was the most common mistake among the interns. General guidelines include:

• Avoid pie/donut charts for numerous categories or nearly identical ratios.
• Use line plots for time series analysis.
• Opt for the right bar plot variations to depict category counts effectively.
• Use scatter plots to visualize relationships between continuous variables only when a correlation exists.

If you utilize other plot types, ensure they are appropriate and convey the intended insights clearly.

> A simple rule for creating visualizations and dashboards: if you find yourself needing extensive explanations, reconsider their inclusion.

Visuals should be self-explanatory, telling their story without excessive text.

Call to Action

That wraps up this overview! I hope you found these insights valuable and feel more prepared to tackle your next data visualization project. If you found this helpful, feel free to show your support with claps so other data professionals can see it.

This article isn’t eligible for earnings on Medium due to my location, but if you appreciate my work, you can tip me using the button below.

Goodbye for now!

Visit us at DataDrivenInvestor.com

Subscribe to DDIntel here.

Have a unique story to share? Submit to DDIntel here.

Join our creator ecosystem here.

DDIntel highlights notable content from our main site and our popular DDI Medium publication. Explore for more insights from our community.

DDI Official Telegram Channel: https://t.me/+tafUp6ecEys4YjQ1

Follow us on LinkedIn, Twitter, YouTube, and Facebook.