The Black Magic of Predicting the Future

When Apple dropped the price of the iPhone by $200 so soon after launch, I tended to be forgiving (at least once they made it up to their early adopters). As I said over at Lewis' blog, I understand the difficulty anyone has in predicting the future -- especially when a product changes the way the game is played like the iPhone.

AdAge's video podcast of 9/5/2007 (subscribe) brings up an excellent example of what I mean, and they do so by using a similarly innovative product -- the DVR. In 2003 Forrester predicted that by 2007, DVR penetration would be at 25.7% in the U.S. However, now that it's 2007 and we have real data to measure against, we know that penetration is only 17%.

If you're an advertiser, don't overestimate what that means, though. As before, people are still ignoring your TV ads. It just takes them longer because they're ignoring them by means other than a DVR.

In general, the further out you try to project, the less accurate your prediction will be, because you simply don't know and can't predict how people will react in the real world... how they will be influenced once they see how people react to the new product.

I know there are very clever mathematicians out there who claim they can predict to within plus or minus three percentage points with 95% confidence what people will do, but that is only assuming that they have all the relevant information. When a product is much better than anything that has come before, it's less likely that statisticians will have all of the relevant information. - Cam Beck

Comments

[Lewis Green]

Cam,

Nice post. Predicting is only slightly more scientific than throwing darts. However, in iPod's case, there price change occurred within months of launching the product. That's just wrong.

Have a great weekend Cam and thanks for the shout out.