A Graph Done Right

An infographic called Unboxing the iPad Data has been floating around since the iPad debuted in early April. It’s well done, for the most part, but one portion of it bugged me. Specifically, this:

This graph is, essentially, meaningless. According to the caption, the iPad sold an estimated 300,000 units “this weekend,” which is a vague timeframe (two days? three days? according to the graph it’s one) that includes preorders. Meanwhile, the iPhone sold 1 million units in “over 70 days.” Which means there’s no commonality between the variables. Which means it might as well be comparing apples to lightning strikes, let alone oranges, in terms of being able to draw any conclusions. I wondered if the size of the half-circles might indicate some sort of accurate info, but their scale (diameters 190 vs. 270) relative to each other is meaningless.* As it stands, it’s a pretty picture drawn to make a couple numbers more interesting to look at.

For this chart to mean anything at all, we need some frame of reference. In order to compare the sales numbers on a chart or graph in a meaningful way, we need to standardize either according to number of units sold or number of days elapsed — or both. This chart does none of the above.

Fortunately, Apple announced today that it had reached 1 million iPads sold, which allowed me to standardize the chart. Sticking with the half-circle of the original, I redrew it to reflect the common total of 1 million units sold. I don’t have Trade Gothic on this computer, so I used Eurostile for that cool computational look.

This still isn’t all that useful of a chart. Growth over time is what we’re really interested in, right? How quickly are those iPads moving off the shelves? To show that, we can’t just do a bubble, we need to do a good old fashioned graph, with units sold on one axis and time on the other. (This time I’ve done it Helvetica.)

To make this graph truly informative, I’d really need to have more than one data point, but this at least gives you a better picture of the trajectory of iPad sales versus iPhone sales, something the original infographic didn’t do at all. I’ve also included that 300,000 in the first weekend as a faint overlay, with an assumed 3-day timeframe. That starts to give you a sense of the sales arc — fast off the line and then a presumably steady climb the next 70 percent.

Interestingly, looking back at the original iPhone, it may have taken AT&T reported 146,000 iPhone activations in the first day and a half. So if we wanted to compare first weekend sales, the iPad is apparently twice as successful an initial launch. The iPhone 3G, on the other hand, sold a million units in a week.

*If you take the combined bases of the graph’s half-circle as a total of 1 million, their proportions would be awfully close to correct — the iPad’s circle is roughly 40 percent of the total. Still off by 10 percent, though. Close but no cigar.

⊕ wsft.us/200