Thomas: Oh, hello there. You’re listening to episode 10 of season two of the Presentation Boss Podcast. Episode 10, Kate, that means we are 10% towards your goal of 100 episodes.

Kate: Yeah. Wow, that means we can start lining season three up for 2032.

Thomas: We could, yeah. How about s- let’s go the next, I don’t know, 10 episodes.

I was gonna say 90 but, but we’re 10% in. How are you feeling? Like, are we, are we back? Is this the podcast how it is?

Kate: I think so. Um, I have forgotten how much I enjoyed just talking to people that we admire, that we love. Um, just cool people and having cool conversations, so that has been- Mm … great. Um, video is hard.

Yes. Especially this.

Thomas: Yeah.

Kate: The ones on Zoom, not too bad, but I’m finding video just difficult.

Thomas: Yeah. It’s a lot more to think about. Yeah. My… I still love nerding out about good talks. Breakdowns are still my favorite episodes.

Kate: Yep.

Thomas: Um, but then also just the amount that podcasting has changed in the last, what we say, five, 10 years.

Yeah, yeah, yeah. Um, there’s a lot more technology and tools, it’s way more accessible, but my entire workflow, so I do, I do the production, I do all the editing, and I do- Mm … a good part of the publishing. That flow is all just different now. But I think-

Kate: Good different or bad different?

Thomas: Uh, just different. Yeah.

It’s just different. Some of the tools are easier. Video adds a whole new element, of course- Yeah … a whole new dimension to it. Uh, y- th- there’s less sort of cheeky edits you can make when it’s video.

Kate: Ooh, yeah,

Thomas: okay. Yeah. Rather than sort of chopping and changing an audio as I used to, you know, do sometimes just for, like, smoothness and flow.

But it’s more real this way, which I guess is the whole point of-

Kate: Yeah …

Thomas: video.

Kate: Yep. Yep, yep.

Thomas: Okay.

Kate: All right.

Thomas: Let’s get stuck into an episode. About data visualization, today we’re talking data viz.

Kate: Yeah, which always good to talk about. Uh, our favorite thing potent- potentially to talk about-

Thomas: Yeah …

Kate: is data viz. Um, and if you work with data, if you work with data at all, you probably create data viz.

You create charts and graphs and, and the whole gamut. And after you’ve spent hours staring at your screen, choosing your chart, and tweaking it, and making all the little, um, bits and pieces just right, it can be really hard to see that chart from someone else’s perspective and, like, how your audience will, because you already know what it means.

It’s the same as- Mm … you know, when you read something that you’ve written, and you will read what you think you’ve written even though you might have typos and wrong words in there.

Thomas: Yep, yep,

Kate: yep. Yeah, so and, and data viz is exactly the same. Um, you know what it should be, and that is the only thing that you will read in it.

So how do you create a chart that is not only obvious to you, but is completely unmistakable and obvious to your audience as well?

That’s our goal and the problem that we’re gonna tackle today.

Thomas: Hm. And I wanna start quite… Is it simple or basic or high level? The question is what makes a good chart? I asked this to all of our audiences- Mm … uh, like last week in that, that keynote. 120 people in the room, “What makes a good chart?” And people give all sorts of answers.

Yeah. And they’re good answers, usually good answers. Um, and I notice the sort of first place people go is they reach for rules, like these rules they have for themselves or have been maybe, uh, kicked around the office type of thing. Things like bar charts are good-

Kate: Mm-hmm …

Thomas: and pie charts are bad.

Kate: Yep.

Thomas: Uh, that kind of thing.

Kate: Yep, almost binary. This good, that bad. Yep.

Thomas: Yep, yeah.

Kate: Yeah, for sure. Um, there’s other rules like you shouldn’t use 3D and they should be colorful and attractive which, again, um, broad sweeping rules.

Thomas: Yeah, yeah. And sometimes as the conversation goes on I get more specific answers like we shouldn’t have too many categories.

Mm-hmm. And somebody will usually say, “I need to use the right chart for the data that I’m presenting” which, excellent, I agree. What does that mean? And these are really good ideas. They’re really good ideas. And so our definition for what is a good chart, a good chart is one that does the job that it was designed to do.

A good chart is one that does the job it was designed to do.

Kate: Nice.

