Observable

Too often, vague requests and messy datasets lead analysts down the wrong path. That's why we put together a practical framework for moving from raw data to clearer, more useful insights. In this guide, you'll learn how to quickly and efficiently discover patterns and trends, develop data visualizations, and communicate your findings so that stakeholders can understand and act upon your analysis: https://lnkd.in/giyYpmsf

When you're working closely with a dataset, insights may be obvious to you but less clear to stakeholders unfamiliar with the data. Check out 8 tips for developing clearer and more interpretable data visualizations, including highlighting key elements, decluttering, adding takeaways in titles, and including useful context. Ready to start building more impactful data visualizations? Explore examples in the Observable Plot gallery: https://lnkd.in/gfKh_hCP

Many of today’s data tools still separate writing, running, and inspecting code. That was manageable when humans were the primary authors, but it breaks when AI can generate thousands of lines of code in seconds. This leaves us with systems that run, but aren’t understood or trusted. If AI is going to be part of real data work, we need new mediums that keep humans in the loop of understanding, not just prompting. Read more from Observable co-CEO Julio Avalos: https://lnkd.in/g26JKYKm

AI is quickly remaking data analysis, but it's still important to ask the right questions of your data. When analysis was slow and expensive, bad questions got caught in review. AI removes that friction, which means that poorly framed questions can now produce polished, convincing, and ultimately wrong answers faster than ever. We wrote about how data teams can build strong inquiry habits to create better outcomes. Learn more on our blog:

AI can write code. But neither you nor it understands what it's written. That's not a model problem. It's an environment problem. For decades, software development ran on a tight loop: write, run, inspect. Comprehension was built into the process. AI breaks that process by generating code faster than any human can review it. Logic you didn't write, variables you don't recognize, and systems you didn't design only add to the complexity. We've dramatically improved "write" without corresponding improvements to "run" and "inspect." The result is opacity at machine speed. Learn more about the fix in this blog post from Observable's co-CEO, julio A. https://lnkd.in/g26JKYKm

🧇 Happy International Waffle Day! Did you know that Observable Plot has a built-in waffle mark? Visit the Plot documentation to see examples and learn about helpful options to customize chart units, adjust spacing, round values, and more: https://lnkd.in/g3PEJQK5

Standard chart types like bar charts, line charts, and pie charts are ubiquitous in BI tools. But, for some tasks, a more unique chart type might be a better fit for the data. In this post we highlight useful alternatives such as ridgeline charts, streamgraphs, and waffle charts, and share reusable Observable Plot code to spin them up on your own: https://lnkd.in/gUJ9zuQ3

🤔 What can data teams learn from data journalism? By adopting some of data journalism's best practices — such as knowing your audience, understanding your data, personalizing visualizations, and designing for interpretability — data teams can deliver more insightful visualizations, and improve how they work with stakeholders. In this blog post, we unpack lessons from recent conversations with data journalists that data analysis and BI professionals can apply to their own work: https://lnkd.in/gvK33YAg

Pi day is tomorrow, which naturally has us thinking about pie charts! Check out 8 options to visualize parts of a whole, from pies to waffles to treemaps. Then, head over to the Observable Plot and D3 galleries to explore hundreds of chart examples that you can quickly fork and customize. 👉 Plot gallery: https://lnkd.in/gfKh_hCP 👉 D3 gallery: https://lnkd.in/eqbsGrw

AI agents are making it easier than ever for people across organizations to explore data and find answers on their own. But broader access makes a strong data culture grounded in learning, transparency, and responsible practices more important than ever. Data teams have a critical role to play in shaping that foundation. Learn practical ways analysts can help build and sustain a healthy data culture within their organization 👇