Join Bill Shander for an in-depth discussion in this video The right data, part of Data Visualization for Data Analysts.
- This is a course for data analysts, so it may seem a little crazy to have a chapter about using the right data, as though you don't know the right data to use to tell your story. But here's the thing, from my experience, often my clients who are closest to the data do struggle with this idea. The struggle comes in two primary forms. You've heard the phrase, "If your only tool is a hammer, everything looks like a nail"? Data people often get stuck thinking about their data as a tool in their toolbox. So the first thing they ask themselves is, "What chart can I show?" When they should be asking themselves, "What story can I tell?" When I do workshops with clients, this comes up again and again.
I'll assign a task and teams go off into their corner, and they immediately look at their data and start sketching charts. Now, it's sort of good in a way, at least they're being visual, but I always remind them, "Take a step back, "think about the story first, "which will always lead you to the right data, "and then to the right visual approach." The second issue is that data people are so used to their data, that they sometimes go into a kind of data haze. Think of it like snow blindness. There's so much data and you fall so in love with a certain view of some really interesting data, that you're overwhelmed by it.
Your mission becomes how can I use this data, rather than thinking about why you're using it, and does it tell the story you're trying to tell? Essentially you get hung up on exactly the data you have. You know exactly what you've measured, so you focus on how you can visualize exactly that. It's easy to get tunnel vision. Now since you're probably a sophisticated and experienced data analyst, this likely works for you much of the time. But, I frequently come across opportunities to change the thinking a bit. For example, I had a client investigating the link between R&D spending and the decline of a company over time.
They had Employee Headcount and R&D Spend as their two data values. And they planned to chart it all together in a chart like this, with two axes, right? Blue is R&D Spending, that's the left axis. Green is Employee Headcount, and that's on the right axis. So in addition to the fact that I would always work hard to avoid two axes charts because they tend to be confusing. I thought there was a better way to get at the correlation they wanted to show. And so, this isn't rocket science, I suggested they show the ratio between these two numbers, right, R&D spend per employee as one line.
Not only is it simpler, of course, and probably half of you watching this had thought of that as soon as I said I looked at an alternative approach. But it also reveals that interesting spike towards the left third of the chart here. There was this spike in the data that the other two charts didn't reveal. Showing ratios and relationships often leads to interesting outcomes. Again, this may sound obvious, you probably do this most of the time, but I bet even those of you who are shaking your head a bit about this being too obvious, if you're being honest with yourself, you realize you do this too.
So the main point here is to always take the time and really think about your data. Think about your story, think about whether the data you're illustrating really is the best data to show. Or, if there's something you can do with it to bring out the best of it for your story. Another problem, and this is related to much of what I've been saying in this course, is that you need to always remember to work hard to narrow down to just what's important. So this slide, which we saw earlier when we were talking about color briefly, there's lots going on here, right? We're looking at The Outlook for Revenue Growth is Dim, as Widget revenues by category and year.
I see numbers, I see segmentation, I see two different types of charts. Sometimes I'm looking at one way of categorizing data, sometimes another way. I see call-outs pointing out these percentage rise and fall, et cetera. We're gonna go through this more in rethinking charts later on in this course, but what are we trying to actually communicate here? If we take the headline at face value, and the point is that the outlook for revenue growth is dim, so we're really just concerned with 2014 and 2015 as far as the main point of this chart.
So why are we going all the way back to 2002 with our data? Maybe there's a reason, or maybe we could have narrowed it to just 2008 plus, just telling the story of the post-recession flatline that will be continuing. Long story short, just remember, when you're working with your data to be humble. Even though you know your data and you know what you're doing, don't be blinded by it. Don't get tunnel vision. Always remember your story, and then rethink whether your data and how you're thinking about it is exactly what you need. Often even small tweaks can make a big difference.
- Why visual communications matter, and how they work
- Communicating via story
- Communicating with color
- Using legends and sources
- Sketching and wireframing
- Rethinking slides, charts, and diagrams
- Rethinking your templates and brand guidelines