Thursday, December 13, 2007

November NPD Numbers

Wii: 981,000
360: 770,000
PS2: 496,000
PS3: 466,000

This is an easy analysis, so it's going to be short. If you look at historical data for the PS2, Xbox, and 360, there's an easy rule: September sales will equal the sales of the following January, plus or minus 15%.

That's why when companies crow about their October-December numbers, we need to be cautious, because those gains really don't carry through to the following year that often.

Here are the years that the rule is correct (year refers to January sales, September sales would be previous year:
PS2--2004, 2006, 2007
Xbox--2003, 2004, 2005

In 2003, the PS2 tracked within 20%. In 2005, sales in January were over 2X what they had been in September of the previous year.

Xbox numbers didn't track in January 2006 because the 360 was released in November of 2005, and Xbox sales fell of a cliff (unlike the PS2, because Sony wasn't selling enough PS3s to put a dent in the PS2 market that soon).

In case you're wondering, the Gamecube didn't track like that--it was all over the map. Nintendo appears to be a special case in general.

As a general rule for Microsoft and Sony, though, I think the +/- 15% expectation is reasonable, unless there are exceptional circumstances. For instance, with the launch of Halo 3 in September of this year, I don't expect 360 sales to be in the 500k range in January (but I do expect ongoing sales in the 230k-250k range).

However, I think it does give us some good information about what we can expect from the PS3.

Here's what we know. I mentioned in last month's analysis that November sales in the U.S. are usually 2.5x the sales in October. Since October sales for the PS3 were 121,000 units, sales in November--even without a price cut--would have been expected to be roughly 300,000 units.

So we're talking about 166,000 excess units in sales that can be attributed to the price cut, not the 345,000 units that Sony is going to claim.

Divide November sales by 2.5 and you get 186,400 units. That's theoretically what October sales would have been at the $399 price point (without discounting unit numbers for the initial price cut buying spike).

So we have 186,400 units in October, and we know that September sales will equal January, plus or minus 15%. So here's the last question: how do October sales usually compare with September?

The answer is "roughly equal"--some years October sales go up, some years they're down, and some years they're the same (I'm guessing it's heavily dependent on the schedule of the big game releases).

So, at the end of the rabbit hole, I think it's reasonable to estimate January sales for the PS3 at 186,000 units, plus or minus 15%.

And for Sony, that's a problem. Yes, 186,000 units a month is a huge boost for them, but that will leave them a distant third in the U.S. market. The 360 outsold the PS3 by 304,000 units in November--that's crushing. And that's with a price cut for the PS3 and all the initial hoopla and promotion.

Certainly, it's positive for Sony that the price cut raised sales 50% over what they would have been expected to sell otherwise--but that 50% increase is from a base so dismal that it was almost irrelevant.

So when you see press releases from Sony talking about how brilliant they are and how great sales were in November, remember holiday sales inflation and what those numbers really mean for the future.

Site Meter