I was reading Geoffrey Moore’s excellent book Crossing the Chasm the other day and it describes the situation between Windows and Linux exactly.
The section on competitive positioning shows that early in a technology products' life the focus is, appropriately enough, product-centric. This is good for getting the techno-enthusiasts and early visionaries to buy into your product. To cross the chasm and get a foothold in the lucrative pragmatist market, the focus has to shift to a market-centric view. The following table is straight from the book (page 137):
| Product-Centric |
Market-Centric |
| Fastest product |
Largest installed base |
| Easiest to use |
Most third party supporters |
| Elegant architecture |
De facto standard |
| Product price |
Cost of ownership |
| Unique functionality |
Quality of support |
Probably the easiest two to compare are product price and cost of ownership. Supporters of Linux like to brag that the software itself is free. What I have not seen is the cost of supporting Linux. If the software is free, then the only way RedHat and others can make money is through support charges. On a related note, lately Microsoft has been touting reports that show the overall total cost of ownership (TCO) is lower for Windows servers than for Linux servers, if you take into account all related costs. See how that falls right into line?
I've also heard that Linux has the best architecture, since thousands of people worldwide have looked over every line of code something like a bajillion times. Whether these arguments hold water or not remains to be seen, but there is this general perception. And on the other side, Microsoft of course claims the de facto standard on the desktop, and is way up there with servers too.
You could continue with fastest product and largest installed base or with unique functionality and quality of support, but the results would be more of the same. The interesting one is with ease of use versus most third party supporters. Here, Microsoft is covering both sides, product and market centric! What is the significance of that?
Well, first Microsoft realizes that ease of use is subject to increasing returns. As Joel points out, making software 10 percent easier to use doubles the number of potential users. And without supporters to fill in the gaps in your whole product offering, you're hosed. To really satisfy the pragmatists, a whole product has to deliver 100 percent. Most people will agree that Microsoft comes closer to delivering 100 percent than any other group does, especially Linux. Second, Microsoft has also been very big on supporting developers and ISVs. Developers and ISVs make the software that make people want to buy your operating system. And now we understand why. This will be the biggest hurdle for Linux to overcome. Without at least one of these factors, Linux will remain a niche player, used by techno-enthusiasts and Microsoft-haters like Scott McNealy.
Linux fans, please don't flame me. I'm not picking on Linux, even Giga says to hold off until 2005. The product just isn't mature enough yet. It's still focusing on itself (product-centric view) rather than on the business problems (market-centric view).