–Dan Kelly, Attorney

A long time ago, I heard an anecdote in which one person asked another, “Why can’t we just mandate that air travel be 99.9% safe?” The response: “If air travel were only 99.9% safe, there would be three accidents a day at O’Hare alone.”

I thought of this anecdote a few days ago when I saw a box of Hammermill paper at the office with a claim on the side: “99.99% Jam Free.” (Turns out, International Paper, owner of the Hammermill brand, views the phrase “99.99% Jam Free” as a trademark.) 99.99% jam free means that no more than one out of every 10,000 sheets would get jammed. That’s one sheet out of every two boxes, or one sheet out of every twenty reams. I was curious to know these numbers to see how realistic this claim is. I’ve never tracked this stat, so I don’t really know, but it seems to me that when the copier jams, it does so repeatedly, and then someone eventually calls for service.  More often than not, the technician does something to the copier, not the paper. I can’t imagine how to test this claim when it may be difficult to determine whether the paper or the machine causes the jam (or even whether the paper causes some sort of wear on the machine that causes the jam). International Paper backs the claim with a guarantee, so I suppose this is their problem, not mine.

But it is probably safe to say that the 99.99% number was chosen more for aesthetics than for particular accuracy. Giving International Paper the benefit of the doubt that there is data to support the number, it is probably not 99.99% on the nose. Even if testing were to support a 99.999% number (only one sheet out of every 100,000 sheets would jam), 99.999% looks a little unbalanced, perhaps incredible. Numbers in between 99.99% and 99.999% would look similarly strange: “99.994% Jam Free” doesn’t really look convincing, even if it is accurate.

But as I’ve shown before, numbers, like words, do matter.  From a legal standpoint, they need to be accurate.  From a marketing standpoint, they should look good, in every sense.