Getting Past the Jargon in Data Visualization
I picked up the Storytelling With Data Ebook a few years back when I was still trying to justify every chart I made to stakeholders who had three seconds of attention before bouncing. It's not the only resource out there, but it's the one that made me stop trying to show everything and start thinking about what people actually need to see. The book is essentially a collection of techniques for stripping away visual clutter and directing attention where it matters. It's written by Cole Nussbaumer Knaflic, who comes from a presentation and communication background rather than pure statistics. The core framework in the book revolves around a sequence: clarify your context, identify your audience, determine what insight you're trying to communicate, and then build a visualization that supports that single idea. It sounds obvious until you're sitting with a messy Excel sheet and twelve conflicting data points. The book walks through real slide decks, showing before and after versions of actual presentations from corporate environments. That's where most of the value lives—not in theory, but in seeing what happens when someone takes a cluttered chart and deliberately removes everything that isn't serving the message. One technique I found myself using constantly is the dot plot over the bar chart. It sounds minor, but it cuts reading time significantly because dot plots force your eyes to scan along a single axis rather than jumping between bar lengths. The book explains why this works, and more importantly, it shows you the exact steps to convert a column chart into a dot plot in Excel without relying on add-ins or complex macro scripts. I had a client who needed a quarterly comparison across forty business units. The original bar chart was a wall of color. I stripped the axes down to zero baseline, removed gridlines, reordered by value, and used color only to highlight the outlier. It went from a page people skimmed past to one they spent thirty seconds actually looking at.
Another concept that's harder to execute than it sounds is the use of white space as a directional tool. You can separate sections with padding alone instead of lines or boxes. Lines draw the eye into regions. White space tells the brain that two things belong together by proximity, which is more subtle but far less noisy. I learned this the hard way when a stakeholder asked me why their dashboard looked like a coloring book. The fix wasn't adding more filters or drill-downs. It was removing borders from tables, cutting fill colors to a single hue, and letting the data speak through position and size alone.
Where the Approach Falls Short
The book works best for business presentations, slide decks, and one-off reports. It's not designed for interactive dashboards or live dashboards where the user controls the exploration. If you're building something in Tableau or Power BI with layered dimensions and filters, the principles translate partially, but you'll hit a wall quickly. The methodology assumes a static format where the author has full control over the narrative. That's fine for a board deck, but it breaks down when you're handing someone a self-serve tool and expecting them to find the right answer without hand-holding. There's also a risk of oversimplification. Some of the examples gloss over situations where nuance actually matters. A stacked bar might look cleaner than a grouped bar, but stacking can obscure differences in the middle segments. The book doesn't always flag when simplifying would mislead rather than clarify. I ran into this with a revenue breakdown where the mid-tier categories were the story. Removing visual elements made it harder to see a meaningful dip in the second segment. The workaround was keeping the grouped bar but thinning the stroke weight and using a muted gray for the non-critical categories instead of hiding them entirely.
Downloading and Using the Material
The Storytelling With Data Ebook is available through the official website at storytellingwithdata.com, where you can purchase or access it depending on current promotions. There are also supplementary worksheets and practice datasets that accompany the course, which I'd recommend grabbing if you plan to apply the techniques rather than just read about them. The practice exercises are where most people actually internalize the concepts. Reading about context and audience identification is one thing. Sitting down with a real dataset and trying to strip a chart down to its essential message without losing accuracy is another. I've seen people waste hours trying to make charts look professional by adding gradients, shadows, and custom fonts. The book flips that approach entirely. Professional doesn't mean decorative. It means the reader gets the right answer faster. I remember a project where I spent six hours redesigning a single slide. The original had a pie chart, two inset images, a legend, and four text boxes. The final version had one dot plot, a single callout line, and a sentence that said exactly what the audience needed to know. It took longer to build than the original, but the payoff was immediate—questions shifted from "what does this mean" to "what should we do next," which is the entire point. If you're working in a team where everyone is producing reports, the book's framework can serve as a shared standard. It gives you language to push back on decisions like "let's add more colors to make it pop" without sounding arbitrary. The principles are grounded in cognitive science and visual perception research, not personal taste. That makes it easier to explain why a chart that looks simpler to you actually communicates more effectively to someone who isn't immersed in the data every day.
