Telling data stories
Last updated
Was this helpful?
Last updated
Was this helpful?
Before you start creating your Juicebox data story, you’ll want to make sure you have a good understanding of your audience’s needs. Everyone who will be using your data story or report has a role in the organization and associated responsibilities. The more your data story or report can help them fulfill those responsibilities, the more value it will have. To do that, make sure you understand your ideal user. Consider these questions:
Who is your ideal user?
What role do they have in the organization?
What makes them successful in that role?
What do they already know about the content of your data story?
What are some ways your data story will influence them?
A good place to start in designing your data story is with understanding the personality of your audience. Our data personality framework gives you a way to understand your target audience.
Like a Myers-Briggs personality test, but for data users, this framework will give you a sense of what you need to emphasize in your data story.
Understanding your audience and their needs is the starting point to gather thoughts for a data story. Now you want to organize those thoughts. One of the best ways to think through your story is to begin by writing. Writing will help you:
Clarify the structure;
Articulate the language and terminology;
Check the flow and transitions;
Test the attention span of your audience.
You don’t need to write an essay; start with a story abstract. A story abstract provides the highest-level description of the key elements of your story. Do you remember those Mad Libs games where you filled in the blanks with nouns, verbs, and adjectives? Here’s a Mad Libs-style template for your data story:
You’ll want to answer the following questions:
What is the topic of your data story?
Who is your target audience?
What are the key questions you want to help them answer?
What are the key measures that will help answer these questions?
What dimensions or breakouts of the data will support the analysis?
What types of insights should they learn from the data?
What actions will your audience take based on these insights?
How will these actions impact the organization’s goals?
How will these actions impact the audience’s personal goals?
The choice of measures (or metrics) can make all the difference in your data story. Measures are your characters. A performance measure represents the thing you want to see improved in your organization, or the thing you want to see reduced or eliminated. It is the thing you want to track how it is changing and why it is moving up or down.
There are four factors to consider when you decide what is the right measure to feature in your data story:
Does the measure have a common interpretation across your audience? In other words, will your audience understand what the measure represents in their world.
Is the measure something that your audience can take action on? The best kinds of measures have clear implications about what should be done as the measure moves up or down.
Does the measure derive from accessible, credible data? Sometimes the most valuable and obvious measures are frustratingly hard to track or unmeasurable.
Can the measure be calculated in a simple, transparent way? Complex measures based on obscure calculations often lead to confusion in the data-driven discussion that you are trying to enable.
Data stories share a similar structure to traditional narratives. The three-act play structure that we are all familiar with sets out a problem (Act 1), allows the characters to change and grow through conflict (Act 2), then reaches a resolution (Act 3). Data stories can take your audience through a similar journey:
Context. Here you are helping your audience understand the background, why the data is important, where it comes from, and what key questions you want to answer.
Heart. Guided exploration is the core of your data story. This is where you provide the visualizations to interact with the data. You want to provide enough flexibility to answer key questions while maintaining focus on the data and insights that matter most.
Actions. The last section of your data story should emphasize the detailed data and results that are most actionable for your audience. This may be a list of individual items or a summary of what you’ve learned from the data.
The following diagram shows how these three stages connect to traditional storytelling.
We want to provide a short checklist of items to evaluate whether you’ve considered important design choices with your data story.
Before your reader dives into the data, you want to set the context. How? At the top of your story, add a slice that gives a title, describes the data, even tells people what kinds of questions they can answer.
Juicebox is designed to do more than just show data visualizations. You can use text to explain what’s being shown in the data. This text could include:
Titles for the visualizations to explain the content. Often the title can be expressed as the question that is answered by the visualization.
Explanations or insights based on the visualization. Is there something you want your audience to see in the data? Tell them. The text areas in Juicebox are an opportunity to be in conversation with your audience.
So much to show...so little attention. Recognize that your audience may be distracted or have a short attention span. Focus is critical in a good data story. This can mean:
Show fewer measures.
Provide fewer filters or dimensional breakouts
Reduce the number of visualizations to get to your key messages more quickly
Let a trusted colleague review your data story and ask them to identify anything that they think isn’t totally necessary to get your message across. If it's not necessary, it should be removed. The cost of more content is high when you risk losing the attention of your audience.
By default, Juicebox will layout your sections and slices vertically. In this way, your audience can read through the content from top to bottom and your filtered selections will pass down the page.
However, sometimes it will make sense to take advantage of the alternative layout options:
Within sections, you can layout slices horizontally. Horizontal layouts will help put more information on the screen at once. They can be useful for displaying groups of measures side-by-side. You might also try laying out a bar chart and trend chart horizontally to visually connect the data in those two charts. Generally, it isn’t a good idea to show more than two or three slices horizontally.
Within slices, you can also choose the layout of the text relative to a visualization. Placing the text to the left of the visualization is an effective way to make your data story easier to read.
A data story will often have multiple sections, perhaps reflecting the three-parts of a story. In Juicebox, you can make visual breaks between sections using color. Find the color selector on the top of each section on the Design tab. By visually breaking up your sections, you’ll help your readers understand where the logical breaks are in your story.
Getting your data ingredient labels right is a small and important design choice. Often when you load data into Juicebox, the data columns may not be written in a way that is easy for your audience to understand. We’ve made it easy in Juicebox to change those labels. What is the simplest and clearest way to describe the data ingredient? Will your audience understand any abbreviations or acronyms you are using?
