Unveiling VizCritique Pro GPT

Adam Mico
Bootcamp
Published in
7 min readFeb 18, 2024

--

+ Celebrating My 100th Blog Post!

Credit | Adam Mico and OpenAI — Showcasing VizCritique Pro GPT

#DataFam and the greater data visualization community — today marks a monumental occasion in our journey together — this is my 100th #DataFam blog post on Medium! Over the years, we’ve explored countless topics on data visualization, shared insights, and grown together in a community passionate about the power of data. For this special centenary post, I’m thrilled to introduce a revolutionary tool that epitomizes the essence of our journey: VizCritique Pro GPT. This isn’t just a tool; it reflects our shared quest for excellence in data visualization. Please check it out here (free with ChatGPT+).

Background

The Inspiration Behind VizCritique Pro

Reflecting on the journey to this 100th post, I’ve sought topics and tools that inform, connect, and transform. The inspiration behind VizCritique Pro comes from this very aspiration — to empower each one of you to create data visualizations that are not just seen but felt, understood, and remembered. This tool culminates our collective learnings, distilled into a GPT that critiques, teaches, and inspires.

Why This GPT Marks a Milestone

It’s more than a “fun” tool. It’s about accessibility, ethical design, engagement, and learning. If you are reading this and do not have access to ChatGPT+ and want your viz reviewed by the GPT, please send me a screenshot of the visualization and indicate whether you want public or private feedback in YOUR language.

Credit | Adam Mico and Open AI — VizCritique Pro in action

Personalized Mentor for Your Data Visualization Needs

What sets VizCritique Pro apart is its ability to offer personalized, actionable feedback. It’s like having a mentor who understands your project’s context, audience, and goals. This personalized touch ensures that the feedback you receive isn’t just generic advice but a bespoke blueprint for enhancing your visualization. Please converse with the GPT if it needs your additional context.

Emphasizing Audience Understanding and Ethical Design

True to our shared values, this GPT stresses the importance of audience adaptation and ethical data representation. It’s a testament to our belief that data visualization should inform, engage, and respect its audience. This tool ensures your visualizations achieve that balance, making your work visually appealing, ethically sound, and universally accessible.

Celebrating 100 Posts of Innovation and Insight

Introducing VizCritique Pro in my 100th blog post is no coincidence. It represents what we’ve been working towards — innovation, insight, and impact in data. This tool is a step forward in our ongoing quest to elevate the standard of data visualization, making complex data not only more accessible but also more engaging. Now, let’s get to the good stuff: the GPT’s capabilities.

Capability Highlights of the VizCritique Pro

Credit | Adam Mico — VizCritique Pro Logo

There are numerous capabilities available with VizCritique Pro. In fact, I learned of the character limit of the GPT based on the comprehensive instructions provided. This tool is designed to provide you with personalized feedback with minimal effort from your end. Below is a non-exhaustive list of its capabilities.

  • Analysis and Feedback on Data Visualizations | Offers detailed critiques on chart types, labels, colors, layout, and other elements based on user descriptions.
  • Scoring System Based on Best Practices | Utilizes a scoring framework inspired by Tableau’s Blueprint visual best practice principles, providing actionable insights.
  • Feedback Across Multiple Domains | Evaluates visualizations on audience adaptation, message alignment, chart usage, layout/storytelling, color schemes, textual elements, and font readability.
  • Adaptation for Target Audience | Tailors feedback to match the complexity with the audience’s expertise level, ensuring clear and effective communication.
  • Strategic Recommendations | Advises on optimal chart usage, effective color schemes, and strategic layout for improved data storytelling and navigation.
  • Enhancement Suggestions | Offers specific suggestions to improve clarity, effectiveness, and engagement of data visualizations.
  • Inclusivity and Accessibility Guidance | Provides recommendations to make visualizations more accessible and engaging, focusing on universal design principles.
  • Professional and Positive Feedback Delivery | Ensures feedback is constructive, offering positive, actionable advice for improvement.
  • Up-to-date with Data Visualization Trends | Keeps knowledge current with the latest trends in data visualization for relevant and effective feedback.
  • Confidentiality in Submissions | Maintains user privacy and security in handling visualization submissions for critique.
  • User Experience Focus | Aims to enable users to understand the impact and usefulness of the feedback provided for their data visualization projects.

For my next step, I want to experiment with one of my visualizations to help you better understand the VizCritique Pro's output.