Thomas: Yeah. And so different charts have different jobs. We know, um, from the data journey there are the two legs of data. One is explore where we take data and with an audience of one, which is ourselves, we use charts that are complex and messy and numerous and we try to find that insight, and that’s the explore charts.

And then when we need to explain that insight to an audience we’re going to use explain charts. So there’s two different broad jobs that a chart needs to or could need to do.

Kate: Yep.

Thomas: So we’re gonna focus on the explain charts, that explanatory step that sits in data storytelling where we’ve got an insight, we’ve got something worth communicating and now I need to hand that to an audience.

Kate: Mm-hmm

Thomas: These are the charts we use in data storytelling. And there’s three tests for what makes a good explain chart, the type of chart that takes that insight and gives it to an audience. Three, three tests. One, the chart needs to be able to be read quickly.

So realistically, can the audience who’s never seen this chart before, can they look at it and understand it in as little time as reasonably possible? A few seconds is kind of what we’re aiming for here. Yeah. Really quick. The second is easy, which means this chart needs to be read by your audience easily using as little cognitive effort as possible to understand it.

They don’t have to sort of stare at it and try and understand the axes and the colors and the… You know how it goes, right? Mm-hmm. Multiple, uh, multiple variables. And the third, which I think is the most important, which is unmistakable, which is it needs to be unmistakable in its, in its messaging, meaning it is as unlikely as possible for a different audience or a different person looking at this chart to pull away a message different to what you intended.

Kate: Yep.

Thomas: And so those are the three rules, quick, easy, unmistakable. That’s what a, a, an explain chart needs to achieve. And so charts that we see given, given out, if they’re not achieving these three things, they can be technically accurate, technically right, and have all the information there, but they can make the audience work way too hard to try and understand what’s going on.

Uh, or they can lead different audiences, different people to completely different or wrong interpretations.

Kate: Mm-hmm.

Speaker 6: oh, hello. It’s just me for a second. Kate’s ducked out. I wanna tell you about the online community that Kate and I have set up. It’s called the Influential Analyst Academy, and in that group is a whole bunch of analysts and technical professionals who are working on their communication skills, aside from just the, just the analytics bits, but how to do things like choose the right chart and have really good data visualizations that are fast, easy, and unmistakable to understand.

Heap of people in there working on a skill per month. This month, the skill has actually been data visualization, so there’s recordings and conversations that are in there and accessible. If you’ve enjoyed this episode, it could well be worth you checking out. So I’ll put a link, uh, in the description as well, so go and check out the Influential Analyst Academy.

Have a look. I think it’ll be right for you. Cheers.

Thomas: so this all starts with choosing the right chart

Kate: Mm.

So . the type of chart you would use to explore is ones that will give you multiple possibilities. It’s going to give you many different interpretations depending on what you’re looking at. Um, and then the ones that you use to explain are generally only going to give you one interpretation, which means they’re going to be simple.

They’re going to be-

Thomas: Mm …

Kate: um, basic charts. Um-

Thomas: Yeah. So let’s have a look at some graphs. Uh, I brought a prop from our workshop that we’re gonna have a look at. So if you’re watching, you can see them. If you’re listening, I will tell you what they are. These are just cards with different graph types on them, and we should probably talk- Yes

I think you were gonna say this.

Kate: Yes. The difference between a graph and a chart.

Thomas: Yeah

Kate: Because we do use them fairly interchangeably-

Thomas: Yeah …

Kate: it seems. Um- But

Thomas: there is a distinction …

Kate: yeah, we’ve got a fairly concrete- Yeah … separation between a graph and a chart.

Thomas: So what’s on the card here is a graph, which is, this is a bar graph.

It is the type of visualization. Mm. There’s no context. There’s no data. There’s no labels. There’s nothing here. A graph is a building block of the chart.

Kate: Yes. So your chart has the labels, the legend, the axes, the context, the-

Thomas: Title …

Kate: uh, all the things, yeah.

Thomas: Yeah. So we’re just looking at the graph type.

Kate: Yeah.

Thomas: So what we’re gonna do here is have a look at the explain type graphs. Mm-hmm. So this first one, Kate, is the…

Kate: Bar chart.

Thomas: Bar chart. This is definitely something you would use to explain.

Kate: This is honestly my favorite kind of chart.

It’s basic, it’s simple, but it’s my favorite.

Thomas: Yep.

Kate: Um, and then jumping quickly into the next one, which is the column chart. It is my pet niggle that people think a column chart and a bar chart are- … the same thing or interchangeable.

