Tech Talks After Hours: Bringing Data to Life

Join Esteban Ruiz de Erenchun and Tom Linton as explore the world of data visualization and address questions such as “How do I get started?" and "What are the various types of visualizations and which to use when?". This series is designed to empower viewers to learn data visualization and creative ways of displaying numbers to various end user audiences!

Key Topics Covered

• Defining data visualization: the goal is to provide a clear answer to a specific question — not to decorate a screen with charts
• Inspiration from David McCandless (author of Knowledge Is Beautiful): a journalist-turned-data-viz expert whose framework anchors the series — know the question, tell the story, be clear
• Common mistakes: screens filled with bar charts, pie charts, and treemaps without purpose or narrative — more visuals does not mean more insight
• The journalist's approach: identify the question the audience is actually asking, then choose the visual that answers it most directly
• Helping clients dream beyond their legacy system limitations — they may be asking the only question their old tool allowed, not the best question
• Principle 1 — Maximize the data-ink ratio: remove anything that doesn't add meaning; less is more; every element that isn't carrying information is noise
• Principle 2 — Know your end user's thought process: design visuals that follow the natural narrative (overall → region → product → location → customer)
• Principle 3 — Use perception wisely: if a chart requires extensive verbal explanation, it's the wrong chart; the visual should speak for itself
• Bar charts: when and why they work; avoiding overuse
• Pie charts: best for five or fewer slices; six or more slices signals the wrong chart type; switching to a donut doesn't fix an overcrowded pie
• Bubble charts: encode three variables simultaneously — position (x/y), size (volume), and color (price or category)
• Choropleth / GeoBi maps: for geographic distribution; must be normalized to avoid misleading conclusions — flagged as a future deep-dive topic
• Treemaps: best for showing component distribution across many categories simultaneously
• Gauges: designed for two measures in a single element (e.g., actual vs. budget, or two KPIs side by side)
• This episode is the first in a series — future episodes will deep-dive individual chart types and OneStream-specific implementation