Use Case | Personal and Experimental Visualization

Credit | Adam Mico — A screenshot of my data visualization

Context Added

This is not a straightforward business dashboard, so context is constructive, as mentioned in the GPT description. However, the GPT is trained to adapt to your context (including audience, reasons for experimentation, and whether it’s an infographic vs. a dashboard tailored for interactivity and business use). Here is the brief context I gave with the image:

Please score this as an infographic. Although it is personal, it’s intended for public consumption. Each dot represents a month in my life, and life starts at birth, and I use my life expectancy to end the months. I used the curvy timeline of dots with intent. Also, the larger purple dot represents the current month.

Scoring (sample output from the DataViz Screenshot GPT):

This is the total output generated within 30–45 seconds (normally) after submitting a prompt. The output concludes immediately before the “Conclusion” section below.

Audience Adaptation

  • Score: 8.0/10

The infographic is tailored to a broad audience with its universal theme of life’s progression. Its simplicity makes it accessible to individuals of varying expertise levels.

Message Alignment

  • Score: 7.5/10

The message of life’s fleeting nature and the motivation to make the most of it is clear. However, using the timeline to reflect on past, present, and future could be made more explicit to reinforce the message.

Optimal Chart Usage

  • Score: 7.5/10

While the timeline is an engaging way to represent life’s journey, there might be other visual elements that could complement the dots to provide additional context or enhance the emotional impact.

Strategic Layout

  • Score: 8.0/10

The layout is clean and facilitates a natural flow from birth to life expectancy. The decision to use a curvy timeline adds an artistic touch that reflects the unpredictability of life.

Effective Color Schemes

  • Score: 8.5/10

The color choices are effective, with a soothing blue representing life and a bold yellow marking life expectancy. The purple dot for the current month stands out, although its significance could be made clearer.

Textual Elements

  • Score: 7.5/10

The title and explanation are thought-provoking, yet the text could be more prominent to immediately draw the viewer’s attention to the visualization’s purpose.

Font Readability

  • Score: 7.5/10 The font is readable, but increasing the size could improve visibility, particularly for viewers with visual impairments.

Accessibility

  • Score: 7.5/10 Accessibility seems considered, but enhancements such as a clearer legend or symbols might aid in interpretation for those with color vision deficiencies.

Overall Impressions (still part of GPT output)

The infographic is a poignant and personal visualization of a universal experience. It elegantly captures the essence of life’s passage through time with its simple yet profound design.

  • Final Score*: 7.9/10 (Visual Viz-ionary | An incredible score! The visualization is informative and emotionally engaging, demonstrating strong design skills and effectively representing a deeply personal yet universally relatable concept, with minor room for improvement.)

*Final score: Please note the score is strategically weighted so it does not score each section with the same weight as they do not carry the same impact. In addition, with over 60 tests I ran, most dashboards and visualizations are in the 5.5–8.0 range with large deltas between each 0.5 scoring incremement. For reference, a low intermediate score would be 5.5–5.9, intermediate would be 6.0–6.4, high intermediate would be 6.5–6.9, professional practitioners range between 7–7.9, and 8.0+ often requires the professional practitioners touch with a sensational use case.

Recommendations (still part of GPT output)

  • Consider adding a legend or a brief description to explain the color coding and the significance of the purple dot.
  • Increase the text size for the title and description to enhance readability.
  • Introduce interactive elements, such as tooltips or clickable dots, that could reveal more information about different life phases, for viewers who wish to explore the concept more deeply.

Conclusion

We’ve arrived at an exciting juncture where the tools at our disposal have transformed learning into an experience that’s not only immediate and enjoyable but also respects the privacy of feedback. I encourage you to seize these tools with vigor, broaden your reach, and embark on a path of continuous learning. My affinity for Tableau may shine through, as I’ve added instructions partially based on Tableau Blueprint’s reservoir of best practices. However, let’s not overlook the versatility of this tool; it is designed to complement a wide array of data visualization platforms, inviting both novices and veterans to craft compelling data narratives in THEIR language (see sample below).

Credit |Adam Mico and OpenAI — VizCritique Pro GPT using Japanese to communicate feedback.

Special Shoutout

After publishing my work, I realized that Jeffrey Shaffer, who is a Tableau Visionary Hall of Famer and has extensive knowledge of data visualization and data science, had already published a similar tool with a different experience. This tool is great as it provides expert feedback based on its extensive knowledge base. Please follow this link to check it out.

Adam Mico

Twitter | LinkedIn | Tableau Public | Data Doctor GPT (focusing on Tableau) | VizCritique Pro GPT

Note: My book, “Tableau Desktop Specialist Certification,” is available for order here.

--

--

Data Visualization and Enablement Leader | Data Leadership Collaborative Advisory Board Member | Tableau Visionary + Ambassador | Views are my own