Thomas: All right.

Kate: It-

Thomas: So let’s talk about… So we’ve got the bar chart here where the bars run horizontally, and we use this type of graph for?

Kate: Categories. So categorical data.

Thomas: Yeah. ‘Cause in English we naturally read lists most important, least important, top to bottom. Categorical.

Kate: Yep. Definitely.

Thomas: And then we have the column chart where the columns are vertical, like columns would be- Up and down … in a building. Yep. And we use this for?

Kate: Time-based series.

Yeah. Although, I will say, I also really like a column chart when we do money because to me, in my head, money stacks up.

Thomas: Yep.

Kate: So, uh, there’s no hard and fast rule, but it makes the most sense to me. Yeah. Three part science, one part art, you’ve gotta make some decisions.

Thomas: Yeah. So- Yeah … columns, bars. Good. Next, uh, next graph card here is this one, which is just-

Kate: Is the simple line chart.

A line chart. With one line.

Thomas: Yeah. So this is where you have, uh, a, a relationship between each of the data points across a time series.

Kate: Mm-hmm.

Thomas: Simple, people understand it.

Kate: Nice. Yep.

Thomas: And the fourth, final graph, ex- explain graph type is this guy, the pie.

Kate: I like a pie.

Thomas: Ooh.

Kate: There’s, there’s two types of people in the world.

Thomas: There’s, there’s two types of people in the

Kate: world. There’s pie lovers and pie haters. Yeah. I like a pie. I’m- With caveats, of course.

Thomas: Yeah. So let’s have a talk about the best practice for a pie chart, which is we see them overused. I th- Yeah … I think that’s safe to say. These are overused, and what tends to

Kate: happen- And that’s why people hate them

Thomas: yeah, ’cause they get requested where they shouldn’t be.

Kate: Mm.

Thomas: So there’s two rules for when to use a pie chart. Number one is when there is two, maybe three categories. And, and there’s no getting around like, “Oh, what if I use a donut?” No, they’re the same thing. Donuts and pies are functionally the same. Two, maybe three categories.

Mm. This one’s got two categories. And the second rule, which is a little bit strange, but trust us, is when the number doesn’t matter, when you’re just looking for a vibe. So this one, it’s roughly one third blue, roughly two thirds gray. Does it matter if that is 33, 37, 29%? No. I can just see that it is roughly a third.

Yeah. And everybody can understand these charts.

Kate: Oh, absolutely. Like, my kids from when they were tiny, like my two-year-old would understand when he was getting a smaller piece of … cake than his sister. A

Thomas: smaller piece of pie even.

Kate: Yeah. Uh, yeah, it is intrinsic to us, uh, from a very young age.

Thomas: So that’s, those are the four chart types.

Mm. See if I can hold all of them at once here. The bar, column, line, pie.

Kate: Yeah.

Thomas: And you’re thinking like, “I’m listening to this, you know- Data visualization … data visualization episode and you’re saying to use these basic chart types.” Yeah.

Kate: Mm.

Thomas: Because these four chart types everybody understands quickly and easily, and with the right s- uh, the right, um, shrubbery around them when you make your chart unmistakable as well.

And I appreciate in the real world, uh, that data is messy, people are messy, culture happens, and you can’t always get to, like, a neat little bar chart.

Kate: Yeah.

Thomas: But we aim as simply as possible so that an audience can understand as quickly as possible. Let’s have a look at some of the graph types- Mm-hmm … that we see used as explain, but kind of don’t work.

Kate: Yep.

Thomas: Or don’t work sometimes.

Kate: All right. What do we got? What do we got?

Thomas: First one, here we go. This is the- Oh … stacked line chart. So just a bunch of lines. They’re all very pretty colors. Um-

Kate: Here it looks like we’ve got seven lines. We ran a master class, uh, last week, the week before. Mm. Um, but there was a lady in there, we were talking about line charts and she said, “Oh, I saw one used at our management meeting that had 100 lines.”

And we were like, “Ha ha ha. Yep, I’d love to see that.” Yeah, of course you do. Yeah. Um, and then she posted a, a screenshot in the chat, and she was correct. There was 100 lines. It was-

Thomas: A blur. It was

Kate: a blur … one of the most insane charts I’ve ever seen.

