Data slides: an example of what not to do
Let’s look at another example that’s even more typical and problematic. Here’s the slide as I saw it in a presentation in early 2009.
Keep in mind that this was the real deal—real slide, real presentation and the person had a real reason to be doing the presentation. This slide fails so utterly it’s hard to know where to begin, but I’ll just start with a list of simple flaws.
- Anybody want to guess what the units are on vertical axis? Or what the numbers are measuring?
- The 3D effect is again just decoration and should be removed.
- If you’re going to label every bar with a number why bother with the axis on the left at all?
Now for the more serious issues here.
- That starburst is silly decoration, but the text inside it is the real problem. We don’t know what “emerging markets are” but if we assume it’s the 1st three bars, then the real growth is over 3X to almost 5X.
- How about that label on the right with the (unnecessary) circle? The starburst compares in multiples, this one in percentage—pick one and stick with it.
- And if the 66% growth for Asia-Pacific is important, then why isn’t the 95% growth in Western Europe worthy of attention?
But the most serious issue here is that there’s no way to figure out just what the point of this data slide really is because it’s the wrong format. The visual appearance is an increasing set of values left-to-right, which we immediately expect to be a timeline. It’s not a timeline and the left-to-right slope is completely irrelevant, a serious distraction in fact.
A data slide like this should be horizontal bars so the brain isn’t fooled into thinking there’s time series data on the chart. If growth is the real point to the slide, then the sales numbers aren’t needed—a growth number is what’s needed either in percentage growth or numeric value of the change over the period. That presents a more subtle issue: is the percent growth what’s important which will emphasize relative growth, but might hide where the real growth in units is, or vice versa?
Here’s the same data but now graphed properly showing percent growth:
So, Latin America is what’s important? Well maybe and maybe not, here’s the same thing, but we’ll graph the actual growth in units over the time period:
Now it’s clear that the real growth in sales volume is Asia-Pacific.
So which of those is “right”? Both or neither, depending on what the presenter wants the point of the slide to actually be—something we already figured out was completely obscured by the original. At least both of the reformatted versions are true to the data, free of decoration and readily understood.
The bad news
Obviously, as originally presented, this data slide fails all five rules miserably. In looking at this example we can see how it illustrates several ways to really confuse an audience and fail to get your point across. It also shows us that the same data can often be presented multiple ways and tell multiple stories depending on how it’s formatted, so it’s important to pick the format that suits the point you’re making.
The good news
The mechanics of fixing this chart aren’t hard—the hard part is knowing how badly it needs fixing. The underlying failure with the chart isn’t that the presenter didn’t have good data, the failure is that they didn’t use it to make a point the audience could take with them. Too bad, because the data does have some good stories in it—I wonder which one(s) they wanted to tell?
Don’t make this mistake.
- Know your data and what story point you’re making with it.
- Follow the Five Data Slide Rules.