Macopinion

Marketshare Myths Part 1

by Marc Zeedar macopinion@designwrite.com

Apple's marketshare is on a huge upswing in the past year, due to the iPod halo and the fact that Macs now use Intel processors. I'm delighted by all the good press, but one thing consistently rankles me: I keep seeing backhanded compliments and jabs at Apple's "low" marketshare. Journalists will write something like, "Apple's marketshare has increased, but..."

Or like this quote from a recent article:

"Macintosh computers have long had proponents among business users, such as those in the art department, for instance. But in terms of pure numbers, Macs barely make the charts."

Arrgh! That kind of dweeby reporting drives me nuts.

It's not that it's inaccurate in terms of specifics -- in lawyer-speak the percentage of Macs sold is a fraction of PCs -- but it's inaccurate in terms of informative information. By "informative information" I mean details that actually matter to real people reading the article.

In other words, what is the reporter trying to say? We can presume he or she is attempting to provide some sort of information. But what, exactly, is the information, and was that properly conveyed?

If the intent is to bash Apple, that's one thing, but if the intent is to balanced journalism (which should be every objective reporter's goal), the above statistic does not qualify.

Statistics alone are meaningless. Statistics only have meaning in context, in relationship with other data.

Just because I can reasonably prove that 92% of people who died of cancer in the last year had eaten pickles during that time does not prove that pickles cause cancer.

In the same way, saying Apple has a "3% global marketshare" is meaningless. It doesn't prove anything.

You can't even tell if the statement is positive or negative without adding your own perception to it. Most people would assume that 3% sounds small -- especially if the number on the other side is 97%.

But if we put that number into automobile sales, 3% is actually a huge chunk of change. If we subdivide automobile sales to only include luxury cars, the percentage is even more significant.

So having a percentage of a market -- any percentage of any market -- may or may not be good. There's literally know way to know without knowing more about that market!

But most reporters don't bother to explain it like that. They don't put the marketshare numbers in context. In fact, most imply that the statistic is a negative. They do this subtly, by adding words like "only" or "just" before the amount. This leads readers to assume that the number is low and bad, when the reality might be completely different.

That's what bothers me about reporting like that. I have no problem with reporting facts or statistics when they are reported in context. But it infuriates me when numbers are thrown out and implied to be negative with no information to back up that connotation.

Let's just look at one simple aspect of this situation. Continuing with the above example, articles about Apple will often state Apple's marketshare. (This will be included regardless of the actual topic of the article.) But do you ever see the marketshare of other computer companies listed for comparison?

Think about it. When you see Ford's marketshare listed, the number only has value when compared to GM or Toyota or other car makers. If Ford has 10% when Chrysler has 40%, then you know that Ford is selling a lot less cars than Chrysler. But if all the makers are holding similar percentages of the market, you know they are selling similar amounts of vehicles.

The same should hold true for Apple: Apple's market percentage is really only relevant compared to HP and Dell and other computer makers.

But we never see that. Instead, Apple is on one side of the line and all the other computer makers in the world are on the other.

That's why we see things like 3% against 97%, which really makes Apple look tiny.

In reality, of course, we should be comparing the marketshare of individual companies:

Global PC Marketshare (Q1 2007)
HP	17.6%
Dell	13.9%
Acer	6.8%
Lenovo	6.3%
Toshiba	4.1%
Apple	3.0% (estimated)

See what a difference this makes to the interpretation of "facts?" With this chart we see that Apple's computer sales are on par with Toshiba. Yet does anyone complain that Toshiba "only" has 4% marketshare?

Yes, HP and Dell sell five or six times as many computers as Apple, and those are significant differences, but 3% compared to 18% is far better than the 3% to 97% you usually see bandied about. If we're going to use marketshare as a dividing factor, let's at least use actual comparative numbers!

But we're dealing with statistics, so of course, the story doesn't end there. The above is a chart for global marketshare, which includes millions of low-end computers in foreign markets Apple doesn't even want to compete in. Including those in the comparison is like looking in the "under $10,000 car" market for BMW sales and concluding that obviously BWM is failing because they sell so few cheap cars!

We can look at just the U.S. market to get a slightly more accurate picture of things, since the American market is not only where Apple sells the most products but also tends to support the higher-end products Apple specializes in:

U.S. PC Marketshare (Q1 2007)
Dell	27.9%
HP	25.7%
Gateway	7.7%
Toshiba	5.4%
Apple	5.0%

These percentages give us more accurate information, but what does it tell us? The above "facts" tell us percentages of new computers sold in the last quarter, but it gives us no information about which computers. For instance, a lot of companies sell cheap $300 PCs with barely a hint of a margin. Ten $300 PCs with a 5% margin means only $150 total profit. Apple, in comparison, might sell a single Mac for $1500 with a 20% margin for a profit of $300... for just one sale! (And Apple's margins are as high as 28% on some products.)

So maybe Dell does sell five times as many computers as Apple. But are those sales as profitable?

Next Time: Reading between the lines

by Marc Zeedar macopinion@designwrite.com