Thomas: It was Near impossible to pull any kind of meaning out of it

Kate: Yep Yeah?

Yep, yep. But we see it all the time. We see people- Yeah … use these complex-

Thomas: Stacked lines …

Kate: stacked lines. Yeah

Thomas: Now, a stacked line can be useful. It can be.

Kate: Mm.

Thomas: If I look at these, we’ve got seven different colors, and generally speaking, the lines are going up. So my question is always, can I simplify a stacked line down to a single line? Mm. Can I simplify that down? Is that the message? Or if I have to use different stacks, we need to do some extra work when we convert this to a chart to highlight the line that we’re looking at. But generally speaking, when we see this stacked line, it’s pretty colors, and there’s between 7 and 10, Mm

oh, sorry, 7 and 100 lines. It’s asking our audience- Yeah … to do too much work.

Kate: And I think on another episode, let’s go through this and exactly how you can work with this to make it- Yep … um, to make it unmistakable. But, uh, we don’t have the time to go deep on that today.

Thomas: Here I have the heat map.

Kate: Oh, the heat map

Thomas: Now we wouldn’t put this in front of an audience because it’s a lot of work to look at.

Yeah. And in this one, if I hold it far enough from the camera or what have you, you can see there’s vaguely a red patch here, there’s vaguely a green patch here. I could, as the analyst, as the audience of one, maybe start to interrogate exactly what’s going on. But putting this in front of an audience and giving them just a few seconds, it’s-

Kate: It’s really hard.

Thomas: Mm.

Kate: And I think people, um… they use this with the best of intent because it is i- maybe the most visual of them because- Yep … it’s such a block of color. Yep. But there is a lot to separate. You’ve got to separate blue, uh, nope, this is green. Green from yellow from orange from red. They’re four quite- Yeah

distinct colors, and you’ve got to kind of process all of those colors. Mm-hmm. Look, I don’t want to say that these are, like, great to put in front of people, but if you need to, the best way to use a heat map is to go from light to dark, as a general rule. Mm. Um, still, if you can possibly avoid it and put more explanation around it, do, but if you have to use a heat map, light to dark, whether that’s light blue to dark blue or whatever, whatever color is going to best represent-

Thomas: Mm

Kate: um, your company or that message.

Thomas: Still a lot of work, still a lot of processing. Yep. So give people time. Yep, yep. Okay. What have we got next? Oh, we have the- Oh … stacked column. So this one is our columns, but there’s one, two, three, four, five different colors in each column that all categorize out different.

Kate: This is the most controversial of them all. Yeah. When we ask in workshops-

Thomas: Yeah …

Kate: um, we ask people to separate them, separate these charts into explore, explain, this always ends up on the line.

Thomas: On the line, yeah. So the rule, the rule I have for this… Sorry, the rule I have for all charts is charts are just like dropped food.

So you know the kind of rule where if you drop food you’ve got a few seconds before you just go, “Ugh”? It’s the

Kate: five-second rule. It’s, it’s science. It’s fact.

Thomas: Okay. Thank you, Dr. Kate. Uh, look, you’ve got those few seconds before you just go- Yep … “Ugh” and you throw it out.

Kate: Yep,

Thomas: yep. Charts are the same. An audience has a few seconds before mentally they just go, “Ugh”, and give up and throw it out.

Kate: Yep.

Thomas: This chart is a really good demonstration of that. If you are 100% sure that your audience is going to understand this in a few seconds, please go for it.

Kate: Mm.

Thomas: Most of the time, though, an audience isn’t- Mm … and I would say can you simplify it down to, let me grab the just normal column chart.

Kate: The column.

And I think, um, again, we’ll go through in another episode about how you can use a stacked column and get it more towards a plain column, um, as much as possible. I really, I don’t love a column, a stacked column being used- In general

Thomas: Yeah …

Kate: w- it needs a lot of work before it’s presented to an audience. Even if you’re like, “Oh, I’m so sure that they will understand this,” I think it needs a lot of work before it’s actually- Yeah

presented.

Thomas: And if you want the test, uh, there’s, there’s a light, there’s a light blue color here. Um, it’s the middle category. Mm. My question is, what is that, what is that category doing? Is it getting bigger, smaller? It’s really hard to compare. Mm. This takes a lot of cognitive effort. It is not easy to understand for most audiences.

Kate: Yeah. If possible, avoid the stacked column.

Thomas: Yeah.

