Why Do We Visualize Data?

Find and communicate actionable insights

See for yourself how preattentive attributes like color help you instantly answer business questions. Ryan shares his personal exercise for demonstrating the power of data visualization and his answer to the question, “Why do we visualize data?”.

Hi, this is Ryan with Playfair Data TV, and in this video we’re going to be discussing why we visualize data in the first place. And I’m also going to be sharing my exercise for helping educate my end users on the value of visualizing data. Asking people why they visualize data is always one of my favorite questions to ask whenever I’m presenting, and the reason is I always hear something different. It kind of fascinates me.

But you have to remember that the practice of data visualization really is subjective to a certain degree. People are doing this for different reasons. So I’m always interested to hear why they are personally visualizing data. But I’ve also put a lot of thought into it myself, and if I had to boil it down to one sentence, I would say that we visualize data in order to find and communicate actionable insights in data.

Visualization helps us do those things, and I’m going to show you why with this exercise. What I do is I show this table. This cross tab. It doesn’t really matter which text you use, but this one happens to be looking at Profit by Sub-Category and Month in the Sample Superstore dataset. And I just pose this to the group, and I say– or whoever, whatever stakeholder I’m trying to help educate, and I’ll say let’s try to answer the most basic business question I can possibly think of.

I’ll say what is the highest number in this table. And I see how long it takes them to answer it. If they’re feeling brave they’ll call out a couple of numbers, usually we hear a few incorrect answers. And they might finally land on the answer 15,864.

And the reason that this takes so long for them to find is they’re having to go either row by row or column by column to try to find the highest value. Once they find the highest value, they’ll kind of commit that to their mind, and they’ll keep scanning, looking for the next highest number. If they find something higher, they replace it. Keep going, if they find something higher, they replace it. All the way to the end. They had to look through in this case 203 different numbers in order to answer that question, what is the highest number.

And if they do it too fast, by the way, I have a backup. I ask what’s the lowest number. But after we go through that exercise, I then convert this text table to highlight table on the next screen, and you’ll see the punchline right away. If we’re asking what’s the highest number, you may have narrowed that down to two values, 15,864 and 11,044. On the low end there might be a couple of more options, but even then you might be comparing three or four numbers instead of 203 numbers.

And this is the power of visualization. This is a pre-attentive attribute called color that helped us find and communicate the insight in this data. There’s about three dozen of these pre-attentive attributes, which we’ll get into more of them in separate videos, but they’re called pre-attentive attributes because we are so good at processing them that it happens almost subconsciously before we even pay attention to the view.

To take this a step further in Tableau we could even put a little bit of extra calculation on this to just spell out exactly what was the highest number and lowest number, if that truly was our business question. So we went from trying to answer the question in this raw text table of data to answering it within less than a second, here. When we had the pre-attentive attribute of color applied as well as a little bit of calculation just to tell us– or just to color the highest number and the lowest number.

This was inspired by Stephen Few’s famous count the nines example. I really liked that example. If you haven’t heard of it he shows a block of numbers, and he just asks the audience to count the number of nines in that block of numbers. It always takes a while, because you’re having to scan every row to count the number of nines.

In the second view he uses the same pre-attentive attribute of color to color the nines red, and he asks you to count them again. I really like that example. It illustrates the value of pre-attentive attributes and data visualization, but I prefer this example because this first view is what corporate reports look like in 80% of situations, in my observation.

And I find that when you do the exercise with a report that looks similar to what they’ve been using in their business, something kind of clicks. They think to themselves, oh, we’ve been looking at data like this for quite a while. And if we had just used pre-attentive attributes like color we would be able to understand our data a lot easier.

So for me a big part of the reason we visualize data is to find and communicate insights. That’s what this pre-attentive attribute of color allowed us to do. To make the analysis actionable we’re going to use design, and we’re going to talk about some other techniques in other videos. But as far as my answer on why we visualize data, it’s to find and communicate actionable insights.

This has been Ryan with Playfair Data TV – thanks for watching!