It's not just the sunk cost issue, though that is bad enough. It's the overhead cost issue as well.
Say you've got a product... to pick an extreme case... that costs $100,000 a pop. You have to have a high-priced sales staff to support it, and a high priced development team. To house them all, you've got fancy offices. And to manage everything, you've got high priced CEOs. Your high end customers wouldn't want anything less. Remember the old saying? "You can't get fired for buying IBM."
Then you've got another product, that costs $10 a month, sold by world-of-mouth and run by some guy and his rented servers out of his mom's basement -- to pick something on the very other extreme.
And there are only a few customers for that $10 product.
What possible business sense does it make to throw out your $100,000 offering and sell your customers the $10 thing instead? Even if you decided to ignore the money you spent in developing it, you'd have to lay off your entire sales force, all your senior management, sell your office building, and move into YOUR mom's basement to get costs low enough that you can afford to do it.
So either you say, like Apple did, damn the costs, we're moving ahead anyway, and we'll figure out a way to make money with it later.
Or you start up an arms-length skunkworks -- like Hulu.com -- and then let them compete with you.
And they'll start taking away your customers -- $10, even if for a lesser product, sounds a lot better than $100,000. They'll start with the small customers, and move up. And then your existing customers will suddenly decide that they don't need all your bells and whistles anymore -- or the startup guy will add those bells and whistles -- and the old company will be out of business.
Remember the days of enterprise-grade CMS systems? A company I worked at spent around $1 million on Vignette. Today, Drupal and Wordpress do a lot more -- for free. This blog here, for example, is run on Wordpress, using a free template, on a low-cost hosting company and is easier to use and has more functionality than a million-dollar platform offered 12 years ago.
Which reminds me -- there's probably going to a a period, of about ten years give or take a few, where you're going to be able to sell million-dollar 3D content management systems before the bottom falls out of that market. If I had any programming skills, I'd be writing that CMS now!
