TECHNOLOGY LAW CORNER
The Folly of Predicting New Products From Apple's Patents

As a user of Apple (Nasdaq: AAPL) products, I relish the mystery of what's coming next, eagerly consuming the rumors and speculation dished out by the blogs and Web sites.

As a patent lawyer, however, I can't help but to eye with skepticism each new patent application served up by an Apple watcher as potential evidence of the next big thing.

A published patent, by itself, is an unreliable predictor of a new product or feature.

Patent, Patent, Patent

It's not that patents fail to give an accurate picture of what Apple works on. Apple appears to be aggressively patenting all of its innovations. The latest figures indicate that the annual rate at which it has been filing patent applications has roughly doubled each year since about 2003.

The number of published patent applications has grown from 64 to 405 from 2004 to 2007. Granted patents, which lag applications by a number of years due to the time it takes to examine them, went from 0 in 2004 to 185 last year.

Furthermore, if one looks at the classifications of the patents, Apple's patents cover a wide range of technologies. Indeed, no one area dominates. Roughly 14 percent of the inventions relate generally to computer graphics, and approximately 12 percent to "data processing."

Five percent of the patents involve recording, communication and information-retrieval equipment, and an equal percentage relate to database and file-management and data structures. The rest of the patents are in a fairly wide smattering of technical areas.

It therefore seems reasonable to conclude that anything new from Apple is likely to end up in a patent application.

Bad Omens

Why, then, would a patent not be a good predictor of a future product? There are several reasons, but the primary one is that by the time the patent application is public, so, too, will any product implementing it.

Maybe I am giving Apple's development and manufacturing prowess too much credit -- but if an invention has not yet appeared in a product by the time the patent application on the invention is published, there's a good chance that Apple has dropped the invention or significantly altered course.

New patent applications are kept secret for 18 months, at which time they are published unless the applicant states that it has no intention of filing a patent application on the same invention in any other country in which it will be published and requests that it not be published.

Furthermore, the invention will likely have been made at least several months before application. It takes time to decide whether to patent something and to write the detailed description of the invention that is necessary for obtaining a patent.

Therefore, depending on Apple's internal patenting processes, patent applications represent at least an 18-to-24 month trailing window into Apple's cubicles and labs. They are old news.

Apple's patents are, in many cases, probably also capturing ideas early in the development process. In an aggressive patenting program, there is a natural bias toward identifying candidates for patents early. A patent application represents the state of thinking at the time of application. Once filed, new ideas cannot be added to a patent application.

It is, of course, normal for substantial changes to occur during development. Ideas that seemed good at first sometimes prove to be too expensive, or too difficult to implement, or just not that good. Markets and business strategies change -- and, of course, there is the Steve factor. Therefore, one should expect at least some of the ideas contained in patents to never appear in a product.

Patents may also describe things Apple might have no intention of making. Patents frequently include alternative, less desirable implementations, and mention extensions or additional applications of the basic idea that do not figure into immediate plans. There are a number of reasons for including this additional material, but they basically boil down to broadening the scope of a patent and making it harder to design around.

Looking Deeper Into the Patent Oracle

On the other hand, even two years may not be enough time to develop a product or to fit its introduction into Apple's carefully crafted marketing plans.

Take for example U.S. Patent Application No. 2009/0066648, published March 12, 2009, which I've picked somewhat randomly from among Apple's most recently published applications.

The application describes using an Apple remote control outfitted with motion sensors to navigate the menus of an Apple TV or Front Row. Think of using a Wii controller to make gestures to flip through screens rather than one's finger on a touchscreen, or arrow keys and the center button on an Apple remote. The application's priority date suggests Apple first worked on it at least almost two years ago.

As of Apple's World Wide Developers Conference in 2009, we have not seen it. Is this a feature that has been discarded or simply delayed?

On the one hand, it is part of a theme running through numerous patent applications that I have seen relating to gesture control, which have regularly turned up in products. Incorporating remote gesture control into Apple TV -- or any computer, for that matter -- seems to me to make a lot of sense.

Based on the description in the patent application, I would say that a fair amount of thought had been given to how it would be implemented. The nature of the illustrations suggests a prototype may have even been developed.

Given how infrequently Apple TV receives updates, its failure to debut in a product may just be a timing issue. On the other hand, perhaps Apple has decided that there are even better ways of remotely navigating a media center interface.

Beyond Mere Curiosity

Although a particular patent is unlikely to give me any guidance on whether I should buy now or wait a few months, Apple's patents, taken collectively, might provide useful business intelligence to competitors.

Inventor information not only identifies key employees, but also -- when combined with a careful analysis of the subject matter of the patent -- gives insight on how employees are deployed, how product development teams are formed, and possibly how they work together.

Apparently discarded inventions, or inventions with substantially different commercial embodiments, may give insight into Apple's development process. Deeper analysis may uncover recurring ideas, such as gesture control, being applied in different contexts.

Trends in development of new ideas, even if they do not appear in products, reveal Apple's interests and strategies. This information can be used by competitors as input to their own strategies, and may even, with some extra thinking, lead them to file blocking patents that will give them leverage should they ever need to negotiate with Apple for a patent license.

Though it might not predict the next product, a thorough analysis of Apple's patents could nevertheless give competitors a boost toward getting in front of the curve rather than remaining perpetually behind it.