Who is your chart for?

People often come to me wanting to know which visualization is the best one for their situation. Before you pick a chart type or visualization approach, it’s important to understand the context in which the chart will be used. The context for a visualization can be defined in many ways, but ultimately it comes down to understanding both the audience and the purpose for a chart.

Audience

A visualization can speak to many audiences and groups, and each one will have a unique set of needs. I like to include both the person and the high-level task/context in my audience description, to help narrow things down. Notice that the what the audience needs in a chart for the first two bullets could be quite different, even though both focus on experts.

• Experts reading a scientific publication
• Experts analyzing their data to find something new
• Students learning a new concept or idea
• Decisionmakers (in a business, people shopping, etc.)
• Patients in a hospital
• People on the street
• Museumgoers interested in something memorable
• People who love charts
• People who don’t understand or are afraid of math

How broad is your audience?

A figure in a typical scientific journal article will be read by tens or perhaps hundreds of people, most of whom are experts in the field and have made hundreds of similar charts in the past. A poster in the subway will be seen by people from all walks of life, some of whom may not even know how to read a chart, and might be afraid to try. A visualization for an art gallery has quite a broad audience, but they will all at least be united by the fact that they are all people who go to art galleries, which implies some level of cultural interest and leisure time. You generally want to define your audience as narrowly as possible, but you should also make sure that you are not over-restricting and ignoring common needs.

Who are you excluding?

Good design should be inclusive and accessible. If you find that your choices are excluding a large portion of your audience, you might want to re-think your approach. Usually, making a design more accessible benefits everybody, not just the subset of people who depend on the information that you provide. For instance, adding information and annotations that support understanding and increase interest for one segment of your audience will usually help others as well. Accessibility can be thought of in several different ways:

• Legibility/Display factors – people must be able to read your visualization to benefit from it. This includes people with disabilities, who may have very specific needs.
  • Text contrast is high enough that you can see the words clearly
  • Text is large enough to be read at a glance, or by people with impaired eyesight
  • Colors can be seen by color-blind people
  • Colors can be photocopied or printed out in black and white
  • Screens can be read outside in the sun, or while driving after dark
  • Display height (for a poster or piece of art) is such that it is comfortable to read
  • Alt-text is available to support screen readers
  • Interactions do not rely on fine motor control and the use of a mouse
• Literacy/Graphicacy – people need to understand your visualization to gain value
  • Clear labels, legends and annotations support understanding
  • Uses familiar or standard chart types, includes explanations and tutorials for less common or unfamiliar charts
  • Step-by-step, logical progression builds understanding and introduces the chart
• Interest – people need to care about your visualization to be impacted or influenced by it
  • Visual and aesthetic appeal attract viewers
  • Good content and an explicit message give readers a clear takeaway
  • Users can see themselves represented in the data, or there are connections to familiar contexts
  • Tone, use of evidence, choice of language/jargon, and interesting topics support the user in engaging with the piece

Context

Where will the visualization be shown?

Once you understand your audience, it’s important to consider when they will use the visualization, and what their needs are at that time. You need to match the chart type to a user’s actual needs, and those needs may depend on what else is going on.

The best choice of chart often depends on how it will be used.

• At a scientific conference
• In a museum gallery
• On a dashboard or instrument control panel
• As a supporting figure in a news article
• As the centerpiece of a data journalism project from the graphics desk at the NYT
• In an information dashboard
• As the centerpiece of a weekly or monthly report
• As the key feature in a website or data product

You might also consider how the viewing context is likely to affect a person’s mental state. Thinking about the expert at a scientific conference, consider how many different things they could be feeling, and possibly all at the same time.

• Exhausted after a long day of talks
• Excited by new ideas
• Overwhelmed by the noise and bustle in a busy conference hall
• Worried that this concept might invalidate their own theories
• Eager to debate the fine points of their ideas with someone who will actually understand and be interested in their work
• Looking for insights that will crack open new ideas and areas to explore

Understanding even a few of these factors can help you think more deeply about what that person needs.

All of these considerations are helping to build a user persona for your audience: a sketch that helps you to imagine your user more clearly, so that you can tailor your design for them. This specificity can also eliminate people, though, so be careful to choose a persona that is truly representative of your audience. Done poorly, persona can devolve into simplistic stereotypes. Thinking through a range of identities for that core persona can help you to examine your own assumptions and gives you a much deeper appreciation for your audience, and how its needs might vary from one person to the next. How does your picture change if the person is: male/female? Black/white/brown? European/American/Asian/African/Middle Eastern? Able-bodied/disabled? Young/old? Educated/uneducated? Rich/poor? A well-known leader/a newcomer?

