Key Takeaways:
• Exploration vs Explanation: The difference between the analysis phase (exploration) and the presentation phase (explanation) is crucial. Don’t show the audience messy, complex charts. Present only what’s necessary, simplifying and guiding the reading.
• Avoid biases: Two psychological barriers to avoid are “False Confidence” and “Curse of Knowledge.” Don’t assume others see the data as you do. The communicator is responsible for clarity.
• W.A.Y.E.D. Test (Where Are Your Eyes Drawn?): Check if the audience’s focus is on the key message using contrast and color with a simple test.
• Adapting the message: The same data set can tell different stories depending on the audience. The presentation design must be tailored to the specific needs of the listeners, customizing data for different groups.
• Words matter: Don’t be afraid to use words to support numbers. Descriptive titles, annotations in charts, and verbal explanations are essential to avoid misunderstandings and guide the audience in reading the data correctly.
• AI and human value: While AI can automate data analysis, human connection makes the difference. The Data Storyteller of the future must humanize numbers, interpreting and contextualizing data empathetically to inspire actionable outcomes.
Those who work with data know how easy it is to get lost among tables, charts and reports. But presenting an analysis is something else: you need clarity, intention and the ability to guide the listener’s gaze towards what really matters. This is where my conversation with Cole Nussbaumer Knaflic, author of Storytelling with Data and one of the most authoritative voices in the visual communication scene, started.
Through the Lean Presentation Design method and the book Presenting Data, I have tried to give practical tools to help those who communicate with numbers to be more incisive, more aware and more effective. Cole followed a different path, which led her to work at Google in people analytics, but we immediately found ourselves aligned on one thing: the value of data only emerges when we know how to make it understandable and usable for those who receive it.
The discussion was full of insights. In this article, I have collected the six key lessons that emerged from our conversation: six principles for turning numbers into stories capable of generating understanding and, above all, action.
1. The quantum leap: from Exploration to Explanation
Many professionals fail because they project their backstage on the screen. In other words, they show the audience the graph as it came out of the analysis, with all the clutter of exploration. Cole draws a clear line between two distinct phases of working with data:
- Exploration: this is the analytical, disordered and complex phase, in which you look for patterns and test hypotheses. You do this work for yourself, to understand what the data says.
- Explanation: it’s the act of translation for the audience. Once you’ve found the insight, you need to clean up the chart and simplify the message. The audience does not have to interpret; You, as an expert, must provide your own interpretation directly.
Lesson: Don’t delegate to the audience the effort of figuring out what the point of the slide is. If they have to analyze the chart to figure out what you mean, you’ve already lost their attention. You make the leap from exploration to explanation, presenting only what is needed and explicitly guiding the reading.
2. Defeating Bias: False Confidence and the Curse of Knowledge
We discussed two psychological obstacles that frequently derail technical presentations:
- False confidence: The illusion that because we are rational people and have drawn some conclusion from the data, others will inevitably come to the same conclusion when looking at that same data. Wrong: everyone brings with them different experiences, perspectives and prejudices; therefore, it can interpret the data differently.
- Curse of Knowledge: Once you know a dataset well, you no longer remember what it was like to know nothing about it. You are so immersed in the information that you take for granted what isn’t obvious to the audience. This leads you to create dense and technical slides, thinking that they are “obvious”, while for the listener they may be indecipherable.
Lesson: the responsibility for clarity lies 100% with the communicator, never with the public. In other words, if others don’t understand a graph, it’s the communicator who has failed in translation, not the audience who is “Uninformed”. Cole stresses that it’s our responsibility to recognize these biases and actively counter them by putting ourselves in the listener’s place. We need to present data in a way that is easy to understand, without assuming prior knowledge.
A great data storyteller’s real strength is the ability to view the information from the audience’s perspective and keep a fresh outlook, even when they are highly knowledgeable about the subject.
3. The W.A.E.D. test (Where Are Your Eyes Drawn?)
Cole introduced a quick and practical test to check the effectiveness of a slide. It’s a simple but revealing method: close your eyes, open them again and notice where your gaze falls in the first two seconds. Basically, you simulate your audience’s first look at the slide.
During the live, we did this experiment on a dashboard crowded with elements. The result? The eye immediately fell on the logo in the corner and on an often decorative black border, instead of on the critical datum in the center. Not quite the desired effect!
You can think of this test as quickly glancing at a slide while squinting. By doing this, you filter out the small details and immediately see what stands out most.
This simple test allows you, even if empirically, to significantly improve the design of the slides by attracting attention where they are really needed.
Lesson: Use visual contrast and color strategically to guide the audience’s gaze to the key message. For example, Cole showed us how to highlight the most important data with a bright color (her “strategic blue”), leaving everything else in gray in the background. In this way, attention is forced where it’s needed. So, test the W.A.E.D. on your slides: if your eyes don’t go where they should right away, revise the design until the visual focus matches the focus of the message.
