Selling software has its upsides -- and downsides.

Look no further than Microsoft (NASDAQ:MSFT) or Google (NASDAQ:GOOG) for proof positive that the fat margins inherent in not building and shipping a physical product can make a lot of money. That's the upside.

An article in the current issue of Forbes points out the downside, calling it the Netflix (NASDAQ:NFLX) Law. The million-dollar Netflix Prize demonstrated with painful clarity that it is very, very hard to make software measurably and significantly better than what's already available. It took three years for a worldwide competition with a big, fat cash incentive to shake out a 10% improvement over Netflix' own movie recommendations system. Even then, the winner was a consortium of leading individual players who banded together to squeeze every last drop of efficiency out of their combined mental efforts. It seems like early software efforts get it mostly right, most of the time.

So chip designer Intel (NASDAQ:INTC) may be able to keep up with the "double performance every two years" maxim called Moore's Law for the foreseeable future. But software outfits like Netflix, Google, or Oracle (NASDAQ:ORCL) would count themselves lucky to move at even a fraction of that improvement pace.

This means that the business models for hardware and software must be fundamentally different. In hardware, you can always hope to invent your way out of any jam and keep the growth engines running. In software, you need to get an early lead and then hang on for dear life while the competition tries to catch up to your methods.

Netflix itself has proven to be remarkably good at doing exactly that, beating back all comers for years on the strength of a business and software framework that is incredibly tough to improve upon. You could argue that Microsoft is another example of this principle, though the Apple (NASDAQ:AAPL) camp might protest that their company sure has out-designed Microsoft by a wide margin in recent years. Apple's reward for that innovative tenacity? A growing share of the personal computer market and total domination in portable media players. When you beat the odds, the payoff at the end of the rainbow can be remarkably rich.

And then you have the likes of IBM (NYSE:IBM), which is straddling the line between software and hardware. Microsoft is moving in that direction with hardware efforts like the Zune. All three approaches seem valid to me, but for very different types of investors. Me, I'm betting on Netflix defending its hard-won moat well into the all-digital entertainment age.

Where do you see the greatest opportunities in technology investments -- software, hardware, or somewhere in between? Let us know in the comments below.