Charts can mislead

In most cases, bad charts are accidental. Perhaps someone was over-eager to show good results, or maybe, just did a sloppy job of formatting. Whatever the cause, it’s bad mojo to put together analyses or charts that mislead. Here are some bad examples from Starbucks’ recently investor conference. You can see all the slides here.  After some time, you will see many misleading graphs.

#1. No axis label

This is no-no. A chart without a labeled X&Y axis is like a car without an odometer. Not a good idea.

Bad graph - no axis

#2. Not drawn to scale 

Below, you can see that Starbucks is comparing its revenue with operating income. Since the scale is different, the operating income actually looks bigger than revenue. That ain’t right.

Bad graph - Not drawn to scale

3. False or unnecessary comparisons 

Same problem below. The coffee executives compare US & Canada & Latin America but use different – seemingly random – scales. Misleading graphs, I tell you.

The 2012 revenues are shown with red dotted lines. When you line them up, they look similar in size – when in reality – it should look like the graph at the bottom right. The US is where the current revenues come from and Latin America is a rounding error.

Bad graph - inaccurate comparisons

Why make the comparison? 

There was no reason to compare the US, Canada and Latin America.  They are at different stages of their growth. Why force the comparison?

Comparing the US and Canada makes sense

They are both mature markets with similar GDP per headcount and analogous cultures. Taking the population for the US and Canada here, you can see that the average American and Canadian spends about the same on Starbucks annually.

Starbucks per American

Analyze Latin America by itself or against emerging markets 

Looking at Latin America over the last 3 years, looks like they went from $76M to $115M to $143M. Nothing shabby about that. Why not focus on that story separately.

Clarity is your job

At the core, a consultant’s job is to drive clarity – through the data, analysis, and presentation. Anything you do to over-simplify, obfuscate, or muddle the issue is bad.  The client can be confused by themselves – without paying your fees.

 

Related posts:

Share This