Heat maps with artificial intelligence
When I want to know how my audience’s attention is distributed or I want to test how the design affects by changing areas of interest, I use AI to create heat maps. In the example below I used Microsoft Copilot.
To do this, I use a structured prompt template to ensure the statistical validity of the heatmap.
4. The same graph, three different stories
One of the most powerful moments of the conversation was seeing how a single set of data (in this case, the volume of customer service calls) can generate radically different stories and interpretations based on the audience to which it is presented. It was an excellent demonstration of adaptation to the public:
- For the operations manager: the graph highlighted the hourly peaks of calls, to identify the moments of greatest traffic. Insight and action: Maybe stagger your team’s lunch breaks so that you always have enough agents during peak calls.
- For the HR team: The same dataset was reorganized to show differences between days of the week. For example, it turned out that Friday is the day with the least call traffic. Insight: This can help you plan vacations, remote work or team-building activities on Fridays, when the impact on customer service is less.
- For the IT team: The focus was on technical anomalies, such as highlighting sudden drops or suspicious spikes due to system issues. Insight: Those quirks in the graph direct IT to look for and fix possible bugs or malfunctions in the calling platform.
Lesson: Design is not decoration but it is the alignment of data to the specific needs of the listener. The same graph should not be presented in the same way to everyone, because each audience has different information needs.
In this example, we’ve “sculpted” the data story tailored to different interests. When preparing a presentation, always ask yourself: who do I have in front of me and what really matters? Then adjust your visualization accordingly. Sometimes, if the audiences are very different, it may even be worthwhile creating separate versions of the same presentation, each focusing on the aspects relevant to each group.
5. The power of words: beyond the descriptive title
We discussed the famous phrase: “A picture is worth a thousand words”. Cole added a fundamental corrective: “Yes, but only if it says the same words to everyone who looks at her.” In other words, a graph alone is worth a thousand words only if it is clear and unambiguous to anyone who observes it.
If a chart is mute, meaning it lacks context and explanation, everyone will draw different conclusions from the data. And that’s a big deal! During the session, Cole showed how the same graph, presented without a title or labels, was interpreted in discordant ways by the various observers. However, it was enough to add a few targeted words, an explanatory title at the top that summarized the message, clear labels on the axes and maybe a few notes directly on the data series and suddenly everyone was reading the same story in that graph.
Lesson: Don’t be afraid to use words to support numbers. Use executive titles, slide titles that already contain the message or recommendation (e.g., “Increasing sales require more staff in stock” instead of the bland “2023 sales trend”). Add annotations directly on the graph to explain peaks or troughs, highlight with text what you want to point out. Words are the bridge that prevents misinterpretations: they guide the audience in interpreting the numbers exactly as you want. In short, images and words must work together to tell a single clear story.
6. The AI Era: The Return to the Human Factor
Obviously, we talked about Artificial Intelligence. If AI today can automate data analysis and even create charts or drafts of presentations, what is left for us humans?
The human connection: Cole’s answer focused on what AI can never replace. An analysis can be automated, but a handshake, an empathetic look and the trust that is created between people. In an increasingly technology-driven world, human added value lies in the ability to connect with other human beings.
Lesson: the role of the Data Storyteller of the future will be to know how to read what is not written in the data. It means understanding the human, political and corporate context around numbers, grasping the nuances and implications that go beyond the numbers themselves. It will be our job to humanize the numbers, contextualize them and use them to inspire confidence and push for change.
Ultimately, communication, empathy and critical thinking skills will become even more important: they are the ones that will transform a cold automatic analysis into an engaging story that ignites concrete decisions and actions.
Conclusion
Presenting data in a lean way means eliminating the unnecessary to leave room for what really matters: insight. As I always repeat, less is more. But as we learned from Cole, that “less” must be chosen with surgical precision to be effective.
FAQ – Frequently Asked Questions
Why is descriptive title important in a chart?
A descriptive title helps to clearly communicate the main message of the chart, avoiding misinterpretations and guiding the audience towards reading the data correctly.
What does “Data Storytelling” mean?
Data Storytelling is the art of transforming data and numbers into an understandable and engaging narrative, combining visual and textual elements to convey a clear and persuasive message.
What is the role of words in data analysis?
Words provide the necessary context to the numbers, explain trends, highlight anomalies and make the message accessible to everyone, avoiding misunderstandings.
How does the presentation of data change with Artificial Intelligence?
Artificial Intelligence can automate analysis and charting, but human value remains critical to interpreting, contextualizing, and communicating results in an empathetic and effective way.
How can I improve the effectiveness of my data submissions?
Use clear, descriptive titles, integrate explanatory annotations into charts, eliminate the superfluous, and always accompany the data with a narrative that makes it easier to understand and guides the audience towards the desired decisions.
What is lean data presentation?
Presenting data in a lean way means reducing to the essentials, selecting only what is really needed to convey the main insight, for more effective and incisive communication.
Comments on 6 key lessons on Data Storytelling: what I learned from my conversation with Cole Nussbaumer Knaflic