User experience

Once we have a sense of our persona, we can start to think about the experience that we want to create. I like to think about experience in terms of coffee shops. If I’m going out for coffee with a friend from out of town, I probably want a coffee shop with nice ambience and a little flair. Because this is a rare social event and we have lots of time, I’m in this for the experience all the way. A little bit of ceremony and some thoughtful touches will go a long way toward making this a memorable event. Because I’m likely to be deep in conversation with a friend I don’t get to see very often, I’m not likely to notice if it takes a bit more time, but I am likely to care whether it feels like a special treat for this person that I don’t get to see very often. We’re probably going to enjoy looking through the menu and commenting on the different choices, folding that into our dialogue as the conversation warms up. We’ll look around at the decorations, notice whether the chairs are comfortable, and we might even take photos of the foam art on top of our lattes. In this case, the experience is what matters, and I’m actively looking for an environment that helps me to slow down, enjoy the moment, and soak it all in. (Notice that this description assumes that I have leisure time to meet up with a friend but maybe only a few hours to spend, that we can afford and would feel comfortable in a fancy coffee shop, that we’d prefer to meet in a public place rather than my home, and that I live or can get to a place where such a thing exists.)

Compare this with a different person coming into the same coffee shop. They’re on their way to an important meeting. They didn’t get any sleep the night before because they were so stressed out over their presentation. The train was late and they’re rushing to get to the office, but they really need a cup of coffee to carry them through the first part of the day. This person is likely to experience the exact same coffee shop completely differently. They’ll be annoyed that there are so many choices, tapping their foot while the barista chats with the person in front of them in line. They probably don’t want foam art, chit chat, or fanciness of any kind: they just need to get their coffee and get out.

Both of these situations are completely reasonable, but they lead to vastly different decisions in terms of how you approach designing a coffee shop. At the end of the day, both persona just want a cup of coffee, but how you get there is completely different depending on which set of needs you choose to support: in one case, you go for the long, slow, over-the-top experience, and in the other you want a ship shape production line.

Visualization is very similar: at the end of the day, people just need to access the data, but how they choose to do so will vary greatly depending on the context and their specific task.

If you’re planning to spend an entire Saturday going to a museum or an art gallery, you’re probably going to take the time to explore, and you’ll be delighted to watch a complicated story unfold as you work your way deeper and deeper into an interesting visual narrative. The desire for a memorable experience is what brought you here, and you’re hoping to find something captivating that will draw you in. Emotional and aesthetic gratification will be critical for helping you to stay engaged and interested in that piece, especially when there are so many other interesting things to see.

At the other extreme, imagine that you are a stock broker or a fighter jet pilot whose career or survival depends on making split-second decisions under lots of pressure. You don’t have time to carefully peel back the layers of information to expose a deeper story; you need the important information laid out in front of you, all at once. The cleaner and clearer, the better: mistakes can be fatal. (Whether this is the best environment for decision-making is a whole different article, but there’s no question that this situation presents a very different set of user needs.)

In both cases, a single person could match both persona, on different days and at different times. Depending on a user’s experience in that moment, they might have entirely different needs. Forgetting to consider the experience and needs of your user – and how they are situated in that context, and that moment – is a recipe for failure.

Understanding the context of a visualization also helps to create some important boundaries for the form of the visualization, all of which affect our choices for how to design.

• Will the visualization be printed or digital, static or interactive?
• How long are users likely to spend on the piece?
• Will people be looking at this in person, on their phone, on a laptop computer or movie screen?
• Will the visualization be viewed close up, or from far away?
• How many people will be viewing it at one time?

Designing for your audience

All of these considerations can affect chart selection and design in many different ways. Understanding your audience needs or wants can help you begin to frame the different choices for your chart display.

• Chart literacy, or graphicacy. Does your user know how to read a chart? Are they fascinated by data, or permanently scarred from a bad math class experience in middle school? How “advanced” a chart are they likely to understand?
• Need for quantitative accuracy. How critical is it that your audience is able to read exact numbers from this chart? Are they likely to know how to do so, or do you need to help them out?
• Amount of detail required, or desired. Is your audience interested in digging through all of the information and details, or are they mostly interested in getting to the conclusion?
• Annotations and supporting information. Do you need to include annotations to explain what’s going on? Would it be helpful to explicitly point out specific facts?
• “The reveal.” Is your data best presented as a logical argument, a slow unfolding of information with increasing complexity, or a clear statement of a simple fact?
• Display specifics. How much detail will your audience be able to see and interact with? Based on the display context that you’re considering, will users be able to see what they need in your chart?
• Tone or approach of your piece. How will you engage your audience? What tone or language is appropriate, and what’s likely to be interesting to them?