(C) 2017 All Rights Reserved. Visme is a product of Easy WebContent, Inc.
Designed by HindSite Interactive. Proudly made in Maryland.
A well-known data viz expert, educator and blogger in the community, Cole Nussbaumer Knaflic is deeply rooted in her mathematical background when it comes to storytelling with data.
Cole had a robust career in data analysis including a People Analytics role at Google. It was at Google that she started training employees at Google to tell stories with data. Her popular blog Storytelling with Data is a must read for anyone who is particularly interested in data visualization in the business context. She also authored the book “Storytelling with Data: A Data Visualization Guide for Business Professionals,” a handy manual packed with practical lessons on how to make data the pivotal point in storytelling.
Cole now travels around the world training organizations and individuals to tell better stories with data. We recently reached out to Cole for deep insights on data storytelling so that you can learn from her best practices.
1. In your work, you talk about visualizing data in the professional context such as presenting business analytics. A lot of our users see the need to visualize financial/business data because no one wants to read lengthy reports anymore. Do you have any insights to share with us on how to keep the business data/charts simple yet visually interesting?
The challenge with any sort of regular reporting is that people tend to lose interest over time. That said, regular reports serve a great purpose, because you can use them to identify stories.
If you’re the analyst, you should look at your reporting in detail—take note of where are things in line with expectations and where are they not. Where you see things that are surprising or interesting, these are nuggets of potential stories to dig into and better understand.
Once you identify interesting takeaways or stories, you don’t want them to get lost in the report, so my recommendation is to pull those things out up front and use the lessons covered in my book.
You can think of this like the cover for your regular report. This allows you to say to your audience something like, “All of the usual data is there and you’re welcome to look through it, but I’ve already done that and here are the interesting things you need to be aware of or take action on this time.”
My view is that it isn’t The graph that makes the data interesting. The data is interesting when you’ve taken time to think about your audience and what matters to them and presented the data in a way that will make that clear and actionable. Visualizing the data effectively and building a robust narrative around it to explain to your audience why they should care will set you up for success.
2. What is your approach to chose the best format to visualize a Data set?
There isn’t a single best approach for visualizing a data set. What you focus on and how you analyze and visualize it will vary for different situations. Once I’ve identified a story and I’m trying to figure out how to visualize it to create that magical “ah ha” moment that graphs done well can do, I allow myself time and freedom to iterate and look at the data multiple ways.
Consider what you want your audience to be able to do with the graph or see via the graph and design with the goal of making that as easy as possible. Then seek feedback. Create your graph and show it to a friend or colleague. Have them talk you through their thought process—where they pay attention, what questions they ask, what observations they make—all of this will provide insight on whether your graph is serving its intended purpose, or if it isn’t will give you pointers on where to concentrate your iterations.
“Never simply show data; rather, make data a pivotal point in an overarching story and use it to drive your audience to action.” Click to tweet
3. Once one has visualized the data, it doesn’t necessarily mean we can get everyone to pay attention. There is still the job to “present the story”. How do we craft a data stories in a way that truly engage our audience?
I’ve been thinking about this a lot lately as I continue to hone my own storytelling skills and test out different ways of teaching others how to tell stories with data. Typically, when presenting analytical results, we follow a linear structure. We start with the hypothesis, or question we set out to answer, then talk about the data we collected, then the analysis we undertook and methodology we used, and then end with our findings.
This is the most natural way to “tell the story,” because it’s the process that we went through. But that doesn’t mean this is the most effective way to present our analysis to our audience. It’s not a story. Stories have a shape, they have a plot, a rising action, a climax where tensions reach their highest, a falling action, and a resolution. Stories told well are naturally engaging and we can use this structure for the stories we want to tell with our data for great benefit.
This means rethinking the way we present analytical results. Also I find it’s critical to think about telling stories not for ourselves or for our data or analysis, but for our audience. What’s the tension that’s going to make them pay attention? It’s by keeping the audience in mind throughout the process and designing everything—from the graphs to whether/where to include data, to the story—with them in mind that we put ourselves in a position for success in getting our point across and motivating action.
4. What is your suggestion on visualizing massive amount of qualitative data?
There are a lot of potential directions to go with this question. I’ll outline one specific strategy. When you have a lot of qualitative data, I find one often effective approach is to quantify some of it and pair this summary with some specific verbatim comments.
For example, if you’re analyzing customer feedback, you can summarize into the main positive and negative themes, showing relative frequency numerically, and then pull out a couple of pithy quotes as makes sense to provide additional insight as an example of what people are saying.
Verbatims are awesome because they have emotion, they come from a person and we can often relate to what someone else is saying. The challenge is that when you put a lot of text in front of your audience, it can be visually overwhelming.
So think about making the text easily scannable by pushing some to the background and emphasizing only the most important words or phrases (via bold, color, etc.) so the information can be more quickly processed by your audience.
5. How do we go about avoiding false expectations from fake data?
I’ll use this question to bring up the golden rule of data visualization, which is don’t lie with data. A discerning audience will see through one-sided data or data that is bent to serve an agenda, so resist any urge to twist your data to make it say what you want it to say!
When it comes to expectations, I think sometimes people falsely think that data will always provide the “answer.” Yes, data can often help direct us in smart ways. But we can also get into a cycle of analysis-paralysis, where data becomes a crutch for postponing decision making.
If you’re in a situation where someone keeps asking for more data and more data and more data, try to get to the core of what they are trying to solve for. If more data won’t help or change the decision, other factors may need to be considered.
6. How much data is needed to translate visuals? In other words, there is infinite amount of data to chose from. How do you select data sufficient to tell a story but not so much that the audience get overwhelmed?
The answer to this will change depending on the situation. For me, the biggest factor to consider is the audience and what the right amount of data for them will be.
Some audiences will be convinced on minimal data or are more influenced by other factors. Other audiences will need to be convinced of the robustness of the data, the underlying methodology, etc.
Start by considering your audience and their needs, then use this to guide when and how and how much data you incorporate into your presentation. Just because you analyzed a ton of data does not mean you need to show a ton of data.
There are tradeoffs in all decisions related to communicating with data. The goal when it comes to “how much data” is enough to convince your audience without crossing the line into “too much” and your communication becomes visually overwhelming.
I’ll briefly describe two strategies you can employ that can help with the latter without sacrificing the former. First, in a live presentation or meeting, build your graph/visual piece by piece so you can talk your audience through what they are looking at and have them focus exactly where you want them to as you’re talking through the relevant piece. In this manner, you can build up to a final visual that won’t feel as complicated as if you had put it in front of your audience all at once.
If you aren’t presenting live, another strategy can be to push a lot of details to the background (make them small, use grey font/lines) and focus your audience’s attention (via color, size, placement) only on the most important parts of what you’re showing.
Also use words (via titles and text on the page) to make it clear why you want your audience to focus here. Each of these are strategies for providing detail without it becoming visually overwhelming. Also sometimes it’s helpful if you’re struggling to visualize something complicated or have a really dense page or graph, think about whether it might make sense to break it into multiple pieces.
Turning boring information into beautiful visuals doesn’t have to be a daunting task. At Visme, we help you use the drag-and-drop visual content tool with the best help content.
We will bring more expert content like this Q&A with Cole above. Stay tuned!