Thursday, February 26, 2015

annotate with text

In keeping with my prior post, I'm sharing another "most-photographed slide" from my recent workshops.


My voiceover typically sounds something like this:

When it comes to storytelling with data, one very important component of stories is words. There are some words that absolutely have to be there: every graph needs a title and every axis needs a title. This is true no matter how clear you think it is from context. The only exception that comes to mind is if your x-axis is January, February, March, etc., you probably don't need to title it "months of the year." You probably should make it clear what year it is. Any other axis needs a title. Label directly so your audience doesn't question what they are looking at.

Don't assume that two different people looking at the same data visualization will walk away with the same conclusion. Which means, if there is a conclusion you want your audience to reach, you should state it in words. Use what we know about preattentive attributes to make those words stand out: make them big, leverage color and/or bold, and put them in high priority places on the page like the top.

Speaking of which, that title bar - stories have words: annotate with text - is precious real estate. It is the first thing your audience encounters when they see your screen or your page, so make it count. Use this space for active titles, not descriptive titles. If there is a key takeaway for your audience, put it there so they don't miss it. It will also help set up the content that is to follow on the rest of the page.

When you are communicating with data, there are some words that usually need to be there: data source, as of date, and perhaps notes on assumptions or your methodology. These are necessary words but they don't need to cry out for attention. Use what we know about preattentive attributes to emphasize the important parts of your visual and also to de-emphasize less critical pieces. Footnotes can be small, the text can be grey, and they can be in lower priority places on the page like the bottom.

Use words to title, label, and explain; they help make your data visualizations accessible!

To see this and other storytelling with data lessons firsthand, attend one of my upcoming public workshops.

Monday, February 23, 2015

consulting for context

Upcoming workshops: Details have been set for my Chicago public workshop - it will be on 3/23 and you can register here. I'm offering a full day workshop in London as a pre-conference session ahead of Tucana Global's People Analytics conference in April (it can be registered for separately from the conference and content will be made relevant to non-HR as well), details here. I'm also in the planning stages of a public workshop in Dallas in early May - to be notified when it is scheduled or suggest/volunteer a venue, click here.

Speaking of workshops, I've conducted a lot of them over the past three weeks. It takes nearly all of my fingers to count them. I've packed so many into a small time period that I've started to observe some interesting patterns across them. I'll tell you about one such pattern today.

I tend to begin each workshop by asking attendees to do something that I think may make some people feel uncomfortable: take whatever electronic devices are within their reach and place them slightly out of reach. (There is little that is more distracting in a shared learning environment than someone typing on their laptop or texting on their phone!) 

Still, somehow, people end up with phones in their hands.

To my amusement, however, for the most part they don't appear to be using their phones to check their email or update their Facebook status, but rather to snap a quick pic of the slide I'm discussing. This is one use of an electronic device in my workshops that I might be willing to condone! The pattern I referred to earlier is that it seems to be the same handful of slides that prompt said picture-taking. I believe this is an indication of the usefulness of the content, so thought I'd share some of that content with you here.

One popular slide is about consulting for context and lists the following thought-starters:
  • What background information is relevant/essential?
  • Who is your audience? What do you know about them?
  • What biases does your audience have that might make them resistant to your message?
  • What data is available that would help strengthen your case? Is your audience familiar with this data, or is it new?
  • What would a successful outcome look like?
  • If you only had one minute or a single sentence to tell your audience what they need to know, what would you say?
My voice-over of this slide usually sounds something like the following. Often, when you are putting together a communication, it is at the request of someone else: a client, a stakeholder, or a manager. Sometimes, the person requesting the work has things in their head that are important to understand that they may not think to say out loud. The above are some questions you can use when that's the case, to try to tease the full context out of the person requesting to make sure you have a robust understanding of the need to communicate before you start building the actual communication. Being clear on the context up front can drastically reduce iterations down the road.

I find the last two questions in particular can be really useful for getting at the main message you want to communicate. I often use these when I'm working with clients to get clarity on what they want to say. What would a successful outcome look like? Or, if you had a really finite amount of time or number of words to say what you need to say: what would that sound like?

I use this to set up the concepts that are typically covered next (that I've blogged about here previously): the 3-minute story and Big Idea.

Interested in other parts of my workshops that are prompting people to photograph what they see? Stay tuned here and I'll highlight some others in upcoming posts!

Wednesday, January 14, 2015

Jon and I chat about non-zero baselines

Earlier this week, I got a rare chance to chat in person with my friend Jon Schwabish (policyviz.com) who was in town for a quick trip. We recorded a short chat on a topic that's been generating some buzz in the data viz world lately: non-zero baselines (which I've previously blogged about here and here) and whether it is ever ok to start the y-axis of your graph at a value other than zero.

Check out the following to hear our chat. Leave a comment here or on Twitter (me | Jon) to join in the conversation!



A couple visuals we referenced while talking:

The graphs we had up on Jon's computer as we were chatting
From Drew Skau's recent Visual.ly blog post