Kate: Think, um, yeah, I see this one a lot and I would love to… This would be the one that I would love to delete from people’s reports- Mm … and presentations and dashboards.

Thomas: Yeah. We see this a lot with clients. People bring a stacked column and you put a lot of work into- Mm

breaking down categories and what is the actual message and how do we simplify?

All right, stacked column. We’ve kind of beaten that one. Uh, let’s move on to this, the bubble plot.

Kate: Oh, the bubble. Yep.

Thomas: Now, the bubble plot, there’s a lot going on, and if we have a look at it, what’s happening is we have the X and the Y axis.

We have the color of each bubble, and also the size of each bubble. Mm. Which is four variables, which is two more variables than you really want in an explain chart. Yep. It is so much to process, so it takes so much time or mental effort or both to understand what’s going on with this relationship type chart.

Kate: Yep. Yep, yep. This is overwhelming. I, I would like to never see one of these in a report, in a presentation.

Thomas: I’ve seen one bubble pl- plot that I like and I’m not gonna talk about it today. We’ll come to it.

Kate: Yeah, I think I know what you mean. We will talk about it.

Thomas: Next, last, we have the scatter plot.

Kate: Oh, the scatter.

Yep.

Thomas: Yep. Where we just have dots across the X and Y. It looks like a cloud.

Kate: This is a straight up analytics chart.

Thomas: Yep.

Kate: Yep. The end. End of discussion.

Thomas: Yeah. You could fiddle around and put like quadrants in it- Mm … and put things in regions, but again, it still takes a lot of effort. Okay. So that leaves us with where possible you’re going to push towards the bar, the column, the line, or a pie chart.

Like we said before, sometimes this might feel simplistic, but these are the chart types everybody understands and where possible we push towards the simplest type of chart so it’s quick, easy, unmistakable, and has just two variables.

Kate: all righty, so that is the basics of choose your chart- Mm … and how to select the best chart, the most appropriate chart for your data story. Again, we wanna just be pushing towards the simple explain charts as much as possible so that people can understand them quickly and simply and unmistakably.

If you go for a more complex graph, something that is more in the explore category, then you really run the risk of your audience, um, doing the analytical work for you and coming to a different conclusion to what- Mm … you have come, come to. And we really want them to be of the same understanding of you, as you, of your data.

Thomas: Yeah, and we, we use these chart types, these simpler chart types before- Mm … because they are clear rather than complex. Mm. We’re chasing clarity over, like, complexity or cleverness or adding in more context. And sometimes, I know, it can be a request of us to, can we add that bit of data? Mm. Can we add that axis?

Can we add something? And w- that request comes when there’s a lack of understanding around- Mm … exactly what this chart says, but instead adding that complexity pushes us away from what we’re trying to achieve, which is understanding, which is clarity. Um, I should say, uh, choose your chart, having the right graph type is the first step in a four-step process that we teach everybody who stands still long enough.

Uh, certainly all of our clients. First step in the StoryChart process, it is, uh… We’ll get to the other three steps in other episodes.

Kate: Which is only gonna be one more episode to go through the final three steps. Mm. Because choose your chart is the biggest one. It’s maybe not the hardest one, but it’s maybe the most important to get around, to get your head around.

Thomas: Yeah, yeah, and it comes with, uh, props from the workshop. Um, what I will do is I’ll put a link to a download to a summary sheet- Mm … of which of these graph types are the ones to use- Mm … which are the ones to avoid, and exactly when and why. I’ll put a link to that, that you can download in the show notes.

Kate: Yeah. All right. Thank you so much for being with us today. Head to presentationboss.com.au/podcast where you will find the resources, the downloads, um, the links from today’s show.

Thomas: For sure. If you have any questions about specific presentation or data storytelling skills that you would like us to discuss here in this corner of the office, uh, please get in touch.

Email podcast@presentationboss.com.au. We are always happy to hear from you, hear your thoughts and requests.

Kate: Yeah. And of course, if you liked this episode, if you found value, please recommend us, recommend us to a friend.

Thomas: Lovely. Have a great week

​

Thomas: Mm. And we will talk about that in the next discussion episode.

Kate: We will. This is a big piece. The, um, the other three steps

Thomas: Who’s playing music? That’s in this house dude.

Kate: If I was home alone, I would be freaking out right now

Thomas: I’m just gonna pause this.