Less iteration, better charts: rethinking how we create visualizations

"Creating a chart is usually an iterative process. We build a rough first version, then tweak the colors, change the chart type, adjust the title, and refine the styling until we are satisfied. Thanks to AI coding tools, this cycle has become even faster."

But what if we approached it differently? What if coding were the final step instead of the first? This talk explores a design first, code second process that starts with the "why" before moving to the "how."

We will discuss why, if you want to create great charts, you should stop starting with code and instead clarify your goal upfront. This approach not only leads to better visual outcomes but also changes how we write code and how we use AI coding tools. When the intent is clearly defined, prompts become more precise, the generated code is closer to the desired result, and the number of revisions drops significantly. By defining intent before implementation, you become stronger in both visual storytelling and software development while shortening the iteration cycle.