The Final Data Story

Changes Made Since Part II

Since completing Part II, I made several key changes to improve the flow, design, and overall storytelling of my data story:

Removed Chapters for Better Flow: Based on feedback, I removed the rigid chapter structure to create a smoother, continuous narrative. This made the story feel less like a report and more like an engaging, data-driven journey.
Strengthened the Call to Action (CTA): I revised the final CTA to make it more direct, actionable, and empowering. The updated CTA focuses on encouraging readers to reflect on their energy usage while highlighting practical steps they can take.
Improved Formatting and Design:
- I simplified the text formatting by reducing dense paragraphs, increasing white space, and using callouts for key data points.
- I adjusted background colors and contrasts to improve readability and reduce visual strain, especially in areas where users noted issues.
- Visual elements were aligned better with the text, using a side-by-side layout to make the story more cohesive and easier to follow.
Refined Data Visualizations: I made visualizations simpler and clearer, with enhanced annotations and proper citations to strengthen credibility.

These changes aimed to make the data story more engaging, reader-friendly, and aligned with audience expectations.

The Audience

The target audience for my final data story primarily includes:

General Public: I aimed to make the story accessible to anyone, regardless of their technical background. This focus was driven by interview insights, where several users highlighted the importance of simple, clear explanations without technical jargon.
Industry Leaders & Policymakers: The story is also intended for decision-makers in tech and energy sectors who are responsible for developing sustainable AI solutions. Specific sections on local grid strain and renewable energy investments were added to highlight policy-relevant insights.
Adjustments for the Audience:
- I used relatable analogies (e.g., comparing ChatGPT queries to household energy use) to make complex data more digestible.
- I ensured that the call to action offered realistic steps that both the general public and policymakers could take.
- Feedback from user interviews influenced my decision to highlight local impacts of AI energy consumption, not just global statistics.

Final Design Decisions

Color Palette & Visual Style: I chose cool blues and greys to reflect energy themes, with orange accents to highlight key data points. Dark mode was used throughout to align with the slide themes and improve readability.
Typography: I stuck with Poppins, Shorthand’s default font, for its clean and modern aesthetic, which enhances readability.
Data Visualization Choices:
- I used a blue color palette for charts to maintain consistency.
- Visualizations were simplified and paired with direct annotations to help readers immediately grasp key insights.
- Side-by-side layouts were utilized to make complex data easier to understand.
Interactive Elements: Scroll-driven storytelling and animated charts were used to maintain engagement while guiding the reader through the narrative.

References

Note: All detailed references are included directly within the Shorthand story. Below are any additional references specific to this write-up:

de Vries, A. (2023). The growing energy footprint of artificial intelligence. Joule. Link
International Energy Agency. (2024). World Energy Outlook 2024. IEA. Link
U.S. Energy Information Administration. (2023). Annual Energy Outlook 2023. EIA. Link
Strubell, E., Ganesh, A., & McCallum, A. (2019). Energy and Policy Considerations for Deep Learning in NLP. ACL. Link
Jones, N. (2018). How to stop data centres from gobbling up the world’s electricity. Nature. Link
Masanet, E., Shehabi, A., Lei, N., Smith, S., & Koomey, J. (2020). Recalibrating global data center energy-use estimates. Science. Link

AI Acknowledgements

For this assignment, I used ChatGPT to help brainstorm and refine narrative structures, as well as to draft sections of text, which I then revised to match my voice and style. I also used ChatGPT to generate initial data visualization sketches, which I later refined using Datawrapper.

Additionally, I utilized Grammarly to proofread my work, ensuring proper grammar and improving the overall flow of the writing.

All visualizations were finalized using Datawrapper, and the storytelling was built using Shorthand.

Final Thoughts

This project was both challenging and rewarding. One of the most exciting parts was translating complex data into relatable stories, like comparing ChatGPT energy use to household appliances — it made abstract data tangible.

The story itself went through a major evolution — I started with a broad topic on AI adoption for business, but it lacked focus. I then shifted to AI’s energy consumption, which provided a clearer direction but was initially too data-heavy and felt more like a report. Through feedback and iterations, I transformed it into a narrative-driven story, making the data more approachable and engaging.

If I had more time, I would have explored more interactive visualizations, like line graphs showing trends over time with scroll effects to make the data feel more dynamic and immersive. I also would have loved to conduct more user interviews to refine the story further.

What I’m most proud of is how the story evolved — starting as a data-heavy report and transforming into an engaging narrative that blends data, storytelling, and actionable insights. It taught me the power of designing for the audience, and how the smallest design tweaks can make a big difference in readability and engagement.

Overall, I’m excited about how it turned out and grateful for the feedback that guided me along the way!

aritra-dataviz-portfolio